An Alternative to Aggro

[Flatfingers]

I'm on record as opposing the mindless cloning of the "aggro" mechanic into any MMORPG, and especially to unthinkingly inserting that tired mechanic into this particular game. In fact, I question the value of the aggro mechanic itself and its follow-on design features.

But I haven't really taken the time to suggest an alternative, which I think is a necessary element of constructive criticism. So this essay is an attempt to draft such an alternative. I don't think it's perfect. I just think it could be a workable starting point.

Comments, as always, are welcome.

...

"Aggro" in current MMORPGs works basically like this: when an NPC needs to choose which character should be attacked next from a group of player characters, it consults an internal list of "aggression" values. For each player character in the group, the attacking NPC calculates an aggression value based on various qualities of and/or actions by that PC. It then aims its next attack at the PC with the highest aggression value. That player character is then said to have "aggro'ed" the attacking NPC.

The natural outgrowth of this aggro concept is that players then try to do things to "manage" the aggro of attacking NPCs. The "tank" role comes first, because it's obvious that if players can control who gets aggro'ed, they'll want that aggro to stick to the character with the best defenses, leaving weaker characters unharmed and free to do other things like apply damage to the NPC or heal the tank. Group combat, then, becomes focused around aggro management -- carefully choosing and timing actions to do as much harm to the enemy or as much help to one's group as possible without shifting aggro away from the tank PC.

My question is: when did people start confusing "managing aggro" with having an interesting tactical combat experience? What in the world does "managing aggro" have to do with letting a group of players make intelligent and cooperative use of a rich set of environmental phenomena to achieve tactical superiority? How does the artificial and arbitrary gameplay of "aggro management" make any use whatsoever of the IP, the setting, on which a MMORPG is based? How is "performing actions intended to control the internal aggro calculation of an NPC" anything like "combat?"

If people think they like the aggro game, that's fine. People can like what they like. But the fact that some people like one particular solution to a game design question does not imply that it's the only possible solution. As gamers, we should be expecting game developers to look for more enjoyable solutions to game design questions, to try to create new and better solutions, not to merely clone mechanics that might work for some other gameworld. Importing the "aggro" mechanic from ground-based fantasy combat games into Star Trek Online's space game presents the appearance of laziness; it looks exactly like the lack of effort that other posters have been condemning when it comes to implementing full movement in space.

So instead of aggro, I'd like to propose an alternative approach to NPC combat decision-making, which for lack of a better name I'm calling "cultural tactics."

Cultural tactics would assign cultural qualities to every NPC (in avatar mode) and every ship's NPC commander (in space mode). All individual NPCs will belong to a primary culture. While some individual variation may be possible, those cultural qualities will tend to determine the choices that an individual NPC makes in any situation. Those choices may be about combat actions, or they may be about diplomatic actions or anything else the NPC is able to do; all possible forms of interaction with player characters would be produced by a goal-generating system whose rules would take as inputs the cultural attributes of the decision-making NPC, relevant aspects of the local environment (including perceptions of player character resources), and a desired goal state.

What's important to see here is what's not listed as an input to this decision-making system: player character actions. Getting good combat behavior out of an NPC actor does not require allowing players to directly manipulate that decision-making process. As for offering it because it provides gameplay, that's true... but when that gameplay takes over, shifting the attention of players from interacting with elements of the gameworld to the manipulation of arbitrary rules that have nothing whatsoever to do with combat, then if that mechanic isn't required, it doesn't need to be implemented.

Cultural tactics would allow NPCs to have an appropriate and interesting degree of autonomy. Instead of being the pawns of players in gameplay that distracts from the gameworld, NPCs whose actions are based on their cultural attributes would choose combat targets in a way that tells us something interesting about who they are. A mindlessly aggressive culture might just target the nearest ship. (So maneuvering into and out of an NPC enemy's range would be a viable combat tactic for groups of player characters up against ships commanded by members of such cultures.) An honor-culture NPC might always try to target and destroy the strongest (however that's defined) player ship; a ship commanded by an NPC from a victory-at-any-cost culture might seek to destroy the weakest ships first. A nasty pirate might go after the ship that appears to have the most/best weapons. A daring privateer could be culturally inclined to attack the ship that might carry the most interesting advanced technology. Members of an cybernetically enhanced culture that shares a hive-mind (you know who you are!) might simply attack randomly -- they're big enough not to care what the typical opponent looks like -- or they might look for whichever ship acted like the leader in order to disable the group's command hierarchy.

The point behind all these examples is to show that aggro is irrelevant. Aggro is not necessary for aggressor NPCs to be able to make interesting choices about whom to target. And getting rid of aggro serves the useful function of eliminating the bizarre focus of players on withholding their gameplay actions in order to avoid being hammered instantly into pulp by a vastly stronger foe, a condition that is unlikely in Star Trek Online (even against the Borg) because one-shot-kills in MMORPGs are not fun.

Without being forced to play the Aggro Management Game, players are free to engage in actual combat-relevant tactical decision-making: should I try to maneuver to my target's rear facing, or would it be better to try an alpha strike now? Can I use the particles in the nearby nebula in some interesting way? Is there something cool I can do with a tractor beam or another of my mighty starship's very cool systems right now instead of having to hold my fire because it might make an NPC mad at me?

In summary, if the aggro mechanic works for other games, fine, but it is not required for every game and IMO it is definitely not required for this one. It can be discarded with no loss, and considerable gain, since not having to withhold one's combat actions for fear of attracting attention lets players participate more frequently in the fun. It also means they don't have to have all of their actions squeezed into the subset considered appropriate by some developer for a particular and narrowly-defined combat role like "tank" or "crowd control," allowing players much more freedom to play the combat game the way that's most enjoyable to them. Roles are still possible; the beauty of getting rid of aggro is that those roles can then be defined in ways that make more sense for the setting of a particular MMORPG. And even without aggro, aggressor NPCs are fully capable of selecting their targets in fun and meaningful ways.

If all that is accepted, then yes, I find it disappointing that STO's designers appear to have cloned the aggro mechanic for this game. If they really believed it was necessary, that's a shame.

...

I'm under no illusion at this point that STO's producers will read this and think, "Say, you know, that Flatfingers has a point -- right, everybody stop what you're doing; we're going to re-do combat even if it means shipping four months later than planned!" I assume that the aggro mechanic and its tank/DPS/support handmaidens are in, period, end of story, kthxbye.

The point of proposing and explaining this alternative is not to try to change Cryptic's collective mind. It's simply to contribute to the conversation, and because it's fun to explore game design. I expect there'll be people who disagree strongly with what I've proposed here -- perhaps if they make the effort to explain, in a rational and constructive way, why they consider copying the aggro mechanic from other games into this particular game, it'll help people (maybe even me) gain a better appreciation for why STO's designers chose the NPC targeting mechanic they did.

Naturally, anyone who feels there might be some ideas of value in what I've suggested here are free to speak up, too. :)

To those who've given fair consideration to this design concept, even if you prefer not to comment, thanks. I appreciate it.

--Flatfingers

[jhem99]

Does this have anything to do with anger management?

[thefreshjedi]

Long read Flat...jeezus, are your fingers bleeding yet? But succinct as usual, and I appreciate your depth. But the problem is I'm worried about the younger, more restless in the group losing interest half-way through it all.

But I get what your saying. And I too can appreciate this. I am also tired of the same old rag in every MMO. And it would be nice to see a different approach. But let me postulate this: Because every MMO does use the same old rag, my fear is that since aggro management has become the staple by which all of these players are able to organize themself into a party. Aggro management seems to give something to hinge the group around. The tank holds aggro. The rogue stays behind and backstabs. The mage stays long-range and launches magic-missles. The druid helps off-tank, or supports the rogue. The healer can heal without interupting. It's cliche, but it works for a reason. It gives the group something to work with. The system that your proposing is more akin to what you find in an FPS environment. It's disorganized to a degree, but each class of character can identify with a role they can do effectively to get the job done. (Take Battelfield Bad Company for example). There you have everyone not only picking the class they associate with best. Be it sniper, support, assault, demolitions, etc. Everyone falls into something they can enjoy. But they still have to work together for the same goal. If loosely.

My only thing here is that the traditional MMO looks to ground everyone into a particular role for the sake of the party structure, achieving the same goals. The organization focus attempts gives everyone something to focus on together, and when run tightly, and organized well, there is far less a chance of the party to disolve as a result of frustrating losses at the hands of an epic NPC.

The point though with Star Trek, and you're right, is that you're dealing with a different element altogether. (Except for maybe the ground combat portion). So who knows what would be effective in this style of game.

-avery

[LordDave]

Why would STO need aggro?
Aggro management only comes into play when you have one very large enemy being attacked by many not so very large ones and this large one can only hit one guy at a time.

In STO it's likely that a target will be picked based on where it's hitting and how much damage it's doing. A borg ship, for example, would attack the defiant blasting it's weakened shields instead of the sovereign blasting it's good shields. It's also likely that the Borg ship would keep attacking the defiant until it couldn't, then switched to a new target to hit until it could attack the defiant again.
Or even just kept on the Defiant until someone else posed a greater threat.

But I suspect that in most assaults, enemy ships will pick a target and attack until it's dead, cloaks, runs away, or is deemed not an immediate threat. (ie. it won't chase down a medical ship while a defiant is blasting away it's aft section)

[CaptHiliner]

I believe you may be on to something here. I have had the pleasure of tanking for many years in ... and
have often wondered how to articulate my feelings reguarding aggro. This sums it up much better than
"could you please hit somebody else for a change?":eek:.

[SIMONLEV]

Hmm... so are u suggesting that the NPC's would use their cultural advantages to the best of their ability?

And that these advantages are randomly generated just like the mission creator?

For example if they have a +2 to hit med ships then they try anything to use that bonus efficiently.

I may be off the mark on this one.

[indigowhale345]

Well, the whole point of aggression mechanics is to get players to control the battle. I can agree its not the greatest or most interesting mechanic, but the need to control the battle can be really quite important depending on the game and the combat systems. I'll give a couple examples here.

Anarchy Online. These days, at the highest levels, the aggro game is largely irrelevant except in raiding situations. Most endgame characters that can get aggro can also take it. If the character is twinked out enough to do the kind of damage to get the attention, they probably also are twinked out similarly defensively. And all they have to do is live long enough till the next heal is cast. Of course in raid type situations, aggression management is very important, because if the tank loses aggro, then the mob is usually capable of killing most classes very quickly, which can lead to a domino effect and raidforce wipe.

Now look at Neverwinter Nights. Not truly an MMO, and not a game that actually has an aggro mechanic. Yet with many of the improved AI scripts that can be found for the game, you can make the AI mobs quite intelligent, as in zeroing in on the lowest HP characters that are trying not to be noticed in the back. Meanwhile they ignore the much deadlier fighters, and simply walk right past them to go kill the mages or rogues in a few hits. And the fighter types can't do anything about it in most cases, not even block the path of the monster. Its that utter lack of control over the battlefield that can make that game extremely frustrating for weaker characters, and leads to certain needs regardless of the class, like having high AC.

So with those examples, you kind of have two extremes. One where aggro exists but its not entirely important. Tanks have their roles become largely superfluous and unimportant, and everyone is just a different kind of damage dealer, some with more tricks than others. And another example where the AI does certain things 'intelligently', yet it becomes extremely frustrating because the players have zero control over the battle. The poor mages and rogues might as well wait for the fighters to finish killing the enemies and just leech the way some AI performs.


Now, to be sure, I agree with you that the aggro mechanic, as it is exists in most MMOs, is a dinosaur of a mechanic that I hope not to see in STO. However, the point I am making in return is that we still need a way to control the battlefield. It doesn't need to automatically succeed, but when we see a ton of ships go after the weakest link, there really should be something that the rest of the players should be able to do to either get their attention, defeat their assault, or protect the weakest ship.

Whether its an episode mission to protect a taco freighter, or just a raiding fleet, there needs to be ways to alter and control the battlefield, to manipulate the AI to some degree. Else I can imagine it can potentially get frustrating for weaker ship classes, or certain ways of speccing ships. And while you have a good start of the system, the point of aggression mechanics is not just AI targeting management techniques, but control. And you did not address the control part of the equation.

[drays917]

I like your idea a lot, it sounds like it makes a lot of sense. My only thought is, maybe you shouldn't necessarily know upon first meeting an NPC what its "cultural tactics" are. That is, if you knew every single time what the ship you were fighting was going to do, I don't see how it's an improvement over regular aggro management. That said, I do like your system because of the diversity in fighting styles it provides - something relatively new each battle.

I think Captain Bob is right too, that there does need to be some way (to some extent) to control the battle so that the poor Oberth doesn't die immediately every battle. But, I think that can be resolved on the players' end, in that it's up to you to have a contingency plan regardless of the enemy's attack pattern. Unless I misinterpreted you, in which case-my mistake.

Of course, I plan on PvPing it as much as possible, so it won't make much difference :)

[justynhuculak]

I like your idea a lot, it sounds like it makes a lot of sense. My only thought is, maybe you shouldn't necessarily know upon first meeting an NPC what its "cultural tactics" are.

Exactly. With well-established species, i.e. Klingons, you'll have some idea of what their tactics will be (aim for glory), but if you're coming across a less well-known species for the first time, you won't know what their tactics will be. The more you fight a particular enemy, the better you'll get to know them (perhaps their tactics could be recorded in a database that'll display vital info and such next to the ship when you scan it?).

Another great post Flat :)

[RookActual]

Flatfingers, I think you just broke your own record, but still haven't broken the forum like many of us think you are intending to do!

[Loekii]

Imo, Aggro is simply 'target' priority, in STO.

We are basically fighting 'military and combat trained' vessels.

Just as players set up a 'target priority' --- ie take out the healer first, etc -- so do NPCs.

I think the key is to create a better system that can be less subseptable to 'gaming', as well as adding some 'randomness' to the formula, to further skew predictability.

Ie, shoot the most damaged ship, focus fire, and generally operating how intelligent players would play, etc, instead of being 'Taunted' into attacking the obviously worst target in the group.

"Cultural Tactics" will just result in a 'cheat sheet' for players -- Oh, its a Romulan, they will attack this way, oh its a pirate, they will attack that way, etc. We need LESS PREDICTABILITY, not more.

[DanSeale]

Flat Fingers: Dude you know how to put pen on paper so to speak !

In simpler terms for me I really do hope that STO turns to something else besides aggro. If they use "aggro" then you maybe rest assured that other items similar will follow: ie "TANKS" ETC. I really don't want another WoW or EvE .. I really want a cool game that is Trek and "space" travel etc .. oriented.

Without getting side tracked on to other issues .. let's just say that I hope that STO finds an alternative to Aggro ...

good post BTW.

[SIMONLEV]

There are lots of ways to protect the little guy. Cryptic has mentioned extending shields and converting power

and im sure there are others they've mentioned but i dont know any more on hand. There may even be an

ability to combine these defensive effects with other players who knows

[billybob442]

Loekii some good points. Aggro is really just how the AI determines target priority. The problem is that aggro, as it's handled by most MMO's, makes for STUPID enemies who are easily manipulated into fighting the fight on the players terms. Of course some people love the aggro mechanic for this very reason, it demands little of the player and makes for very predictable battles. All you have to do is repeat the same basic process a few zillion times and your at level cap. Bleh!

So back to what Loekii was saying, we need less predictability not more. The idea of having different sets of targeting priorities is a good start. However which set a ship(s) uses should be semi-random. Of course certain strategies lend themselves to certain type of missions and enemies (someone out to assassinate an ambassador is likely to target the ambassador first even when his guards threaten him, Klingons are likely to attack the largest enemy ship they can reasonably kill because that brings the most glory... etc). So the initial attack may be somewhat predictable. The real "feather in a devs hat" would be to develop a system that switches between these priority systems in a semi-random fashion. Favoring one strategy over another based on circumstances while still not being totally predictable.

What this all boils down to is: You need a system where the players can influence the flow of battle without completely controlling it.

[Manx]

While I despise the aggro mechanic in all its forms, I have to agree with Captain Bob. Players need to have ways to exert some control over the flow of battle, or group tactics will have no role to play in STO; that means, at the very least, being able to manipulate the NPCs targeting priorities somehow.

I try to tell myself that there must be a way to do this without anything that could be mistaken for aggro management; but, really, what would you call a character that has some clever trick to draw enemy fire, if not 'a tank'? :(

[billybob442]

I try to tell myself that there must be a way to do this without anything that could be mistaken for aggro management; but, really, what would you call a character that has some clever trick to draw enemy fire, if not 'a tank'? :(

A target!! :D

It just hit me why the aggro system bugs me so. It's the exact opposite of what you really want to do in a combat. In real life a warrior wouldn't want to make his enemies more aggressive, instead he wants to suppress their aggression. That's why they call it suppression fire! Seriously though, that's the idea behind suppression fire. A machine gun pours lots of bullets on a location to either make the enemy loose his nerve and hide or drive him from that location. Though a capable enemy just displaces and tries to come at you from a new angle. You of course then coordinate with your buddy's suppression fire, driving the bad guy out of one place after another until you force the enemy into that quant little kill zone your team set up for just this occasion.

Normal aggro systems pulls the enemy to some point (we'll call it point A), which makes his motions very simplistic and predictable. Under this system, instead, you drive him away from point A. Of course he has the option of choosing any other point on the map except that one. Which point he picks is going to be a combination of his target priorities and a little randomness. So while you still have the ability to influence the flow of battle you would not be able to predict or control it.

Under this system your big warships wouldn't tank, they'd terrorize. And who doesn't like to do that? ;)

[dyvimtorm]

Hey Flat, superb text wall, I almost made it through that one :p J/k always interesting stuff, even if I am a bit too inattentive.

Basically, I agree I'd like the NPCs to manage combat priority in a more complex fashion than just a counter for aggro. Don't know it the devs can manage it, but a more interactive AAI would definitely raise the challenge level of the ships being faced. Or in that sense, more options for the computer to choose from in a combat.

[Urantia]

As alway Flatfingers well done...you have my vote.

[bridgeburner18]

If it is a stipulation that this dissertation be intricate and verbose, then I submit my application in concurrence with previous iterations to provide a level of sophistication, in a forum overrun with imitations, to simply assault your senses with a wall of incoherent text meant to illicit some form of appreciation that I too know big words.

[drays917]

If it is a stipulation that this dissertation be intricate and verbose, then I submit my application in concurrence with previous iterations to provide a level of sophistication, in a forum overrun with imitations, to simply assault your senses with a wall of incoherent text meant to illicit some form of appreciation that I too know big words.



...We are Pakled.

[Gerrard]

I am a firm believer that "chaos" in a battle is counterproductive. I think that a well organized group should be well rewarded for that. In the past "agro management" has been a viable solution. If there is ANY other way that works, I'm ALLLL for it. No matter what it is. So long as it rewards you for thinking and working together and not just "mobbing" in.

[Flatfingers]

Thanks for the comments, one and all.

I wanted to let a little time go by before responding so that everyone would have a chance to speak their mind. I hope a few comments now will be OK.

1. How can players control combat without aggro management?

Those who felt that a complete answer to "aggro management" as a system for PvE combat needed to include some discussion of how players could respond to cultural tactics are correct. While I would say I did talk about how players would control combat in a game that uses something other than "aggro," I didn't go into it in much detail. That weakened my proposal.

On the other hand, imagine how long my post might have been if I had really gotten into that.

But check out the paragraph I wrote that started with "Without being forced to play the Aggro Management Game ..." The alternative to players trying to control some magic number inside an NPC is for them to look around at the local environment and think about how they can use the systems of their ship to operate on local environmental phenomena to achieve tactical superiority.

I'd say that, as a gameplay system, that interaction with (in-game) physical reality comes a lot closer to letting players experience what's really interesting about combat than trying to control a magic "aggro" number. (Note that this is a statement from a game design perspective. Combat is "interesting" in other ways, but even I agree that STO should be a game, not a combat simulator.)

2. Aggro is simple and well-understood.

I agree on both counts. My argument is not that aggro is worthless; it is that aggro is just one possible mechanic for organizing PvE combat -- there's nothing holy or perfect about it.

IMO, aggro is not the best that game designers can do. It's important for the health of the MMORPG industry for designers to try other things that might turn out to be even more fun than aggro management. Less altruistically, it's also important to make sure that every mechanic for a particular game helps to achieve the desired player experience for that game... and I don't think "aggro management" is the best choice for how to design space PvE for groups in a Star Trek MMORPG.

3. The aggro mechanic creates easy-to-understand roles for party members in combat situations.

I agree with this as well. However, I'm not sure that it's necessary for there to be an explicit linkage between the method that NPCs use to select PC targets and the ways that players organize themselves in groups in PvE combat.

By breaking that linkage, game designers are free to establish other roles that might be more appropriate for the IP. In the case of Star Trek Online, that's exactly what I have in mind when I talk about wrapping STO's gameplay around Engineering, Tactical, and Science roles. (Here (http://forums.startrekonline.com/showthread.php?p=607263&highlight=Tactical#post607263), for example.)

I understand the value of roles. Many players interested in group combat appear to prefer that there be a very small number of very well-defined roles that spell out exactly what they're going to be expected to contribute to group combat. I don't think there's anything wrong with supporting that gameplay interest... as long as it doesn't crowd out other equally valid interests.

This is why I tend to favor systems that are more freeform (such as skill-based character advancement) but which clearly offer a short list of distinctive templates -- roles -- from which players can quickly select the kind of thing they want to focus on in combat. This allows players who what to know exactly what to do to just jump in and start contributing, while still permitting players who enjoy having more latitude in their action to come up with their own successful individual and group tactics. (This is the approach taken in the MMORPG I've been slowly designing.)

In the case of Star Trek Online, I favor a middle way between aggro-based roles and freeform-with-templates roles: namely, Engineering and Tactical and Science roles. These aren't compelled by the NPC's attack selection AI, leaving that AI free to be something other than aggro-based (like "cultural tactics"), but these roles are still highly appropriate for this gameworld based on Star Trek which always had clearly defined Science, Engineering and Tactical roles. (For Starfleet, anyway.)

I think this is an appropriate design for this particular game, but it's definitely a good question. I'm certainly open to other opinions on it.

4. Are "cultural tactics" specific bonuses against certain opponents?

I was actually thinking a little more broadly than that. A cultural attribute would be something more like "aggressive" or "xenophobic" or "manipulative." In a combat situation, NPCs would decide what to do based on how the elements of the local environment applied to their cultural attributes.

For example, a Romulan NPC would likely have the "manipulative" cultural attribute. Romulans don't take on opponents directly -- they lay traps, and traps within traps. Rather than just coming at you with all guns blazing like some Klingons might, a Romulan is more likely to find ways to use the local environment to initiate indirect attacks against you.

5. Wouldn't all NPCs with the same cultural attributes always use the same tactics? Boring!

I think they'd tend to have some similar tactics -- which is desirable -- but there'd still be some room for variation for a couple of reasons.

First, I did say "[w]hile some individual variation may be possible ...". NPCs wouldn't all be absolutely controlled by their cultural attributes; there'd be some individual randomness. (Think about the Romulan commander in TOS: "Balance of Terror" and how he argued with his subcommander over what the Empire expected.)

Also, different environments would lend themselves to different tactics. A Romulan NPC involved in a skirmish in a nebula might do one thing, while the same NPC in a mass fleet action in an asteroid field might do something different there.

But it would still be sneaky. :)

6. Unknown aliens would have unknown cultural tactics.

Yep. All just part of the fun.

Actually, one of the follow-up ideas I had was that every sector might have several hostile groups (xenophobic aliens, pirates, etc.) whose home locations are hidden. These aggressors would pop up from time to time, allowing players to get to know them through their battle tactics. Eventually players would be able to figure out from their patterns of attacks where their base is located. Then player fleets could, if they wanted, get together to assault the Bad Guys where they live once and for all.

At which point a new hidden group of Bad Guys would be spawned, with their own unique set of cultural attributes and hidden base(s).

Just a thought.

7. Too... many... words... eyes bleeding....

OK, guilty as charged... but seriously, you folks think this was long-winded?

Have you forgotten about or not seen the really hardcore insults to the 11,000-character limit that I've written?

Some terrifying examples related to Star Trek Online:

Starship Operations in a Star Trek MMORPG (http://flatfingers-theory.blogspot.com/2007/04/starship-operations-in-star-trek-mmorpg.html) [external link]
Combat Modes and Player Ranks in a Star Trek MMORPG (http://flatfingers-theory.blogspot.com/2007/04/combat-modes-and-player-ranks-in-star.html) [external link]
Comparative Rankings of Starships in Star Trek (http://flatfingers-theory.blogspot.com/2007/07/comparative-rankings-of-starships-in.html) [external link]
Sensors in a Star Trek MMORPG (http://flatfingers-theory.blogspot.com/2007/10/sensors-and-star-trek-online.html) [external link]
Proposed Questions for a Star Trek Online FAQ (http://forums.startrekonline.com/showthread.php?t=5854)
Technology Levels of New Worlds in Star Trek Online (http://forums.startrekonline.com/showthread.php?t=9180)
Survey: The Game of Galactic Exploration (http://forums.startrekonline.com/showthread.php?t=13123)
Who Is Exploration For? (http://forums.startrekonline.com/showthread.php?t=15238)
"Scotty Crafting": A Design Proposal (http://forums.startrekonline.com/showthread.php?t=16676)
Breeding Better NPC Starships (http://forums.startrekonline.com/showthread.php?t=18305)

:D

My writing tends to run long not because I'm trying to bury people with a flood of words -- that would be counterproductive -- but because I sincerely enjoy exploring ideas. I like seeing what's good and bad about an idea, so I throw out a lot of thoughts about it to see what helps to understand it and what doesn't, and to see what other thoughts those comments spark in other people who are part of the conversation.

I thought this particular essay was pretty direct by comparison with some of my posts. But as I've said before, I'll try to keep the massive walls of text down to a minimum.

(Hmm. I have said that before, haven't I?)

--Flatfingers

[wootage]

Well, the whole point of aggression mechanics is to get players to control the battle.

I see aggression mechanics as having everything to do with making the game easier to create for the devs. Bear in mind they have to plan experiences, and the more choices the mobs can make, the harder it is to do that. The established method all the way from EQ1 is the aforementioned hate list, since it's easy to predict what your mob will do in any situation.

Flat's suggestion is essentially a different kind of hate list, and that would work as well for the developers. Add some unique, tricky tactics such as NPC teamwork and support, and you've got a winning recipe for making PvE combat fun.

Even Darkfall's rather lame approach of tying the NPC actions to the weapons they have (archers kite, melee closes and then runs when too hurt, mobs switch weapons if they possess them) is a major upgrade from the dumb-doornail approach to NPC combat, and it's nowhere near what can be done.

Great sugg. Flat! I for one hope Cryptic does stop, drop, and rework the AI :D

[wootage]

While I despise the aggro mechanic in all its forms, I have to agree with Captain Bob. Players need to have ways to exert some control over the flow of battle, or group tactics will have no role to play in STO; that means, at the very least, being able to manipulate the NPCs targeting priorities somehow.

I try to tell myself that there must be a way to do this without anything that could be mistaken for aggro management; but, really, what would you call a character that has some clever trick to draw enemy fire, if not 'a tank'? :(

Question: why does a player need an artificial game mechanic (something that is not related to their equipment, possessions or other character-related factors, including in-character communications with NPCs) to control their opponent so that they can win? Isn't that the definition of an exploit?

Let's do a quick check and see how it fits:

Using pathing to place a mob where it gets stuck so you can kill it without any risk. Exploit.
Finding places where pathing will not let a mob reach you so you can kill it without risk. Exploit.
Using a bug to reset a mob's aggro to 0 so you can kill it without risk. Exploit.



Using an internal mob behavior mechanic to make a mob concentrate on a specific player regardless of other attackers, so that the best-defended player can draw the mobs attacks while those other attackers kill it without risk to themselves.

Sounds like an exploit to me.

And just because game developers have not only let players get away with it, but actively encouraged it by putting special player aggro-management abilities in the game, does not make it anything other than what it is.

Seriously, for the sake of having fun in games, we need to lose the crutch of "aggro management" and just play our ships and characters. Any enemy is going to size up their attackers by the simple factor of who is the biggest threat, and who has the biggest weakness to the NPC's attacks. It's just silly that a heavily armored, low damage warrior is able to taunt an NPC off of the weak, rapidly dying mage who is the actual greatest threat to the NPC, or prevent said NPC from going right for the most threatening and most vulnerable targets to cut down the potential threat to itself.

[carnagefiend]

It's just silly that a heavily armored, low damage warrior is able to taunt an NPC off of the weak, rapidly dying mage who is the actual greatest threat to the NPC, or prevent said NPC from going right for the most threatening and most vulnerable targets to cut down the potential threat to itself.

It's not that odd actually.

I can't imagine how many times I have to tell people to "KILL THE HEALERS" in World of Warcraft. And this applies to both AI and PvP...

Anyways, the only real way to manage this sort of game is to allow all ships to engage in all roles.

That's literally all there is to it.

And now for Penny Arcade's input on this discussion!

http://www.penny-arcade.com/images/2007/20070921.jpg

[Flatfingers]

just because game developers have not only let players get away with it, but actively encouraged it by putting special player aggro-management abilities in the game, does not make it anything other than what it is.

Seriously, for the sake of having fun in games, we need to lose the crutch of "aggro management" and just play our ships and characters.

That's pretty much the heart of my argument right there. Nicely put, wootage.

A word about /taunt: that one command/capability sums up everything that I perceive as bad about wrapping combat around the aggro management mechanic. The very idea that I can taunt any opponent into doing what I want instantaneously destroys any possibility that an enemy NPC can behave in an interestingly clever way.

Instead of allowing NPCs to vary according to their skill at adapting to their environment (pretty much the definition of intelligence), which allows the PvE experience to capture some of the challenge of PvP, /taunt turns every NPC into an easily-programmable robot:

do until NPC_hitpoints <= 0
{
if aggro < 250, /taunt
}

That's not NPC tactics. That's the elimination of NPC tactics.

Wouldn't it be more interesting -- i.e., more enjoyable as gameplay -- if game designers created NPCs that are different in more ways than just their hitpoints and available attacks?

...

I want to point out again that this point of view is not a personal attack against gamers who enjoy the aggro management mechanic. I enthusiastically support the existence of MMORPGs offering that approach to combat.

What I'm saying is that I think there are other, and better, alternatives to that mechanic, and that Star Trek Online's designers should explore those alternatives instead of just running with "aggro" because it was convenient and there's an arbitrary launch date looming.

If they've made that effort, if they truly gave serious design consideration to something other than aggro management as a basis for group PvE (especially in space), then that's all I can ask for.

Not all I could hope for, obviously, but so it goes.

--Flatfingers

[Loekii]

Simply, just continue to make AI more like Players.

There is 'Aggro' in PvP.

Players place their 'Aggro' based upon proven tactics. So NPCs/Mobs should do the same, rather than responding to 'parlor tricks'. If the 'trick' doesn't work in PVP, then it should not work in PVE. For example, how many people suddenly jump on the 'tank' when he /taunts them. None.

In STO, target priority just needs to be complex and not 'predictable'. It should be 'scripted' in such a way, that NPCs/Mobs basically act like players.

Moving in that direction, it reduces the exploitability of PvE combat, and actually keeps players on their toes.

As the game progresses, if players start to recognize patterns - CHANGE THE PROGRAM.

Overall, I would want to see a Romulan Patrol, basically fight differently and randomly -- but making logical choices -- just like how fights vary from PvP encounter to PVP encounter.

And on a side note - Games make PVE AI far too easy. There is value in 'losing' a good fight. It makes you a better player. I would rather get killed daily - even have to work at completing a simple Fed Ex mission, than cruise through PvE content like a Varsity Soccer player playing against a bunch of kids.

[Manx]

Question: why does a player need an artificial game mechanic (something that is not related to their equipment, possessions or other character-related factors, including in-character communications with NPCs) to control their opponent so that they can win? Isn't that the definition of an exploit?

Let's do a quick check and see how it fits:

Using pathing to place a mob where it gets stuck so you can kill it without any risk. Exploit.
Finding places where pathing will not let a mob reach you so you can kill it without risk. Exploit.
Using a bug to reset a mob's aggro to 0 so you can kill it without risk. Exploit.



Using an internal mob behavior mechanic to make a mob concentrate on a specific player regardless of other attackers, so that the best-defended player can draw the mobs attacks while those other attackers kill it without risk to themselves.

Sounds like an exploit to me.

And just because game developers have not only let players get away with it, but actively encouraged it by putting special player aggro-management abilities in the game, does not make it anything other than what it is.

Seriously, for the sake of having fun in games, we need to lose the crutch of "aggro management" and just play our ships and characters. Any enemy is going to size up their attackers by the simple factor of who is the biggest threat, and who has the biggest weakness to the NPC's attacks. It's just silly that a heavily armored, low damage warrior is able to taunt an NPC off of the weak, rapidly dying mage who is the actual greatest threat to the NPC, or prevent said NPC from going right for the most threatening and most vulnerable targets to cut down the potential threat to itself.

Even in real life battles, combatants try to manipulate their opponents. The whole point of tactics is to make battles more predictable; but since NPCs are not all that bright, they can't be tricked, provoked, scared, or surprised. It is necessary that some mechanic be in place to simulate these things; or the NPCs will just keep on mindlessly chasing down their assigned targets, regardless of what you do to them.

I'm not saying that Flat's idea is a bad one, I quite like it; but even with a system like he suggests, there still needs to be some way to influence the behaviour of the mobs.

Well... unless Cryptic has developed some awesome AI for STO (who knows?) :cool:

[SIMONLEV]

Wow! Flatfinger i just glimpsed at thoes links and looks like good reading material. I really respect how you are

displaying these ideas to the STO cause. My recommendation.. call up Cryptic and ask them what kind of

education or experience u need to work there. They always have suggestions ( in the dev chat) ,on people

who are interested in the field .. its weird how they do it so often

Also arnt u fearful of lurkers from other companies who could "acquire" ideas?

George Lucas has a pleasant face but Star Trek beats out Star Wars any day :eek:


Q:What came fist the chicken or the egg?
A: PvP

[Dr._Sskarno]

I tried reading your post on a Saturday at midnight,,,,but I'm sure it was a good idea...:)

[Spacemanspiff]

TLDR

/10 char

[Loekii]

Even in real life battles, combatants try to manipulate their opponents. The whole point of tactics is to make battles more predictable; but since NPCs are not all that bright, they can't be tricked, provoked, scared, or surprised. It is necessary that some mechanic be in place to simulate these things; or the NPCs will just keep on mindlessly chasing down their assigned targets, regardless of what you do to them.

I'm not saying that Flat's idea is a bad one, I quite like it; but even with a system like he suggests, there still needs to be some way to influence the behaviour of the mobs.

Well... unless Cryptic has developed some awesome AI for STO (who knows?) :cool:

Again, if a Player is not 'surprised', 'provoked', etc, why should they code the AI to stroke players egos?

AI should not be 'dumbed' down, so it is predictable or to force the AI to respond to 'parlor trick' tactics.

AI should be ramped up, so when a Player makes poor choices, they lose and the AI wins. Don't like losing? Then learn better tactics.

This is one of the reasons PvErs tend to get crushed in PVP. They grow accustom to not having to really think about tactics, and expect the other player to 'let them win'.


Two Warbirds, should fight wisely, rather than get 'baited' into attacking the wrong target, or get 'fooled' into making poor tactical moves -- aka being manipulated by the player.

Players should win PvE, by playing better than the Two Warbirds, rather than having the Warbirds coded to be 'manipulated' by the player. Thats a 'last place trophy for Bobbies team, so Bobbie doesn't cry' approach, which just helps Bobbie remain a bad player.

However, if Bobbie has to learn to play well, to win -- even if that means he gets beat a couple of times, Bobbie becomes a better player.

[Tekterra]

I think the suggestion from the OP was interesting. However, the system would make it too predictable for players. By experience, players will know exactly what the npc will do. With aggro system, players have to manage the aggro, but with such system sugggested by OP, AI will be too predictable which will make it too easy on the players. "Know your enemy, know yourself and you shall be victoriouse." If you know exactly what the NPC will do that's not a good thing for the game.

AI isn't to hard to program, but depends on what kind of game experience the developers want to bring to the players. Certainly, they can program a smart AI, and teach that AI the best tactics to use. For example, directing all firepower on a single player ship, start with the smallist ship in range. This simple tactic will make it a nightmare for players. Let me use this example here. A group of 5 player ships of various size fighting a group of NPC ships slightly greater in number. Normally, the players would be able to win by spreading aggro. However, when the AI decided to focus fire on the nearest ship, it's going to be hell for that player to survive the overwhelming firepower of all the NPC ships. Chances are the player ships will start to explode one after another. The point here is that the developers have to be carefull what to do. Just like other aspects of the game, it is all about balance. If you could create an unpredictable AI that can throw in some supprise for the players but not making the encounter too hard that will be great.

[Loekii]

I think the suggestion from the OP was interesting. However, the system would make it too predictable for players. By experience, players will know exactly what the npc will do. With aggro system, players have to manage the aggro, but with such system sugggested by OP, AI will be too predictable which will make it too easy on the players. "Know your enemy, know yourself and you shall be victoriouse." If you know exactly what the NPC will do that's not a good thing for the game.

AI isn't to hard to program, but depends on what kind of game experience the developers want to bring to the players. Certainly, they can program a smart AI, and teach that AI the best tactics to use. For example, directing all firepower on a single player ship, start with the smallist ship in range. This simple tactic will make it a nightmare for players. Let me use this example here. A group of 5 player ships of various size fighting a group of NPC ships slightly greater in number. Normally, the players would be able to win by spreading aggro. However, when the AI decided to focus fire on the nearest ship, it's going to be hell for that player to survive the overwhelming firepower of all the NPC ships. Chances are the player ships will start to explode one after another. The point here is that the developers have to be carefull what to do. Just like other aspects of the game, it is all about balance. If you could create an unpredictable AI that can throw in some supprise for the players but not making the encounter too hard that will be great.

I certainly prefer that approach to mindless tactics of 'tricking' the AI to attack the strongest ship, and ignore all the softer targets.

It also turns the game less into a 'tanking' game, but rather into a run and gun scenario, where players need to figure out ways to weaken the opponent before they are destroyed, instead of how to 'game' them into doing something stupid.

[billybob442]

A lot of people seem to have the impression that not using an aggro mechanic means that the player has NO way to impact the NPC's actions in battle. There's no reason for it to be that way. If the system is designed properly you can influence the battle without perfectly controlling it.

Flats, please correct me if I got the wrong idea here... If I understand the Flats correctly there would be a toolbox of tactics for NPC's to use. Each type of enemy would have a preferred way of choosing targets and a preference list of tactics to use against those targets. Each tactic would have some metric to measure if the tactic was working, such as the "suppression" metric I suggested earlier (patting self on back for finding a way to shamelessly plug my own idea) Player actions could influence the enemy by "breaking" the tactic., such as reacting in so that the enemy can't attack his preferred target, he's taking too much damage... etc. When this is tripped the enemy changes tactics, moving down his preference list in a semi-random manner (i.e. things higher in the list are more likely to be chosen over lower ones but will not be chosen automatically). In this way the player manipulates the enemy into abandoning a tactic.

The way this REALLY differs from the aggro mechanic is that under the aggro system your actions determine what the enemy is going to do, while under this system your actions determine what he's not going to do. Which of course still leaves him open to a whole list of other options.

Devs can control how this plays out by adjusting how many tactics are on an NPC's list and how sensitive the "tactic failure" system is for them. Dumb or ill-trained enemies will have very few tactics and not switch off of a tactic until he's effectively dead. Smart, experienced enemies will have dozens of tricks and change between them easily.

This avoids another problem that the aggro system creates. By allowing for "smart" enemies its possible to have a tough fight without creating bad guys that have increasingly more and more (and eventually insane levels of) health and damage.

[Phunix]

A few goods points raised in the thread, but in the end it's just all aggro and reputation in my simple brain (or more accurately, a bunch of numbers that need to be weighed for an outcome).
But yea I would like to see a bit more complex behaviour from NPCs, based on their race and position in their hierarchy. (and his rep, and your rep, and soforth)

In the end I don't care what the mechanism is called, as long as it feels Star Trek and right, and hopefully a bit challenging.

[billybob442]

A few goods points raised in the thread, but in the end it's just all aggro and reputation in my simple brain (or more accurately, a bunch of numbers that need to be weighed for an outcome).
But yea I would like to see a bit more complex behaviour from NPCs, based on their race and position in their hierarchy. (and his rep, and your rep, and soforth)

In the end I don't care what the mechanism is called, as long as it feels Star Trek and right, and hopefully a bit challenging.

In the end every concept used in a computer program has to boil down to a bunch of numbers that are weighed to determine an outcome. That's just the way computers work. Processors do simple math and basic "and / or" comparisons. If something can't be reduced to numbers then a computer can't operate on it.

The trick for a good combat system is creating one where the players can influence those numbers without making it so easily influenced that they can effectively pick what those numbers are going to be.

[Manx]

Again, if a Player is not 'surprised', 'provoked', etc, why should they code the AI to stroke players egos?

AI should not be 'dumbed' down, so it is predictable or to force the AI to respond to 'parlor trick' tactics.

AI should be ramped up, so when a Player makes poor choices, they lose and the AI wins. Don't like losing? Then learn better tactics.

This is one of the reasons PvErs tend to get crushed in PVP. They grow accustom to not having to really think about tactics, and expect the other player to 'let them win'.


Two Warbirds, should fight wisely, rather than get 'baited' into attacking the wrong target, or get 'fooled' into making poor tactical moves -- aka being manipulated by the player.

Players should win PvE, by playing better than the Two Warbirds, rather than having the Warbirds coded to be 'manipulated' by the player. Thats a 'last place trophy for Bobbies team, so Bobbie doesn't cry' approach, which just helps Bobbie remain a bad player.

However, if Bobbie has to learn to play well, to win -- even if that means he gets beat a couple of times, Bobbie becomes a better player.




But players can be surprised. We expect things, and those expectations can be thwarted; even by game characters. NPCs do not expect anything; they have their programed responses and that's it. They don't think about what the player is trying to do; nor do they try to formulate plans of their own. Tactics, beyond 'everyone shoot this guy', do not work on NPCs; how many strategy games came down to zerg fests just because of the stupid AI?

I'm not advocating the removal of tactics from NPC battles here; in fact I'm actually doing the opposite.

Again, I don't like the standard aggro mechanics either; but they create, at least the illusion, of tactical play. There would need to be something similar to avoid zergy gameplay.

[Peteromir]

The concept of tanking is not one thats alien to the real world, but its difficult to replicate without the use of "artificial" taunts etc. The problem lies in giving the heavy armoured slower da,age absorber type class the tools to do the job. In the real world they tend to work either by phisically blocking the way or by not being vastly damaging to remain engaged with, but extremely damaging/difficult to disengage from.

Aggro incially sprang up as a method of giving a NPC a target prority list. Purely on DPS, it works fine, it even works reasonably sensibly if you factor in healing as affecting the NPC's priorities. But the problem is that a player tooled up to take the damage is also likely to be doing less damage than one tooled up to do damage, thus artificaial taunting is nesary, even that has real world precident, but the level to which it becomes nessary in many games breaks it as a beliviable lie.

While a traditional aggro system based on each ships damage dealing/support using should not be comlpetely removed, it should be modified by other factors, and tactical considerations. Cultural beliefs should be among them, as well as how the ships are relitively armed, and set up. Theres still room for changing such things, but instead of taunts, false sensor readings, jamming beams etc can be used.

Tradtional MMO tanking, would not be beliviably reproduciable, but by such methods as shield extenstion, keeping a stronger shielded ship between you an the enemy etc much of its actual effects can be. If you put an enemy in a postion where he has too expose himself to a more dangerous ship than you, it achieves a similar effect. Rember also that when equivilent damage levels are talked about in such things tougher ships tend to have the bigger guns anyway.

Even if traditional aggro is in there, i doubt that traditional tanks will be, and aggro generational will primerially be a dPS thing.

[Loekii]

But players can be surprised. We expect things, and those expectations can be thwarted; even by game characters. NPCs do not expect anything; they have their programed responses and that's it. They don't think about what the player is trying to do; nor do they try to formulate plans of their own. Tactics, beyond 'everyone shoot this guy', do not work on NPCs; how many strategy games came down to zerg fests just because of the stupid AI?

I'm not advocating the removal of tactics from NPC battles here; in fact I'm actually doing the opposite.

Again, I don't like the standard aggro mechanics either; but they create, at least the illusion, of tactical play. There would need to be something similar to avoid zergy gameplay.

On a very simple level, lets look at computer chess.

Computers are programed to predict and respond to the moves.

They are not programed to be 'tricked', so that the player can manipulate the computer on the chess board.

If a player gets the upper hand, then there would be a form of control, but that is because they are making the correct decisions -- rather than 'kitting' the computer.


So applying that to STO, NPCs/Mobs should not be programed with 'achilles heels' or set up so players can tactically manipulate them to create the illusion of tactical gameplay.

Rather, you approach AI in STO, like how others have approached AI in Chess --- you program the NPCs to 'win', and it is up to the Players to out play them via tactics and proper choices (not because you can 'game' the program --- ie Kit the mob, and exploit the weakness of the program).

[Peteromir]

On a very simple level, lets look at computer chess.

Computers are programed to predict and respond to the moves.

They are not programed to be 'tricked', so that the player can manipulate the computer on the chess board.

If a player gets the upper hand, then there would be a form of control, but that is because they are making the correct decisions -- rather than 'kitting' the computer.


So applying that to STO, NPCs/Mobs should not be programed with 'achilles heels' or set up so players can tactically manipulate them to create the illusion of tactical gameplay.

Rather, you approach AI in STO, like how others have approached AI in Chess --- you program the NPCs to 'win', and it is up to the Players to out play them via tactics and proper choices (not because you can 'game' the program --- ie Kit the mob, and exploit the weakness of the program).

A computor chess program is playying the probabilities. Its not thinking or even simulaiting thinking, and its relitively easy to make a very hard computor chess program, especially if you have enough power spair. You can still trick them as long as you can out think the probabiltity routine, and know at certain critical junctions which way to push it. THe really famous ones cant be gamed that esially as they are running somany possible outcome trees they can more easially spot where you are going.

Why shouldn't things have achillies heals, many things do in real life? All a computor can do is either be set to be mindless, or give the illsion of thinking, this will always come down to some logic routine and number assement, the more complex, gemrally the better the illusion, but illusion it still is. The easiest way to mask the logic routines to hide them from the player is to introduce a random element, but you have to be carefull with that, as it can lead to bizairee happenings.

Tricking things is an important part of dealing with humans, things like klaying smoke on a particular vunrable approach to an enemy position, then attacking from a completly diffrernt angle....... On D-day German radar showed 2 naval fleets attacking, in truth the second, heading for the point the Germans thouught the logical place to attack was made up entirely of bomber aircraft flying in formation dropping tinfoil out of the window to fool them.

To get round the problem of creating better thinking AI many designers cheat and mearly allow harder AI more info than they should have, better reactions or accuracy than is managble through player input devices. Many "trick" tactics that would fool a human AI routines are often specifically programmed to be vunrable to, or else they wouldnt fall for it (the bogus smoke screen example above).

In short somewhere you have to start programming certain tricks and behavoir in or the AI would either be too reourse intensive or alternitavely unbelieviable and/or dumb. Vulcans would be beliviable in a pure logic routine, but klingons would require some artificail changes to make them seem klingon.

[Loekii]

I think you misunderstand what I mean by 'Trick'.

I am all for 'Cheating' in the AI's favor, to make up for the fact that it is just a program.

What I am against is intentionally programing a 'weakness' in the AI, that players can exploit to win.


For example, in WoW, having Boss Mobs AI, always run the same pattern, or always make them the same 'predictable' class, etc, is a bad thing imo. I would rather see 20 types of patterns, mixed with random class and ability asignments.

Another WoW example, in the Dread Mines, after killing some Boss Mobs, a "patrol" wanders in from behind --- every Time. It is predictable, rather than random. You never see the Patrol arrive 'early', or never arrive.

So I think that games should kick up the scale, and randomize the various scripts, so that encounters play out more 'individualistically', where they are more like 'new' encounters, rather than predictable encounters. Sure there is only so much you can do, but we are not even close to that limit currently.

For example, imagine your Away Team is trying to rescue a kidnapped crew member from a Syndicate Base. Instead of the guards always being laid out in the same location (3 in first room, 4 in the next, 1 patrol, etc), it should be mixed up, so that you might not find any guards in the first room, but encounter a patrol in the usually empty hallway.

[cavilier210]

I think this idea is intriguing, and would love to see it defined a little more on how it works. I myself found the aggro system of wow boring, and way to predictable

[billybob442]

The concept of tanking is not one thats alien to the real world, but its difficult to replicate without the use of "artificial" taunts etc. The problem lies in giving the heavy armoured slower damage absorber type class the tools to do the job. In the real world they tend to work either by phisically blocking the way or by not being vastly damaging to remain engaged with, but extremely damaging/difficult to disengage from.

I don't think that 'tank' is the right word for what goes on in real life. At no point does a modern ship, soldier or tank for that matter say, "OK, I'll stand here and soak up bullets while you guys flank the enemy." for the simple reason that bullets tend to kill people, and that goes for ships as well, the battleship Arizona was taken down with 1100+ on board by a single Japanese dive bomber. Even when they don't kill it's still gonna take weeks in a ship yard or hospital instead of a few seconds with a healer.

As you yourself pointed out, the 'tanking' that you're thinking of is the problem of disengagement. During a battle soldiers get behind cover and ships fall into formations that let them protect each other, all of which helps keep them alive. To disengage the soldier has to abandon his cover, ships have to break formation leaving them vulnerable. Throughout history more infantrymen have died trying to flee the enemy than fight him for this very reason. A large part of warfare is based around intimidating the enemy into abandoning his defensive position / formation so the real killing can begin. Usually this is done by concentrated fire on an area (i.e. maintaining a high potential DPS) even when it doesn't always find it's mark. The idea is to cause more fear than their training & discipline can handle, forcing them to break and leave formation.

]While a traditional aggro system based on each ships damage dealing/support using should not be comlpetely removed, it should be modified by other factors, and tactical considerations. Cultural beliefs should be among them, as well as how the ships are relitively armed, and set up. Theres still room for changing such things, but instead of taunts, false sensor readings, jamming beams etc can be used.

You're right. Adding to the traditional aggro mechanic would allow devs to leverage existing ideas easier and reduce the culture shock of jumping to an entirely new system. Seems to me battle involves a combination of two conflicting emotions - aggression and fear. The first is your good old timey aggro that determines how badly the bad guy wants to charge in. The latter would represent his willingness to make a run for it in hopes of living to fight another day. If player actions affected two different measurements that can rise or fall independently you get a far great range of possibilities than just stand off or charge into battle.

Sensor jamming, decoys, displacing your sensor image... etc. would all make ship battles tricky and a lot of fun! If the Feds could use them in PvP too then they could easily answer objections to Klingon cloaking being unbalanced. Just a side thought.

[Manx]

On a very simple level, lets look at computer chess.

Computers are programed to predict and respond to the moves.

They are not programed to be 'tricked', so that the player can manipulate the computer on the chess board.

If a player gets the upper hand, then there would be a form of control, but that is because they are making the correct decisions -- rather than 'kitting' the computer.


So applying that to STO, NPCs/Mobs should not be programed with 'achilles heels' or set up so players can tactically manipulate them to create the illusion of tactical gameplay.

Rather, you approach AI in STO, like how others have approached AI in Chess --- you program the NPCs to 'win', and it is up to the Players to out play them via tactics and proper choices (not because you can 'game' the program --- ie Kit the mob, and exploit the weakness of the program).

Oh I agree, that would be the ideal solution; I just suspect it would be asking a bit much of MMO AI is all. I would be perfectly happy were STO to prove me wrong though :)

[MoppyCGDaniels]

Why would STO need aggro?
Aggro management only comes into play when you have one very large enemy being attacked by many not so very large ones and this large one can only hit one guy at a time.

In STO it's likely that a target will be picked based on where it's hitting and how much damage it's doing. A borg ship, for example, would attack the defiant blasting it's weakened shields instead of the sovereign blasting it's good shields. It's also likely that the Borg ship would keep attacking the defiant until it couldn't, then switched to a new target to hit until it could attack the defiant again.
Or even just kept on the Defiant until someone else posed a greater threat.

But I suspect that in most assaults, enemy ships will pick a target and attack until it's dead, cloaks, runs away, or is deemed not an immediate threat. (ie. it won't chase down a medical ship while a defiant is blasting away it's aft section)

With the expansion of a certain planet in SWG did show the overwelming NPC endenemy average score in how well the PC had to be trainend for winning over a creature with other Jedis. For the opposite of aggro tank losses when a healer or science ship like a Deadalus is being taking apart, the times I was in a dogfight with the draining Tie-Fighters in orbit of Tattooine.

An alternative to aggro PvP (what does effect in random NPC killing) was my routine I followed in SWG: On my route from Corellia to Tattooine was some buisness to conduct. Especialy when my starfighter was being damaged, so that I had to repair it. All though me did happen by erasing a flag the AI did react by doesn't the same corps again defeating on ground. As for a ST version of PvP I'd say joined fleets for given oders from the Sector starbase with other SF officers on ships likely to destroy an Domionion installation on a Asteroid or a Cardessian orbital weaponsplattform would do the trick.

I'm refering also for ground combat or melee combat especialy with encountering Klingons who carry Meg'Lets for example on a ridge, or for that manner Gorn.

[Peteromir]

I don't think that 'tank' is the right word for what goes on in real life. At no point does a modern ship, soldier or tank for that matter say, "OK, I'll stand here and soak up bullets while you guys flank the enemy." for the simple reason that bullets tend to kill people, and that goes for ships as well, the battleship Arizona was taken down with 1100+ on board by a single Japanese dive bomber. Even when they don't kill it's still gonna take weeks in a ship yard or hospital instead of a few seconds with a healer.

As you yourself pointed out, the 'tanking' that you're thinking of is the problem of disengagement. During a battle soldiers get behind cover and ships fall into formations that let them protect each other, all of which helps keep them alive. To disengage the soldier has to abandon his cover, ships have to break formation leaving them vulnerable. Throughout history more infantrymen have died trying to flee the enemy than fight him for this very reason. A large part of warfare is based around intimidating the enemy into abandoning his defensive position / formation so the real killing can begin. Usually this is done by concentrated fire on an area (i.e. maintaining a high potential DPS) even when it doesn't always find it's mark. The idea is to cause more fear than their training & discipline can handle, forcing them to break and leave formation.



You're right. Adding to the traditional aggro mechanic would allow devs to leverage existing ideas easier and reduce the culture shock of jumping to an entirely new system. Seems to me battle involves a combination of two conflicting emotions - aggression and fear. The first is your good old timey aggro that determines how badly the bad guy wants to charge in. The latter would represent his willingness to make a run for it in hopes of living to fight another day. If player actions affected two different measurements that can rise or fall independently you get a far great range of possibilities than just stand off or charge into battle.

Sensor jamming, decoys, displacing your sensor image... etc. would all make ship battles tricky and a lot of fun! If the Feds could use them in PvP too then they could easily answer objections to Klingon cloaking being unbalanced. Just a side thought.


Abonadoning his position is not a terriably large part, more its using fire to keep him in it so much that he loses vision and the ability to fire on things.

There is still alot of what MMO tanking's main role that is persuading an enemy to direct his/her fire somewhere other than where its needed/ would be most benificial to said enemies aims. THe difference is whereas in an mmo it uses forced taunts to direct fire onto a duriable target, in modern warfare it tends to involve tricking an enemy as to whats the biggest threat, prefiriable to a location that has no troops, but if not into troops in hard cover or onto a target duriable enough to take the fire (say an attached armour unit).

In capital ship naval warefarer i dont suppose weight of fire suppression is a simialr tactic, more a deterrant. But damage absorbtion (through the usage of anti missile systems whether missile/gun based or electronic/physical countermeasure etc) is a comman tactic.



Against popular belief the aggro system does feature in peoples target priority system. Even taunts do affect people, but to a lesser extent in trained personel, and theres raerly the force taunt even in civlians once engaged. Its just people having an amazing amount of prosessing power (though less ability to use it in pure numbers calcs than a computer) and so have in effect a vastly more complex logic routine to work out whats the biggest threat, including not having to assign each variable a number (which is one of ther harder parts of producing a number based logic calc based on real world factors). They also have the ability to tell when they are getting the wrong answer, and thus either adapt their decsion process on the fly or overdie it entirely, something that most computor routines are unable to do at all.


Personally I'd have a multi component system.

Component 1 the major part is a logic routine, based on a military tactics manual for that race, witth a modifier for captains personality (randomly picked from a belll curve like distribution of the personalitie types for that race), with a small random element to lower predictibility). It takes into account the other components.

Component 2 Is percived DPS of the enemy craft (to take into account suuport powers and repairs etc, tables for how much each race consdiers these relitively may vary).

Component 3 is Incoming DPS on a particular ship, along with percifed fear inducing effects (interspace affects as an example)

Compenent 4 is Healing, Ive seperated this as a rotuine from incoming DPS as at some point the incoming DPS no matter what the healing may well scare off some races.

Coponent 5 is Predicted effect, or how much damage you can do to an enemy relitive to any repair/heals it can get.

In group situations senior AI can attempt to overide as long as fear not overtaken a ship, and comms are working.

[Jerosh_Skitari]

I have to say that the only problem I had with your original post Flatfingers was the language.

You defined "Aggro" in this manner:
"Aggro" in current MMORPGs works basically like this: when an NPC needs to choose which character should be attacked next from a group of player characters, it consults an internal list of "aggression" values. For each player character in the group, the attacking NPC calculates an aggression value based on various qualities of and/or actions by that PC. It then aims its next attack at the PC with the highest aggression value. That player character is then said to have "aggro'ed" the attacking NPC.

While your "cultural tactics" are a different way of assigning "aggression values," it isn't necessarily a different animal. I'm not in favor of "aggro" as we've seen it in other games, but I just want to make it clear that someone might take your "cultural tactics" idea and call it aggro.

Now, let's examine the post which (I assume) inspired you to make this thread:
In order to fire a weapon you have to have a clear line of sight. Anything with collision between you and your target will block LOS. A asteriod.. space debris.. other ships..

It's not dogfighting, combat moves slower than that, so theres no real way to fly in and block a shot. If you flew in and put yourself between your friend and the enemy. odds are the enemy would try to manuver until it gets a clear shot at your friend (all it would have to do would pitch up a for a second and it would then be above both of you and have LOS), or just aggro you instead. It's pretty straight forward :)
from this thread: http://forums.startrekonline.com/showthread.php?p=609174#post609174

Let's look at the situation and assume that the devs had read your mind (or your previous posts!) and had in place a system which you would call "cultural tactics." Specifically, let's examine the tactic outlay whereby the enemy ship targets the nearest player. By flying in between the enemy ship and the friendly ship, you've become the nearest player and you have thus "aggro'd" the enemy.

Or maybe the AI is smart enough to acknowledge that if the target it wants to fire on (based on traditional aggro calculations or some other variety of aggro calculations) isn't in LOS, then it will change its target.

My point is, just because Lt_Gozer used "aggro" in his post, it doesn't mean that it's aggro as we know it in other MMOs. I could have given other examples of how his statement could have fit within your proposal, but I think that one was enough. I'm not saying don't worry at all, because it could mean STO will have traditional aggro in it, but it's not necessarily the case.


And as a side not, after copy-pasting Lt_Gozer's post in here and running a spell check to make sure I didn't spell anything wrong... Lt_Gozer needs to learn how to spell :D

[Flatfingers]

While your "cultural tactics" are a different way of assigning "aggression values," it isn't necessarily a different animal. I'm not in favor of "aggro" as we've seen it in other games, but I just want to make it clear that someone might take your "cultural tactics" idea and call it aggro.

Oh, they did. :)

I thought my follow-up post addresed that (among other things), but the short version is that the "cultural tactics" mechanic differs from "aggro" in two important ways:

Cultural tactics don't expect players to directly manipulate a single "who I'm going to attack" number that's internal to an NPC.
Cultural tactics operate at a higher level of flexibility than aggro, allowing the environment to play a larger role in NPC decision-making.Let's look at the situation and assume that the devs had read your mind (or your previous posts!) and had in place a system which you would call "cultural tactics." Specifically, let's examine the tactic outlay whereby the enemy ship targets the nearest player. By flying in between the enemy ship and the friendly ship, you've become the nearest player and you have thus "aggro'd" the enemy.

Not necessarily. One additional rule (which I didn't mention in the initial high-level concept) could be that NPCs don't switch targets until their goal is satisfied. That goal might be "destroy the target" or "disable the target's engines/weapons," but it wouldn't make much sense to define NPC targeting to be interruptible by players other than the targeted player unless we want to see aggro management-like behaviors again.

And remember, "target the nearest player character's ship" is just one possible tactic, and probably the simplest. Other tactics could be less amenable to interruption by non-targeted player characters.

My point is, just because Lt_Gozer used "aggro" in his post, it doesn't mean that it's aggro as we know it in other MMOs. I could have given other examples of how his statement could have fit within your proposal, but I think that one was enough. I'm not saying don't worry at all, because it could mean STO will have traditional aggro in it, but it's not necessarily the case.

It's possible. That's why I said I was going to curl up with a good book -- it seems likely to me that Lt_Gozer was describing aggro in its usual sense, in which case (as I said) my interest in STO drops considerably, but I could be wrong so I decided to avoid commenting further rather than overreacting.

So yes, this thread was inspired in large part by Lt_Gozer's offhand comment. But I thought (and think) it's a sufficiently interesting topic to be worth discussing even if it winds up that STO doesn't implement the aggro mechanic in its usual form.

--Flatfingers

[cocoa-jin]

I agree whole heartly with Flatfinger's point. It is no longer acceptable for premium MMOs continue down the path of the same old models considering I can mindlessly do the same thing on other MMOs for free.

For $15 a month we should expect and demand more out of new developments. AI certainly has its limits, but I cant help but believe many of the constraints we are forced to deal with is due to lack of resource investment into modeling AI. Even worse, once modeled, AI needs to continoulsy updated to address player tactics and trends....mix it up, change AI tactics, dont just make more and more and stronger and stronger AI.

There is a wealth of resources and advisors on combat tactics that can be tapped in order to come up with interesting and engaging tactics for the AI. Yes, STO is in teh future, but its based strongly in combat from modern and recent history. The weapons and systems are different, but the relationships remain...intelligent use of these relationship models would apply remarkably well.

Basic knowledge of air combat, surface and submerged naval tactics on a single and group level would provide many AI tactical styles and options.

Lastly, allow AI to change tactics when the one its using is working.

[Awarkle]

I would hope to see unique style of agro managment, however Cryptic have not only got to appeal the game to the startrek crowd they also have to appeal it to the already established MMOrpg crowd.

if for instance you have your unique agro management tool that someone cant pick up and understand instantly. Likewise you cant have someone come from another game used to the agro mechanic and agro control and go. WTF ??

Yes its lame to have some kind of "tank" do the taunting while the "dps" does the damage while the "support" keeps everyone from dying. BUt unless you have a means which means the monster will activly attack the healers if they heal too much, the dps if they nuke too much and pay attenton to the hard as nails tank. Your going to end up with basically everyoen rolling tanks and playing hate pingpong OR having a healer chain heal till they have agro and ping agro between two healers (ive done that before)

The other issue is people for some reason HATE it when the mob is running all over the place they want it nice and stable and standing where its easily accesable and where it can take a pounding.

The problem will always be if you make the AI too smart people will tire of endless dying Or trying to put together a team that can drop it only to find the working solution doesnt work no more times.

Ifanyone remembers battelcruiser 3000 ad (yes it was buggy) it had an adaptive AI probbly one of the most advanced for any game. They discovered that the AI made the game impossible after a certain point. You couldnt keep up with it so the game ultimatly failed (although it was a poo game apart from the AI).

I think cryptic should make certain badies in sto hard the BORG should not be a walk over and they should adapt But you should be able to ultimatly kill them else everyone will stay clear of the super hard mobs for risk of loosing their ship.

[billybob442]

Abonadoning his position is not a terriably large part, more its using fire to keep him in it so much that he loses vision and the ability to fire on things.

There is still alot of what MMO tanking's main role that is persuading an enemy to direct his/her fire somewhere other than where its needed/ would be most benificial to said enemies aims. THe difference is whereas in an mmo it uses forced taunts to direct fire onto a duriable target, in modern warfare it tends to involve tricking an enemy as to whats the biggest threat, preferably to a location that has no troops, but if not into troops in hard cover or onto a target durable enough to take the fire (say an attached armour unit).

In capital ship naval warfare i don't suppose weight of fire suppression is a simialr tactic, more a deterrant. But damage absorbtion (through the usage of anti missile systems whether missile/gun based or electronic/physical countermeasure etc) is a comman tactic.


Against popular belief the aggro system does feature in peoples target priority system. Even taunts do affect people, but to a lesser extent in trained personel, and theres raerly the force taunt even in civlians once engaged. Its just people having an amazing amount of prosessing power (though less ability to use it in pure numbers calcs than a computer) and so have in effect a vastly more complex logic routine to work out whats the biggest threat, including not having to assign each variable a number (which is one of ther harder parts of producing a number based logic calc based on real world factors). They also have the ability to tell when they are getting the wrong answer, and thus either adapt their decsion process on the fly or override it entirely, something that most computor routines are unable to do at all.

Petromir, apparently I really misunderstood what you meant by 'tank'. I assumed you meant someone who endures repeated direct strikes while his team takes other actions (OK, a literal tank could do this vs. small arms but that's the exception not the rule). However I think we both share the idea of someone who engages the enemy so that he's not free to engage the target of his choice. Meaning 'tank' by those terms then yes, I agree with you that it happens in real life.

BTW - when I talked about breaking off attacks I was thinking mostly about traditional naval combat. On land the dominating force can pin his enemy down behind cover, which is what I understand suppression fire is for. At sea you don't have cover so if the enemy has a dominating volume of fire you have two options - disengage, maneuver and find an angle where he can't bring the same volume of fire to bear on you OR bring in more buddies so that now you dominate. Neither is easy to do which is why the navy prefers to kill it's enemies with missiles and air strikes from hundreds of miles away. ECM is great for reducing the enemies aim and communications & can play havoc with guided weapons but against direct fire it acts more like concealment than cover. Once a naval cannon fires there's not a lot that'll stop it so the best you can do is not be there when that shell comes back down.

OK, back on point. Your paragraph about the differences in people and computers points to the heart of the problem with most aggro systems. Enemies in MMO's virtually never rethink their actions. If the devs want to simulate fighting a "real" enemy then an NPC needs to have some way of knowing that his current actions just aren't working for him and it's time to try something different.

[Peteromir]

Petromir, apparently I really misunderstood what you meant by 'tank'. I assumed you meant someone who endures repeated direct strikes while his team takes other actions (OK, a literal tank could do this vs. small arms but that's the exception not the rule). However I think we both share the idea of someone who engages the enemy so that he's not free to engage the target of his choice. Meaning 'tank' by those terms then yes, I agree with you that it happens in real life.

BTW - when I talked about breaking off attacks I was thinking mostly about traditional naval combat. On land the dominating force can pin his enemy down behind cover, which is what I understand suppression fire is for. At sea you don't have cover so if the enemy has a dominating volume of fire you have two options - disengage, maneuver and find an angle where he can't bring the same volume of fire to bear on you OR bring in more buddies so that now you dominate. Neither is easy to do which is why the navy prefers to kill it's enemies with missiles and air strikes from hundreds of miles away. ECM is great for reducing the enemies aim and communications & can play havoc with guided weapons but against direct fire it acts more like concealment than cover. Once a naval cannon fires there's not a lot that'll stop it so the best you can do is not be there when that shell comes back down.

OK, back on point. Your paragraph about the differences in people and computers points to the heart of the problem with most aggro systems. Enemies in MMO's virtually never rethink their actions. If the devs want to simulate fighting a "real" enemy then an NPC needs to have some way of knowing that his current actions just aren't working for him and it's time to try something different.

I think the one you misunderstood was abit vauge and I didnt differentiate when I was using it in the more traditional MMO sense, where I directly compared it to older more medieval even ancient real world tactic (which really did work quite simiarly, a big block of infantry, either well armoured or wiyth enough bodies to absorb the damage) and the more modern equivilent, whichrarely involves it.

[wootage]

Even in real life battles, combatants try to manipulate their opponents. The whole point of tactics is to make battles more predictable; but since NPCs are not all that bright, they can't be tricked, provoked, scared, or surprised. It is necessary that some mechanic be in place to simulate these things; or the NPCs will just keep on mindlessly chasing down their assigned targets, regardless of what you do to them.

I'm not saying that Flat's idea is a bad one, I quite like it; but even with a system like he suggests, there still needs to be some way to influence the behaviour of the mobs.

Well... unless Cryptic has developed some awesome AI for STO (who knows?) :cool:

I agree. But there is an existing method of control for both PvP and PvE opponents. If I had to pick a single word, it would be "threat". Not the EQ2 misuse of the word in its taunt mechanic, but the true possibility of inflicting damage via your weapons and position. This is something that AI can take into account, since the computer knows all.

This can drive the opponents reaction in a realistic way - what short-range weaponed PvP opponent , faced with a ranged-weapon opponent, does not assess whether to immediately attack or change position to improve the odds based on those two factors - position (range) and damage potential? And vice versa, of course.

That's when Flat's cultural behavior would kick in. If a Romulan sees an attractive mechanism for a trap nearby, they would favor heading for it instead of an outright attack, unless they had a decisive advantage. A Klingon, not being the trappy type (unless they were at a decisive disadvantage) would be far more likely to accept the open battle and rely on maneuver and firepower - or just firepower - to give them the victory.

I like the way this is sounding. Hope the devs are reading, 'cause I think Flat has a winning idea here.

[wootage]

I would hope to see unique style of agro managment, however Cryptic have not only got to appeal the game to the startrek crowd they also have to appeal it to the already established MMOrpg crowd.

if for instance you have your unique agro management tool that someone cant pick up and understand instantly. Likewise you cant have someone come from another game used to the agro mechanic and agro control and go. WTF ??


Yes, actually, you can have people who know only the aggro mechanic come into a completely different game and learn how to play it. Very quickly too. I should know, I'm immortalized in more than one Planetside videos doing the most incredibly dumb things based on my previous (EQ1) MMORPG experience. Like trying to outheal the damage from some guy with a rifle, because as a medic I was a "healer". I have to LOL at myself for that.

However, not only did I learn (and fast - it only takes once to learn from any mistake), I learned how much fun it is to have the only decisive advantage be your decisionmaking abilities. When everyone's lifespan is measured in a few seconds when the shooting starts, suddenly your ability to decide, move and do your best no matter what - matters. A lot. And that made playing that game a whole new kind of fun.

But it's actually not that complicated. It boils down to these few rules:

Always be aware of your side's situation and your position in it.
Work with the capabilities of your immediate neighbors whether they're in your "group" or not,
Do your absolute best in every fight - button mashing only wins by total accident.
Stay alive as long as possible - no points for dying with ammo left ;)


Everyone can learn that, and that means everyone can have the kind of fun we used to in ol' Pside :D

[Flatfingers]

It boils down to these few rules:

Always be aware of your side's situation and your position in it.
Work with the capabilities of your immediate neighbors whether they're in your "group" or not,
Do your absolute best in every fight - button mashing only wins by total accident.
Stay alive as long as possible - no points for dying with ammo left ;)


Beauty.

All games should sound like it's that much fun to play them.

In particular, however, I'm a big fan of items # 1 and 2, which I condense into: always be aware of local resources and know how best to employ them to your advantage.

"Situation" and "capabilities" are both forms of resources. The terrain, your equipment (and your opponent's!), ambient environmental conditions, the abilities of your character and the characters of those working with you... all of these things are resources. The good tactician will be skilled at 1) identifying which resources are relevant to a given configuration of forces, and 2) recognizing how best to interact with those resources to one's advantage or to the disadvantage of the other guy.

(Side note: To see tactics as demonstrated by a master, read David Hackworth (http://en.wikipedia.org/wiki/David_Hackworth)'s autobiography About Face. I disagree emphatically with many of the political/strategic viewpoints which he shares in the latter part of the book, but as a tactician the guy was unquestionably a prodigy.)

Which is why I harp on "environment" so often. Without lots of stuff in the local environment to interact with, tactics devolves into little more than spamming "special moves" -- i.e., button-mashing to fire character abilities as soon as a timer expires.

That is not tactics.

Environmental richness is required for tactical opportunity. Here are some ideas I once suggested for both internal and external resources that could make Star Trek Online beaucoup tactical fun:

Internal capabilities at a tactical level are such things as your numbers, your weapons, your maneuverability, your repair/morale status, your consumables (ammunition, fuel), and any special abilities or assets possessed by yourself or other members of your unit.

External environmental features would be things like terrain (mountains, hills, peaks, ridges, defiles, nebulae, radiation zones), static cover (structures, trees, stars, planets, asteroids) and dynamic cover (vehicles, smoke, dust, haze, and weather such as rain, snow, fog, ion storms), and transportation lines (tracks, roads, highways, transwarp corridors).

With environmental richness -- perhaps as suggesed by Jerosh in his excellent Sensors and Scanning (http://forums.startrekonline.com/showthread.php?t=7852) thread -- and cool technological systems allowing us to use those resources (or deny them to our opponents), interesting tactical opportunities arise... such as "cultural tactics."

But now, having said all that to extol the virtues of designing combat in Star Trek Online to favor perception and creativity as well as aggressiveness, let me ask a question: how can such a system be designed so that it's fun for both hardcore grognards and newcomers to MMORPGs just looking for a fun bit of light conflict in the world of Star Trek?

In other words, suppose Cryptic were to implement an alternative to aggro, either in the form I've suggested or following some of the other suggestions in this and other threads: how should that system be structured to provide players (not characters, but players) with differing levels of tactical skill a level of challenge that will be fun for them?

It's to be expected that sectors in STO will work like "zones" in other games, where each sector has a general rating for its level of combat challenge and the hostile mobs in that sector will generally be easy or tough depending on the "level" of the zone. (I might be wrong about that, but if so it would be a pleasant surprise.) If that's the case, then that's one way of matching player ability to challenge level.

But what about the "cultural tactics" system considered on its own? Is there any way to design such a system so that it's possible to match an aggressor NPC's tactical abilities to the apparent abilities of a player?

--Flatfingers

[billybob442]

Flats... just an idea that came to me.

You determine a players tactical strength based off of character level (assuming more talented players will rise faster), or total hours played, or for a more involved system one could track a player's / character's progress in battles assigning him a "tactical score" based on a variety of measurements.

On the NPC side. You could assign a tactical rating for the computer baddies on top of their cultural factors. NPCs with higher scores may get access to more sophisticated tactical maneuvers and / or they may consider more environmental factors when they make their tactical choices. Additionally a better enemy may be quicker to change tactics if his current approach isn't working. Conversely a bad captain knows few tactics, fails to fully consider local terrain (only looks at the closest environmental factor?) and is slow to adapt if his tactics fail him.

So long as their are ways to quantify the tactical abilities of the player and NPC alike then it's a simple matter of adjusting the NPC's tactical abilities to fit the player.

More importantly. Do we want to adjust it this way? Should a boss be easier to fight because you're a bad player? Or should their be a single bar that all players are measured against? Or should it work one way sometimes and sometimes the other depending on the enemy?

[Flatfingers]

You determine a players tactical strength based off of character level (assuming more talented players will rise faster), or total hours played, or for a more involved system one could track a player's / character's progress in battles assigning him a "tactical score" based on a variety of measurements.

That would be one way. (Though by "level," do you mean rank?)

Another approach that occurs to me might be a combination of the tactical power of available tools (weapons and other technology that can affect opposing vessels or NPCs) and character skills associated with tactics. If characters gain skills by doing, then the number and power of skills a characters knows should be a good measure of a character's tactical strength. By also factoring in the tactical capabilities of a character's ship, I'd think that would wind up being a very reasonable approximation of how effective the player can be tactically.

If the game precalculated strength levels for both character tactical power and ship tactical power, a runtime calculation of overall tactical power should be relatively fast to perform.

More importantly. Do we want to adjust it this way? Should a boss be easier to fight because you're a bad player? Or should their be a single bar that all players are measured against? Or should it work one way sometimes and sometimes the other depending on the enemy?

I agree that that's an important question. In fact, it's so big that it goes well beyond the scope of this thread.

I'd enjoy seeing that question given its own thread. For this one, I'd sort of like to constrain the question of how to determine the challenge level of an opponent NPC to the specific concern of how much a player's skill in resolving tactical problems should affect his character's tactical performance. In other words, how predictable should an NPC using "cultual tactics" be?

...

Now, a couple other thoughts about "cultural tactics" come to mind.

One is the notion that, in addition to a race's or group's culture guiding their tactical choices, perhaps it should also govern what kinds of weapons they tend to use. (And of course the weapon effects then become yet another environmental effect, but let's not go there. ;))

A great example of this would be the "pseudo-plasma torpedo" (PPT) used by the Romulans in the venerable Star Fleet Battles board game. This was a weapon that could be fired at the Romulan player's discretion (for a much lower energy cost) from their standard torpedo tube. It looked like a real plasma torpedo, and moved like a real plasma torpedo, but when it hit, it did no damage.

That might sound pretty lame at first, but consider what it does to the opposing player's thought process. Now when a plasma torpedo is heading toward your ship, you have a choice to make. You could assume that the torpedo is real and fire at it to reduce its damage when it hits... but if it's just a PPT, then you've wasted phaser power you might need. Alternately, you could assume that it's just a PPT and let it hit you... but if it's a real plasma torpedo, it's probably powerful enough (especially at close range) to take down your shields or worse.

The pseudo-plasma torpedo is thus a kind of weapon that makes perfect sense for the Romulan culture, which (as I see it) is all about misdirection.

So what do you folks think about the idea of having culturally-appropriate weapons? It seems to me that having such culturally-appropriate weapons would support the use of culturally-appropriate tactics, but I'm curious to hear other views on that.

A second thought concerns minor and random races/groups. Although there are some things we might debate, with known and major races/groups in Star Trek, it's not too hard to imagine what their one or two most important cultural attributes might be.

What about minor races, whom we didn't see too much of in Star Trek? How should their cultural attributes (and perhaps weapons) be determined?

What about random races and groups that are automatically generated as we explore space in STO? What might a list of cultural attributes (which have tactical application) look like? How should those attributes be chosen? And what about weapons -- which weapons should be associated with particular cultural attributes?

Questions, questions. :)

--Flatfingers

[justynhuculak]

Bumped for awesome content.

[Loekii]

I would like to see ships operate on 'target of opportunity' type system, and tend to stick with their intial target unless forcably removed.

I have never believed in the 'Taunt' system.

I think a 'Tank' should work only because they are 'Blocking' a mob from getting to the softer targets, not because they 'insulted the mob'.

In 'ship/aircraft' tactics, Japaneses Zeros, did not stop attacking the bomber target, because a Corsair 'gave them the bird'.

The AI should be designed in such as way that is LETHAL to players, above all else. I don't care if a Romulan Warbird doesn't 'act' like a romulan Warbird, so long as it puts up a deadly fight.

The Borg should not have a 'culture tactic'. The Borg's cultural tactic is that they are simply BRUTAL. They should kick players around like we are 'greys' and they are the 'bored lv60 player'.

For the most part, I think AI should generally go after the closes target that they have the greatest chance of succeeding against. So if their is an almost crippled nebula in firing range, that should be the target -- not the Full Strength Sovereign, that the ship cannot destroy so long as the Nebula is assisting it.

[billybob442]

That would be one way. (Though by "level," do you mean rank?)

Another approach that occurs to me might be a combination of the tactical power of available tools (weapons and other technology that can affect opposing vessels or NPCs) and character skills associated with tactics. If characters gain skills by doing, then the number and power of skills a characters knows should be a good measure of a character's tactical strength. By also factoring in the tactical capabilities of a character's ship, I'd think that would wind up being a very reasonable approximation of how effective the player can be tactically.

...

Now, a couple other thoughts about "cultural tactics" come to mind.

One is the notion that, in addition to a race's or group's culture guiding their tactical choices, perhaps it should also govern what kinds of weapons they tend to use. (And of course the weapon effects then become yet another environmental effect, but let's not go there. ;))

A great example of this would be the "pseudo-plasma torpedo" (PPT) used by the Romulans in the venerable Star Fleet Battles board game. This was a weapon that could be fired at the Romulan player's discretion (for a much lower energy cost) from their standard torpedo tube. It looked like a real plasma torpedo, and moved like a real plasma torpedo, but when it hit, it did no damage.

That might sound pretty lame at first, but consider what it does to the opposing player's thought process. Now when a plasma torpedo is heading toward your ship, you have a choice to make. You could assume that the torpedo is real and fire at it to reduce its damage when it hits... but if it's just a PPT, then you've wasted phaser power you might need. Alternately, you could assume that it's just a PPT and let it hit you... but if it's a real plasma torpedo, it's probably powerful enough (especially at close range) to take down your shields or worse.

The pseudo-plasma torpedo is thus a kind of weapon that makes perfect sense for the Romulan culture, which (as I see it) is all about misdirection.

So what do you folks think about the idea of having culturally-appropriate weapons? It seems to me that having such culturally-appropriate weapons would support the use of culturally-appropriate tactics, but I'm curious to hear other views on that.

A second thought concerns minor and random races/groups. Although there are some things we might debate, with known and major races/groups in Star Trek, it's not too hard to imagine what their one or two most important cultural attributes might be.

What about minor races, whom we didn't see too much of in Star Trek? How should their cultural attributes (and perhaps weapons) be determined?

What about random races and groups that are automatically generated as we explore space in STO? What might a list of cultural attributes (which have tactical application) look like? How should those attributes be chosen? And what about weapons -- which weapons should be associated with particular cultural attributes?

Questions, questions. :)

--Flatfingers

I was assuming that a character is able to level up his various skills. So I meant a total of all of the characters tactical skill levels combined.

As for your two thoughts:
#1 : Starfleet Battles! I'd totally forgotten about that game. Cultural weapons are a brilliant idea and though cultural tactics may be a bit complicated to expect the devs to implement these are much simpler and could reasonably be (though not preferably) assigned on an encounter by encounter basis and / or (better yet) tossed into a later update for species wide usage.

I'm reminded of the scene in DS9 where Kira is describing the difference between the Federation phaser rifle and the Cardassian disruptor rifle. She points out that the phaser rifle is built with some of the most advanced tech in the quadrant, that you can adjust the fire length, beam spread, energy level, particle emissions, wavelength, phase, whether it's lethal or non-lethal... etc. with exacting precision - AND that it's a delicate instrument that must be finely tuned and many people get killed because they're to preoccupied trying to figure out the best setting. The Cardassian disruptor rifle on the other hand is made with decades old technology and has one setting - kill. However it can be drug through mud, used as a hammer or pry bar and still fires true.

This was an excellent way to point out the difference in the mindset between Starfleet and the Cardassian military. Starfleet wants one piece of equipment that can perform a thousand different jobs and is extremely customizable and are willing to pay the price in terms of reliability and the fact that you have to have an advanced degree in phaser theory just to fire the thing. Cardassians on the other hand want a tool that consistently kills, they want a hammer with a trigger on it, and they don't give a flip about any of that other stuff.


#2 : While it would reduce variety slightly, you could design a series of pre-determined traits (Aggressive, Sneaky, Honorable, Peaceful, Greedy... ) each with a series of behaviors attached to them. Each species would be assigned two or three of these traits in order of preference. That species' behaviors would come from a pool of the behaviors attached to its various traits. When deciding between two conflicting behaviors the species is more likely to choose the one connected to the higher preference trait. So an Aggressive Sneaky and a Sneaky Aggressive species will both have the same set of possible behaviors but they'll differ in how often each type of behavior is chosen.

You could even account for different personalities with this system. Say you want to use 3 traits. Set two of them to be species traits and the third to be a randomly chosen personal trait. Actually how much variation exists within a species could be shown through how these traits are ordered. Species with lots of variation between it's individuals would have the primary trait varying from person to person, while those with very little variation could use the tertiary as the personal trait.

In any event, once you have a list of traits and behaviors to pick from it would be fairly straight forward to randomly select traits off the list for automatically generated species and place them in a random order of preference.

[Mighty_BOB_cnc]

As a non-MMO-playing gamer who lives with 2 MMO-playing roommates (one a habitual WoW player with the top Horde DPS in his server, the other who plays any free MMO) the aggro mechanic as is currently exists (the old tired one) seems very unappealing and dull. I'm inclined to agree with most of the posts in this thread that state there needs to be less predictability (be it through randomization, cultural tactics/behaviors, priority trees, etc.) to spice things up. Anything that could be predictable and done over and over in PvE is pretty much a grind (as in the boring daily grind, like working on an assembly line in a factory) whether it has to do with leveling up or not. My brain is not a robot, I like variety.

[Peteromir]

How about this:

A Logic routine based on racial traits, captains personality, traditional DPS based percieved aggro/threat, incomming fire, relitive numbers etc is used to create a weighted table (simplified down to 6 but likely to be at least 12) that a random number genorator then picks from.

E.G You (a) and a smaller vessel (b) attack a klingon BoP.

table produced:

1. Frontal attack (a)
2. Frontal attack (a)
3. Strafe attack (a)
4. Frontal attack (b)
5. Strafe attack (b)
6. Ram (a)

This means that the AI routine is reacting a situation, taking into account many factors and can be predicted roughtly what a particular enemy is likely to do, but even if you are playing the correct numbers games, within the boundaries so a random crit string wont affect things, theres the chance the cpomputor will act differently.

Its important to add the exact routine is kept under wraps, and the table and roll on it are to be kept hidden (and addons accesing it are bannable). The traits etc can eb found out by scans, etc.

[Loekii]

I would like to see AI simply go after the 'logical' target.

If two Ddrex's encounter a Sov and a nebula, they should focus their attack on the smaller and weaker target (as it is better to end the fight against a Single Sov, than to try to fight a weaken Sov and undamaged Nebula).

It would be better, if the AI actually went to 'cripple' the Nebula first (wipping out its offensive and supportive cababilities) and then turned to concentrate on the Sov.

What I don't want to see is an AI, where the Miranda can 'Kite' the Romulans, while the Sov picks them off, etc. Or see the Sov 'Tank' (pulling Aggro). We are no longer talking about 'dumb beasts', but intelligent warriors, so their actions should reflect the better tactical decisions.

And again, I think that having 'random' reinforcements -- sort of like how the patrols showed up in Dreadmines in WoW (but not on a predictable time table) would add flavor to the game and make up for the weaker AI.

[wrussandrews]

What about the aggro from your significant other (or lack thereof) for playing too much STO?

[DanSeale]

hmmmm

folks this is one time I wish I understood some of the "tech-no-babble" better in order to be able to respond better. (one of the reasons I have not posted too much on this thread). I agree that something aside from the normal WoW style of "pulling" the mobs . with a "tank, healer, super zapper (Lock or other caster) and a couple of high DPS dudes (I like DPS my self) ... style of play is needed for STO.

hmmm

let me throw this out:

What if:
For instance play the AI you pulled was based upon the number and type of ships you had for each mission. I've actually seen this done on other Trek games. When your team pulled the mission ... the group was essentially scanned with a series of if/then commands. The individual mission would determin the size of that team .... AND the type of ships there. (In this case that would be determined by the number of players AND what tier ships they had. The AI that would "show up" for that individaul mission (or instance ) would match that group.

It can be done. Just currious if this makes any sense in helping with some of the varriables both from the "AI" stand point .. and consequently maybe even with a degree of varriance with the game play so that it is not always "exactly" the same.

Dunno if this makes sense .. Maybe Flatfingers or Loekii or someone can make better of this idea.

[Loekii]

What about the aggro from your significant other (or lack thereof) for playing too much STO?

That is what TIVO is for:


ON---->TITLE SEARCH--->REAL HOUSEWIVES OF...., TYRA BANKS SHOW... DESPERATE HOUSEWIVES....DANCING WITH THE STARS...---->RECORD SERIES
Plug in Gaming Headset
Have emergency number for Reservations at Firefly Restaurant in phone.


It beats any 'Tank' I have seen in any game. Heck it could pull aggro of Mage that has gone DPS crazy. :D

[wrussandrews]

That is what TIVO is for:


ON---->TITLE SEARCH--->REAL HOUSEWIVES OF...., TYRA BANKS SHOW... DESPERATE HOUSEWIVES....DANCING WITH THE STARS...---->RECORD SERIES
Plug in Gaming Headset
Have emergency number for Reservations at Firefly Restaurant in phone.


It beats any 'Tank' I have seen in any game. Heck it could pull aggro of Mage that has gone DPS crazy. :D

Experienced this have we?

Do not forget the flowers.

[Loekii]

Experienced this have we?

Do not forget the flowers.

Flowers can easily turn into a melee object -- or a ranged weapon if in a vase. :D

[Borlek]

Great opening post and discussion. I agree that aggro is an old and outdated form of NPC AI. Has anyone any idea of what they are considering using in STO? Any Dev leaks here?

[Tech151]

I was wondering... can someone post the cliff notes version to this essay?

[jhem99]

This is a sensible thread