Arcanaville

[ QUOTE ]
Without developer provided numbers, in easy to find places, all we get is a void filled by players taking best guesses or best reverse-analysis. There are some specialists (arcana, for example), who are very good at this. It is noteworthy, however, that she advocates for defense based sets, and I swear she has it in for /regen.

[/ QUOTE ]

Even if that were true, I'm not sure it would be especially noteworthy, unless you were of the mind to think I would lie about either the way tohit works, or the defense numbers I measure, in order to alter people's subjective opinions about them.

A word on my testing and "reverse-engineering" process, which I think is relevant to this discussion. Usually, I'll decide to test something or measure something, or reverse engineer the behavior of something, and then I'll ask a red name if they'll let me off the hook and just tell me what it is. There are usually three possibilities:

1. They give me a break and just tell me. Sometimes, its just a point of detail, and nothing especially secretive. In that case, I accept the number, and I virtually always test anyway. I then either post that I've measured the number with dev confirmation, or I reply to the devs that I cannot confirm their number. Sometimes, this actually turns up a real bug that eventually gets resolved, or at least confirmed.

2. They make me go the long way. Sometimes for whatever reason they don't tell me. Sometimes they don't reply at all. There are lots of reasons for this, but when it happens, I then proceed to test, and give them my test results, and ask for confirmation. Two sub-cases:

2a. They confirm, or refute, the answer. Sometimes I get the impression that they don't want to tell me, unless they know I could have gotten the answer without them. Sometimes they think it will take too long to explain to me how it works. Sometimes they aren't sure if its something that should be released to the players. In any case, if I already know, and its clear I can demonstrate that its theoretically possible for players to determine the information for themselves to high order precision, sometimes that causes them to change their minds, and tell me. Then, like the first case, I can post my results with dev-confirmation.

2b. Sometimes, I just never hear back at all. I suppose they have their reasons, but when this happens, I then post my results, and specifically state that I cannot get dev confirmation either way on them. Generally, when this happens, I dump a significant amount of my data onto the forums, as well as the test process, so that others can attempt to replicate (or not) my results, specifically because I have no independent confirmation of my results.


Usually, the red names are fairly open about discussing test results: it doesn't appear to me that they have a specific reason to keep most things secret. I think what they do *not* want to have happen is to get into a situation where the discussion of the game focuses on the numbers, including the enormous amount of time it can take to educate people on the proper way to interpret the numbers, along with the design intent behind them. This makes them reluctant to be seen as openly wanting to discuss the numerical aspects of the game.

The service I provide, if you want to call it that, is that I sometimes give the devs some cover to release numbers: they know I can measure them anyway, so they cannot be kept secret. They know I can describe the technique, so the work could theoretically be replicated. My results are generally accurate, so mine are almost as good as theirs anyway. I'm not going to burden them with a lot of argument or confusion over what the numbers mean. And if I post them, instead of them, then *I* take the heat over explaining them, rather than them, which is an enormous time saver for them.


I'm not specifically trying to justify the devs not releasing numbers or mechanics: I'm actually an advocate for full disclosure in general. But I recognize that there are very serious practical consequences of doing so that they are justifiably wary of. When the I7 critter accuracy scaler change went in, I posted several hundred separate answers to specific questions about it, most of them duplicates. Multiply that by all the game mechanics that players would have questions about if all the game mechanics were in a published rulebook, and that is a huge potential workload the devs would be creating for themselves.

[ QUOTE ]
[ QUOTE ]
This is blowing my mind. How can people quote a hit/miss ratio that doesn't add up to 100%?

[/ QUOTE ]

Because the orignal question wasn't asking for a hit/miss ratio? It was asking for hit percentage from attackers view and miss percentage of defenders view. These are two totally separate views with two totally separate expectations. Which is why they were separate questions within each category.

[/ QUOTE ]

Yes. And no.

[ QUOTE ]
Wasn't dull pain's +hp non-enhanceable pre-ED?

[/ QUOTE ]

Yes: I believe the ability to enhance the +health was added after ED was released: it was sometime around the I6 release.

[ QUOTE ]
A MoG scrapper however loses most of his hitpoints and rides at the low point for the duration of MoG. Healing does not work on a scrapper in MoG. While at this low amount of hit points, Blasters could defeat the scrapper with one well timed and well buffed Total Focus.

I am sure you are well aware of this already.

[/ QUOTE ]

You aren't factoring in the resistances of MoG. Unless you were playing a psionic blaster, you couldn't one-shot a MoGed scrapper. The amount of non-psionic shots it takes to defeat a MoGed scrapper will also defeat an Eluded scrapper if they happen in rapid succession. Except MoGed scrappers can use Dull Pain under MoG for +health (its only the heal that is blocked): it actually took 40% more damage to knock out MoG than Elude, barring the use of Aid Self or respites.

Not even with 30% unresistable damage could total focus at the 400% damage cap one-shot a MoGed scrapper under dull pain. It would take at least two, and two total focus hits at the 400% damage cap would also two-shot an SR scrapper under Elude.

I never saw it happen, and I do not think it's possible given the numbers.

[ QUOTE ]
Please don't take offense but did you spend a large amount of time on test prior to I4 being live or do you just read the numbers? I ask because I don't recall ever seeing you back then. Then again, I don't know your alts.

[/ QUOTE ]

I've fought perma-elude against perma-MoG. That should answer your question directly. And I still do not know what the tactical difference is between MoG and Elude you are referring to, but not stating. Both require high order tohit buffs to hit through, both have high burst damage as their weakness.

By the time they were made non-perma, though, the days of MoG being an overwhelmingly effective PvP power were strictly numbered.

[ QUOTE ]
Yes, Power Boost was nice to use but wasn't neccesary. Aim with 6 To-Hit Buff SOs + Build up with 3 To-Hit Buff SOs+ Total Focus Joust would do the trick against a MoGger. Not so the Perma Elude guys. It would hit and do the damage, but trying to follow up with another high damaging attack before they got out of range or behind cover was nigh impossible. Then the blaster's buffs would time out and the perma elude scrapper would go back to being unhittable.

[/ QUOTE ]

Prior to travel power suppression, Elude's run speed did not offer any significant jousting advantage over anyone else in PvP. You seem to be suggesting that the difference between MoG and Elude was that MoGed scrappers had lower health: that's not true: in relative terms MoGed scrappers had *higher* health, and were more difficult to kill. That's because the lower health washed against the resistance in MoG, but regen scrappers also had dull pain, and while the heal of DP was blocked by MoG, the +health of DP was not: you could permanently have higher relative health while under MoG, with well-slotted DP (you could even stack it for short periods).

They were really, really hard to kill. Especially those 6-slot membrane bastards.

[ QUOTE ]
Sit down and make a list of everything the Devs did right as far as pvp is concerned. Each power and how it relates to each other power in the game. Then compare that list to what you have above.

[/ QUOTE ]

Prior to when PvP arenas hit test? Ok:

1. Hitting another player with a melee attack did not cause you to explode.

2. Hurricane wasn't invisible.

3. Positron picked rad.


We're talking about design decisions here: the list of powers that happened not to cause serious problems by coincidence is not really relevant. The fact that flurry wasn't overpowered in PvP is not a point in the devs' favor when it comes to PvP balancing, unless there is specific evidence that flurry encompasses a specific design decision directly relevant to PvP. And it doesn't.


[ QUOTE ]
Even in the early days of I4 on test, a regen scrapper hitting MoG against any blaster with Energy Manipulation who was smart enough to take both Aim and Build up was a death sentence for said scrapper.

[/ QUOTE ]

They would have had to also be smart enough to take power boost, and know it boosted tohit buffs. Even then, not necessarily. Also, perma-MoG was at least as strong as perma-elude back then: MoG had higher defense, and was healing to full every 90 seconds or so: about as good as aid self, but with less interruptions.


Perma-Elude vs I3 Regen (non-MoG), now those were some interesting fights: I used to brawl-TK-brawl-TK at high speed, trying to knock out integration and have the stun kick in, then finish off fast, while they would try to land the one-two lucky shot before I could aid-self back to full.

[ QUOTE ]
[ QUOTE ]
They're both meatbags with massive hitpoints and regen.

[/ QUOTE ]

That's fine, but you were using an example of Babbage to reinforce your point. That Regen is critical to taking down AVs.

[/ QUOTE ]

Plus, does Babbage actually contain any meat?

[ QUOTE ]
I just get a kick out of several people claiming "This game wasn't designed with PvP in mind."

I don't think the Devs were/are that shortsighted.
The very first picture I saw of CoH was a screenshot on a gaming website about this new Superhero MMo coming out. The picture sowed 2 heroes attacking each other.
Also, don't you find it odd how closely most of the powersets performed in PvP when I4 hit test? Don't get me wrong, there were some obvious power differences with some sets, but other than Regen getting nerfed and the redo of how defense works, most other things have been "tweaks".

[/ QUOTE ]

There's a difference between "intended to have PvP" and "designed for PvP."

Off the top of my head, here are some of the critical design errors in CoH at I3 that needed to be balanced for when PvP was introduced:

1. Defense could lower tohit below zero, making accuracy function nonsensically. Intermediate floor added after arena combat demonstrated this.

2. Difference in resistive damage mitigation from low to high was very high to support PvE requirement of supporting different archetypal roles, but was too high for PvP combat. Unresistable damage added to lower the resistive gap in PvP without altering resistance relationships in PvE.

3. Stacking mechanics allowed teams of players to drive game mechanics beyond balanced-for limits. Problem never fully addressed.

4. Some powersets completely impossible to kill by almost any reasonable means, because damage mitigation was higher than all conceivable damage sources (except for large numbers of attackers). This specifically includes I3 MoG. Not fully addressed for all cases until I5 global defense reductions.

5. Tohit buffs too strong in PvP relative to PvE. Not fully addressed yet.

6. Inspirations unbalanced in PvP in seemingly random ways. Small lucks stronger than small insights. Large insights stronger than large lucks. Not currently addressed.

7. Imbalances sufficiently high for defensive protections in general to require game mechanical changes. Toggle-dropping introduced, and tweaked many times since then.

8. Mez vs Mez protection severely unbalanced in PvP. Addressed in multiple small ways, including the eventual introduction of breakfrees. Never fully addressed.


To believe that CoH was designed for PvP combat, and not just intended for PvP combat, I would have to see evidence that, for most powers and effects, there appeared to be clear evidence that the devs asked themselves "what will happen when a player turns this against another player?" I don't see evidence of that. There was no repel resistance until repel became ridiculously leveraged in PvP, a serious design omission - every effect should have had a counter effect in a game designed for players to attack players. Those kinds of omissions make me think that such thoughts of what would happen when power X was used by a player on another player were not carefully considered when powers were designed.

[ QUOTE ]
In other words, I shouldn't be asking the devs to nerf shivans, bombs and whatever else, I should be asking the playerbase to quit being so lazy.

[/ QUOTE ]

No, what I'm saying is that whether you nerf shivans and nukes or not, the LRSF is going to stay exactly the same. So if your issue is that the LRSF is poorly designed, you need to suggest altering its design. If your issue is that the people who use nukes and shivans shouldn't be allowed to benefit from do so, then suggest modifying nukes and shivans. But if you think reducing nukes and shivans is a way to affect the design of the LRSF, I believe you are mistaken.

[ QUOTE ]
[ QUOTE ]
This means certain builds are likely to do better, because they are more efficient at what the LRSF requires. but they were not specifically selected by the devs as such.

[/ QUOTE ]

Arcana, I agree with your assessment of how the LRSF was designed. Impossible until it was beaten with no regard for anything. I'll buy that.

And what's the end outcome? A screwball affair that lends itself to the use of shivans, nukes and ideal team makeups, screwing everyone that doesn't want to be involved in that kind of boring grind.

[/ QUOTE ]

Well, shivans add free damage, and free aoe damage at that. Even if they were cut to half the strength, every end game encounter designed to be very difficult and which can be made easier with higher levels of damage will automatically lend itself to encouraging the use of shivans. That's an unavoidable situation for any replenishable temp power that adds damage-dealing capability to a team. It isn't a specific issue with shivans, its a more general issue of the mere existence of content designed to be sufficiently difficult that not every randomly assembled team is going to have the same degree of difficulty in accomplishing it, or even necessarily a chance to accomplish it, without additional capability. Shivans are just one example of that; Vengeance itself is another.

The only way to not encourage the use of Shivans would be to design an encounter that didn't require high levels of damage. That's not easy, and even if its possible it would almost certainly create the "requirement" for some other odd addition.


When the LRSF was +5, and even to a very high degree when it was +4, I would tend to agree that it lended itself to "magic" team compositions. At +3, however, I do not believe that's true. There are issues associated with different archetypes, but highly sub-optimal teams can complete it at +3 with proper preparation and tactics, particularly with regard to the use of inspirations.

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
Get rid of the "I Win" buttons of Shivans and Warburg bombs and then you don't have to design missions with their use in mind.

[/ QUOTE ]

And the notion that they designed the RSF with either or both in mind is an assumption. Perhaps an obvious one, but one we've never seen validated.

And the fact that people do complete it without them reinforce the notion that they might not have considered them a critical component.

[/ QUOTE ]

Then try to complete it with no rads, stoners, vengeance, shivvies, or nukes. Oh, and have at least 1 dom and 1 stalker on the team. None of this "Oh, but these guys did it" cards people throw, I wanna hear of YOU doing it, with builds.

Call me when you're done.

[/ QUOTE ]

What would doing this prove, or contrawise failing this prove?

In any event, my personal experience tells me this is 100% achieveable, consistently, as described. Dark can tank as well as Stone with a reasonable set of inspirations, a sufficient number of cold and kin corruptors can substitute for radiation debuffs, and since you can do it with seven, you can certainly do it with a stalker and a dominator.

But to be honest, this is a seriously high hurdle to overcome just to prove the thesis that the LRSF was not balanced around shivans and nukes. Anyone familiar with how the LRSF was designed should know better: it wasn't balanced around *anything*. It was originally designed with the AVs at +5, where it was clear it was actually designed to be virtually impossible, and then scaled down until people started beating it, by any means. It wasn't *designed* to any standard at all, much less shivans and nukes.

Saying the LRSF was designed around nukes and shivans is like saying the exploding buildings in steel canyon were designed around stone tanks. They just hit really really hard, that's all. The LRSF is no different. It wasn't designed to be beaten with nukes and shivans, it was just designed to be really really hard. This means certain builds are likely to do better, because they are more efficient at what the LRSF requires. but they were not specifically selected by the devs as such.

[ QUOTE ]
Though your opinion is somewhat more vocal I'd have to say, overall, it's in the minority as far as CoX is concerned.

The "eight thousand blades of grass," with a few tougher foes (AV/GM's) thrown in for good measure, is more akin to people's vision of what is typical of heroes and villains while "100 tough villains that take minutes to kill," I'm assuming one-at-a-time, is more typical of every other MMORPG (at least the others I have played).

[/ QUOTE ]

I concede my opinion may be in the minority with regard to what people want to see in the game, but I disagree that most people think defeating a lot of minor things is more in keeping with a superhero concept. Do you think *any* comic book superhero has wracked up the number of defeats the average player takes to get to level 50? There probably aren't enough criminals in all of Gotham for Batman to have levelled to 50. In all the time I've played CoH, I've probably amassed a larger set of kills than Lobo.

[ QUOTE ]
I dunno Thorizdin, I can remember long fights off the top of my head since they have been so few.

Usually you get spiked within 30-40 seconds of entering an Arena map or zone, then you run or hope your support can keep you alive.

I would get a lifetime, yearly sub to this game if fights happened where it took close to the length of a 10 minute match for one kill to be scored. Or if fights were actually two people squaring off and pounding on each other until you at the keyboard got exhausted or it just got so boring you flipped a coin or something lol.


Unfortunately, thanks to all the unresisted stuff there isn't much that could lead to that unless you're playing an Ice/EM tank or something similar. :/

[/ QUOTE ]

I don't want long and boring (no one's health bar is even moving), I want long and interesting, as in at the five minute mark, its still not certain who's going to win at the ten minute mark, but you know one of you will by then (and ten minutes is probably excessive except in an arena match where both players actually explicitly want such a fight: I was thinking on a timescale of 90-180 seconds, enough to defuse the alpha strike to a moderate advantage, not an overwhelming one).

[ QUOTE ]
If they slowed down combat or PvP, I'd quit as well.

[/ QUOTE ]

I don't threaten to quit, and I don't claim to represent anyone's opinion but mine in this regard, but if MUO has combat that lasts more than three seconds, that will be very attractive to me. I would rather combat be designed like a poker game than a hand of blackjack, which is what CoH combat currently is.

My longest running issue with CoH is that I would rather kill 100 tough villains that take minutes to kill and then level, than mow down eight thousand blades of grass.

[ QUOTE ]
First off, this post is *NOT* a precursor of any particular impending change. The topic is strictly informative and for discussion.

I'd like to know about breakpoints for To Hit and Defense and how you, the players, think it should work. I'm not talking about mechanics -- I'm talking about the expectations you have in a fight.

1) You have the default To Hit value (ie no buffs), your target has no Defense value. How often do you WANT to hit him? Conversely, as the defender, how often do you expect to be missed?
2) You have the maximum possible To Hit value, and your target has no defense value. How often do you WANT to hit him? Conversely, as the defender, how often do you expect to be missed?
3) You have the default To Hit value, and your target has the maximum possible defense value. How often do you WANT to hit him? Conversely, as the defender, how often do you expect to be missed?
4) You have the maximum possible To Hit value, and your target has the maximum possible defense value. How often do you WANT to hit him? Conversely, as the defender, how often do you expect to be missed?

[/ QUOTE ]

To answer these questions in the way you want the answers, I have to remove the loaded terms in them that mean something technical to me. I'm therefore going to paraphrase:

1. Given that you've taken no steps to improve your ability to hit your target, and your target has taken no steps to avoid being hit, I expect to hit the target 50% of the time in PvP, and not just because of any specific game mechanical technical issue, but because 50/50 makes sense in terms of player vs player interaction: the chance to hit is equal to the chance to miss. In PvE, this chance can be higher: I have no specific expectation in PvE, except that if its different from PvP, it should probably be higher and not lower.

2. If I've done everything possible to improve my ability to hit the target, and the target has done nothing to reduce that ability, I should hit the target at the maximum possible rate that the game allows. In CoH that would be 95%, but there's nothing special about that number.

3. If I've done nothing to improve my chances to hit, and the target has done everything possible to cause me to miss, I expect to miss the target at the highest possible miss rate, which in CoH is 5%, but there is nothing specifically significant about that number.

4. If I've done everything possible to improve my chances to hit, and the target has done everything possible to reduce my chances to miss, I expect to be in the same situation as #1.

It goes without saying that I consider the above statements symmetric when I'm the target as when I'm the attacker.


I'll add something for reference: I consider the criteria above necessary but not sufficient for balance. You already know what I think is necessary and sufficient for balance, and I've already proposed a couple of ways to do it, including my personal favorite.

One more observation: given the current game mechanics, #4 is currently an impossible criteria to meet, and its at the core of how I want to fix it.


Edit: on retrospect, I need to add that these boundary conditions are not really the issue, because the real issues occur within the boundaries, where the situation becomes more complex. In fact, as stated, the questions have a linear implication to them that is actually part of the problem. The real question is this: for *any* value of defense, what should the *incremental* effort be to increase or decrease the ability to hit for the attacker and the defender. Its a short question, but it encapsulates literally millions of words of debatable tohit issues.

[ QUOTE ]
The AVs are linked. Even if you DO pull one and only ONE, when that one dies, 2 more will aggro immediately.

[/ QUOTE ]

On two separate STF runs I was on, this did not happen. I don't think they are linked, I think they have very large aggro/alarm radii. But if you pull them far enough apart, and kill them quickly enough, its often the case the others don't notice.

[ QUOTE ]
An idea for trying to plum the depths of Pool B.

This is based on something I discovered yesterday by accident. When you mission-drop a door mission, you get the same chance for getting a Pool B drop as you get from normal mission completion.

1) Have several slots open on Test. The more, the better. Remember how many you have.
2) Select a character from live who has an active mission, has not dropped in the last 7 days, and has easy access to the mission's contact (either Cell active or standing nearby.)
3) Copy that character over to test as many times as you have slots open.
4) Begin to log in the different copies, one by one, dropping the selected mission with each.
5) Record whatever Pool B drops you get.
6) As you get close to running out of copies, start deleting them and repeat from step 2 above.

In general, I would think that this would allow for many more mission completions than would be possible in normal play. Really only of use for seeing what is in Pool B, but it would be quite useful for that.

[/ QUOTE ]

My best speed for blitzing heroic police band missions is about 17/hour. Counting login and logout times, the theoretical best possible pace using this method is about 60/hour, assuming you have a relatively fast machine. That's certainly better, but I didn't think it was worth it to basically fill the transfer queue with hundreds of copies of me to get a three times better execution rate on Pool B when it was clear even a three times better rate wasn't going to find all the Pool B drops in a reasonable amount of time, which is why I didn't switch to this method to test Pool B drop percentage rates. My original submission of about 20 or so Pool B recipes when this thread's beta board antecedent was first created was the culmination of over three hundred missions. That's five hours minimum of sustained monotonous copy/run/delete cycles, assuming the character copy tool was running practically instantaneously.

I'm not exactly trying to discourage you from trying this, unless the character copy tool suddenly becomes ten times slower, but I think it will drive you crazy before you hit the two hundredth mission drop.

[ QUOTE ]
[ QUOTE ]
What matters is not direct numerical equality in PvP, but rather roughly equal opportunity for wins and losses. It matters less how badly you lose, as long as you have an equal chance to actually win.

[/ QUOTE ]
Sorry to cut out most of this post, but this statement stood out to me. This shows exactly why I believe Stalkers and Blasters are, in fact, balanced. Whenever my stalker duels a blaster, 90% of the time one of us dies in the first five seconds. (If they are particularly skilled and I can time the placates right, we can stalemate for a long time.)

[/ QUOTE ]

Ignoring the issue of placate mechanics for a moment, while this is not a balance-issue per se, one of the things that I think is problematic in PvP is the time frame of the combat. PvE is balanced around very rapid combat, and because there really isn't any PvP combat adjustments that attempt to alter that, PvP combat is intrinsicly designed around very fast combat also. This doesn't mean you can't have very long fights occasionally, but the average fight probably lasts just seconds before someone is dead or escaping. That makes it difficult to give PvP a tactical feel in my opinion: its over too quickly for tactical decisions to have any real chance to take strategic effect: a tactical decision has to have immediate reward or its likely to be a waste of time.

There are a couple of problems with slowing down PvP combat, though, although most are solvable. Rebalancing for a slower pace is actually not that difficult: it requires basic transform mathematics: calculate immortality lines for everyone, and then descale them. And actually, it really comes down to health, healing, damage and regeneration, and possibly endurance (everything else is transform-invariant under time). Slow down healing and lower damage, and you end up with fights that are exactly the same as now, but slower.

The real problem is travel powers: slower fights not only create tactical opportunities, it can make it impossible to kill anyone if everyone can escape combat. Two things would have to happen to allow for slower and more move-countermove combat: first, travel powers have to be somehow largely negated as major factors in determining PvP pace. Second, in its place, the game engine needs to have a better way to allow for range effects to affect combat, so that slow and lower magnitude movement can still have an effect on combat beyond the obvious - getting out of range. In essence, replace tactical high speed movement with tactical (relatively) low speed movement. This keeps combat more confined, and less jousty and hit and run: more dogfight, and less Red Baron.

How to do this, especially in the case of travel powers, is a bit sticky. There are two ways you can do it. First, you can give travel powers disadvantages in PvP. The devs were probably on the better track when they had flight have a tohit debuff. Each travel power could have an associated debuff that was irrelevant when actually travelling, but was a penalty when running travel during actual fighting. The problem is that its easy to overbuff a penalty away in teams, so this would have to be done very deftly, and probably requires also looking at stacking and accumulated buff mechanics (my #1 game mechanical pet peeve).

The other thing you can do is create more ways in which maximum effectiveness tends to automatically preclude travel, aka granite. Granite, in fact, is a really good example of a PvP-friendly combat stance-oriented power. Which is not to say its balanced or not balanced, but thats a question of numbers: its the mechanics of the power that make it PvP-friendly. It trades damage for defense, and when it makes you able to take more damage, it also makes you less able to run away from a competent set of attackers that bring enough damage to take you down (the exception, of course, is teleport, but that could be solved as well: give granite a very large self range debuff).

[ QUOTE ]
Ahh...I wasn't talking about this game. This one was pretty clearly balanced around a mathematical ideal of "effectiveness" of some sort given the number of similar powers/sets that are variations of each other with tweaks to certain values. Claws versus Spines seems to be a good example, with Spines having higher burst damage to go with higher end cost and slower recharge, both produce X amount of DPS over time but have drastically different outcomes where burst damage is important or factors (like buffs) strongly impact recharge rates and end recovery. I'd be willing to bet that there is (or at least was) a master spreadsheet with each power set totaling so they matched each other to a certain degree.

[/ QUOTE ]

Unlikely. What we know about how the devs balanced this game, in terms of attack sets, is that they performed, in effect, three separate balancing "passes" on attack sets.

In the first "pass" they designed the attack powers around a balancing formula: two, in fact:

(D - 0.36)/0.16 = R
5.2 * D = E

Where D is the damage of the power in damage scale units (1 DS = 100/36 BI), R is the power's recharge, and E is the endurance cost of the power. AoEs have extra formulas that specify how the AoE power will relate to single target powers. There are exceptions, but every power is balanced around these formulas. Implicit in these formulas is an unspoken assumption, that is hinted at by geko when he talked about them: a reasonable starting point to balance attack sets is to balance each power based on the amount of damage per unslotted recharge cycle. The formulas above add an extra wrinkle: they presume that higher damage attacks are worth more, and therefore are made somewhat more expensive in terms of their availability (thus, the slope of the first equation).

But neither equation factors in activation times, and that's a critical error: for attack powers, activation times are actually a sizeable percentage of the recharge time, and therefore the total cycle time. Worse: fast attacks in terms of activation are often given lower damage. You might think that makes sense, but their formula creates a problem: remember: it makes lower damage attacks less costly in terms of time, which is another way of saying they do more damage per recharge second.

But they can also do more damage per activation second, because they are fast. So the first formula above creates the dangerous situation that fast-activating attacks can also be made more efficient by the attack formula. I.e. shadow punch.

But with some exceptions (especially attacks the devs interpreted as being primarily mez and secondarily damage, like stun, cobra strike, etc) that's how the powers were balanced. And there are very few exceptions besides the mez exception. This actually says something: this says the individual power sets can only have equivalent damage output by coincidence, not by design. That's because the formula clearly isn't balanced, and all the attacks follow it. The only way you can balance the attack sets in terms of damage that is left is adjusting activation times or performing very careful engineering of the sets, and they didn't appear to do that. The exception: Claws, which Castle specifically tweaked relatively recently (i.e. long after release).

The second "pass" distributed secondary effects to all the attacks, like stun, knockdown, immobilize, debuffs, etc. These were clearly and admittedly handed out relatively arbitrarily: they were "balanced" in terms of subjective criteria, but not a mathematical one. There's no rule that says "X debuff = Y% knockdown."

The third "pass" was looking at the sets to make sure their overall utility was "reasonable." This is also highly subjective, but it has admittedly happened. MA was changed after release in I1 specifically because of a perceived lack of overall utility and effectiveness. I'm pretty sure there wasn't any specific numerical metric they were trying to hit with their adjustments, though. Why do I think that? Because the original changes accidentally "overbuffed" thunder kick with more damage than the damage formula above states it should have. In I6, specifically because that was detected, TK was adjusted downward. But if TK was adjusted to the correct value in I1 because of a balancing metric, it shouldn't have been changed in I6. It was, because there wasn't any specific balance metric to keep it at that level, but there was a formula that claimed it was set wrong.

I know the devs look at things like burst damage potential, and sustainable damage potential, but I also disagree strongly with their way of looking at those two quantities: I believe they are methodologically flawed. Somewhere, I'm sure there is a spreadsheet that says MA has this much damage and this much burst, and Claws has that much damage and that much burst, but I think those two terms are not defined in ways that properly represent in-game damage-dealing reality. Moreover, even if those numbers were out of whack, they don't seem to be able to influence power numbers, because the two formulas above seem to be almost pre-eminent: even if set A had less burst damage than it "should" the simple solution of increasing the damage of an attack or two can't be used, because it would violate the formula. That's not me saying it, its what I've been told when actually suggesting changes that violate the formula. *They* will break it if they think they need to, but they never think they need to.

[ QUOTE ]
I can't prove a negative so I have to say that its possible you can create a tree that will create balance, but I can say that quite a few smart folks I know (that are paid to do just that) have come up short trying to.

[/ QUOTE ]

Are you sure? Because I've analyzed this game, and a couple of others (with which I have admittedly much less experience) and I don't see the "fingerprints" that suggest its either been done to some degree, or even attempted and abandoned. If it was attempted in CoH, even if it was abandoned at an early stage, it should have left a signature in the powersets and archetypes, but the game lacks such a signature.

If there were a game for which this was tried then abandoned, I would expect that unless the entire abilities system was rewritten from scratch, there would be evidence in the form of design decisions that are apparently arbitrary in the current system, but would have been significant in an alternate balancing system.

Given how Defense and Regeneration were originally "balanced" at release, it seems clear that the mathematical tools to even attempt this, at least for melee archetypes, didn't even exist when the game was being originally designed. I'm not sure they even exist now: balance models for regeneration in particular seem to be lacking.

[ QUOTE ]
[ QUOTE ]
Here's the trivial example that demonstrates the point. Imagine a game with nothing but super reflexes scrappers and regeneration scrappers, and all SR scrappers had high order tohit buffs in their toggles, and all regeneration scrappers had -regen debuffs in their attacks. In this circumstance, even if the devs messed up and made one of them much stronger than the other on paper, there would be an additional safeguard to making too many of one of them. If SR was ten times stronger than regen on paper, and everyone started making them, eventually there would be a strong disincentive to making them: everyone would have high order tohit buffs, and that would encourage you to roll a regeneration scrapper instead, even if they might be less effective in the abstract case: they would do much better than an SR scrapper against all the SR scrappers.

[/ QUOTE ]
Why would the regen scrapper in this case do better against the SR scrappers than another SR scrapper? Isn't there also a form of positive feedback here, in that, if you want to be able to hit all those SR scrappers running around, you have to roll one yourself?

[/ QUOTE ]

There's an incentive to roll an SR scrapper to get the tohit buffs to hit all the other ones, but it isn't a positive feedback unless you're more effective as an SR scrapper against the other SR scrappers as a Regen would be against the SR scrappers in question. That's a question of proper protection balancing: you do need to make sure that it works out that way.

Plus, in this highly simplified example, the situation is actually being a little bit trivialized by having only two choices: in actuality, you'd have a couple hundred, and you'd be comparing many to many. The interesting thing is that so long as players are bewildered by all the choices and pick relatively randomly, it doesn't matter if none of them really are capable of even *knowing* what the balancing counters are that exist in the system. But if some subset of players decides to figure it out, and attempts to exploit the system, whether they are right or wrong the system will act against them by making them much more obvious targets of opportunity. The act of claiming something is an FotM will encourage people to specifically build to *counter* it, not to follow it, and the ways in which that will be possible are always going to be obvious enough for enough players to counter the FotM trend.

In fact, if too many players complain about an FotM, and not enough players realize there is a trivial counter to that FotM, at some point if it begins to damage the viability of the PvP environment, I'd simply *tell* the players what the strengths and weaknesses of that build were, to educate them.

But I'd never have to do that. There are simply too many players that are too good at optimizing for that to ever even be a remote possibility: if there is a counter, and its built into the system, lots of players will find it.

[ QUOTE ]
[ QUOTE ]
Third, create a requirement to commit to combat to achieve maximum effectiveness

[/ QUOTE ]

I'm not sure where you pulled this from, but it's not a principle of 1v1 PvP combat, nor should it be. Part of any battle is knowing when to retreat, knowing when to pursue, and knowing when and how to limit your enemies' abilities to do either. This is true for 1v1 or 1000 v 1000.

[ QUOTE ]
and force the decision to commit to occur prior to gaining complete information about the combatants.

[/ QUOTE ]

First off, that's already in effect because you don't know you are fighting and you never knew what inspirations they carried or what Accolades they possesed. Second, the devs have already subscribed to this idea by removing the ability to look at the powers of Villians. This is something many of us identified very early on, and the devs responded. I also partially agree with Thor, in the getting rid of the Archetype labels...however....I think there is a psychological necessiity to provide the combatants with some information along those lines, it may break the immersion to some degree, but I think it cuts down on some of the frustration. I also think it serves as a form of cross-game advertising. Getting beat by a Brute or a 'troller may motivate many to go and make one.

[/ QUOTE ]

That's insufficient, for reasons hinted at in the original post. Although hiding build information masks information about a combatant, that's not what I was referencing. What I was referencing is actually a concept that the devs originally had in the plan for CoH 1.0: combat stances.

Part of the problem with 1v1 combat, or even teamed combat to a certain degree, is that at some point, there are really only two things a player can do to alter their effectiveness in a material way: hop around, and shoot in a different order. Once I know how you hop around and the order in which you shoot, I can easily tell if I have a mathematical advantage over you or not from an offensive or defensive perspective: do I deal enough damage to kill you through your defenses faster than you can conceivably deal comparable damage through mine. That ratio is largely fixed for any two combatants in PvP, except for running and jousting which, even if you thought was a good idea for combat in an MMO, this game engine doesn't support well.

Combat stances change the equation somewhat. Even if you've fought the same person ten times before, you cannot know whether the particular stance he's in is a high defense low offense one, or a low defense high offense one, or something in between. Its not just a question of whether you have enough offense to beat his defense before he can do likewise, because he can change those ratios, and so can you: prior to firing the first shot, there's no way to be sure. Firing an opener and running gains you no significant knowledge, because that stance can be reset periodically.

But on short time scales, once you commit to a particular balance of your own offense and defense, you should be tied to it for a long enough period of time to make the decision materially significant for combat purposes: enough so that a bad decision has materially significant consequences (i.e. you tend to lose).

Actually, I'm thinking about something much more sophisticated than combat stances, but that is a low-end version of the idea. The high-end version of the idea is to allow for more move-countermove in PvP: allowing, say, SR scrappers to shift defense from ranged to melee to handle a strong melee threat, then having the melee fighter whip out a machine gun and rip through the now denuded ranged defenses (as a simple example). That's beyond the capabilities of the current powers design, but *not* beyond the capabilities of the current game engine itself (the game already supports the only thing required to implement this: exclusive toggles).


As to where I got this concept from, it forms the basis for a population simulation I wrote in 1988. But I have to admit, I don't know if I was subconsciously influenced by Star Trek: although its not really explained very well in any of the series, its interesting that the one tactic that isn't often used in ship to ship combat in Star Trek, given their ability to travel up to eight thousand times the speed of light, is to simply go to warp and run away, prepare their weapons, and then dash back. Anyone with functioning warp drive should *never* be destroyed in Star Trek, because human beings don't have the reflexes necessary to maneuver a starship into firing position if the opposing captain simply orders his ship to steer randomly. *Something* generally unexplained commits starships to combat, but whatever it is, its there primarily for narrative purposes, and not because of any consistently explained tactics.

But it does make for more interesting combat.

[ QUOTE ]
[ QUOTE ]
My first "official" class in games theory was in 1979, but I was pretty young at the time. I've been studying it as an actual mathematical discipline off and on for approximately twenty-two years.

[/ QUOTE ]

Damn I feel young now...Arcana started studying games theory a few years before I was even born.

[/ QUOTE ]

I was, uh, negative seven at the time.

[ QUOTE ]
If I plan a team build with X, Y, and Z in the morning only to find out that Z is now 20% less effective because some moron posted my uber build idea and everyone is running I'd be unhappy.

[/ QUOTE ]

That's not what I'm suggesting: I'm not suggesting that the game periodically tune or adjust powers based on how many people take them. You won't be 20% less effective because some moron posted your build: you'll be less effective because everyone is copying your build and now has adequate defenses to your build. Except they won't do that, because as soon as more than a few people do it, the very reason for copying your build will vanish, because they would be building something that had no particular advantage over you and everyone else that copied you.


[ QUOTE ]
If you can define a mechanism that automatically balances without lowering the effectiveness of existing characters and powers then I might agree, but I can't see a method for doing that.

[/ QUOTE ]

Here's the trivial example that demonstrates the point. Imagine a game with nothing but super reflexes scrappers and regeneration scrappers, and all SR scrappers had high order tohit buffs in their toggles, and all regeneration scrappers had -regen debuffs in their attacks. In this circumstance, even if the devs messed up and made one of them much stronger than the other on paper, there would be an additional safeguard to making too many of one of them. If SR was ten times stronger than regen on paper, and everyone started making them, eventually there would be a strong disincentive to making them: everyone would have high order tohit buffs, and that would encourage you to roll a regeneration scrapper instead, even if they might be less effective in the abstract case: they would do much better than an SR scrapper against all the SR scrappers.

Its an overly simplistic example of designing for negative feedback, but it illustrates the point that it isn't the powers themselves that is being periodically adjusted by the game, its the players themselves that are each faced with a different decision, based on all the previous decisions made by all other players.

By the way, this isn't the only way to create these negative feedback loops: there are much more complex and interesting ones than the trivial case of "I have my own kryptonite." Its just the easiest one to describe, without resorting to decision matricies.

But you could do this on a much larger and multidimensional scale, so that in theory you could add as many things (read: archetypes, powersets) as you wanted to, while keeping the basic web of counterbalances functioning as they should. In fact, the more powersets and archetypal features you add, the more opportunities you have for negative feedback balancing, *if* you keep them under control. The problem is that people think the only way to balance is with direct numerical equality: X defense = Y regeneration = Z offense. But that's entirely the wrong way to do it, because that's too difficult. What matters is not direct numerical equality in PvP, but rather roughly equal opportunity for wins and losses. It matters less how badly you lose, as long as you have an equal chance to actually win. When you define the problem of PvP "balance" to be "everyone has a reasonable opportunity to win, given reasonable build and tactical skill" instead of "everyone's powers are numerically balanced" it becomes much clearer where the balancing needs to be done: at the level of player opportunity, not power effects.

And that's why I say, seemingly paradoxically, that balancing for PvP is actually easier than balancing for PvE. In PvE, "balance" is a question of level progression, and that is a question of defeat rates in CoH: there is a strong coupling of numerical effectiveness to level progression, and that makes PvE balance a much more computationally intensive thing. But in PvP, four wins and three losses can be a reasonably good day: there's no need to balance around the amount of time it takes on average to kill 4800 other players.

This doesn't mean the powers and mechanics don't matter. It means they have to be designed to support proper negative feedback balancing, and they have to be sane so they can be relied upon for simultaneously diverse, and yet predictable (in the balancing sense) behavior. But "sane" is a much different requirement than "equivalent" for powers balance.