MP Imbalance Solutions: No Play Data. What's the point?

[Rman Virgil]

.

I pulled these posts out of Shadow Wolf's "Contingincy Mod" thread.

The points raised merit focus outside the context of that mod and thus this thread. Yet another balance thread, you say? Maybe. Maybe with a difference that makes a difference in the end...? Let's see if that bears out.

========>

The Marketing campaign for the Retail WZ in 1999 emphasized "Strategy by Design" and over 5000 design combo possibilities!!!

Before the summer of 1999 we already realized that while there may be over 5000 design combos only a small fraction would lead to wins in MP. The game was obviously and fundamentally imbalanced.

Over the course of the last 14 years several have tried to identify and address this MP imbalance. As I recall, it is a very short list of peeps: in the begining, Pro player 4nE, then Troman, then Grim, then Zarel, BP and, more recently, those quoted below.

Over the years there has been an emphasis on modifying stats and re-arranging the tech tree on the one hand and on the other, simply playing high oil on flat maps.

More recently there has been attention given to the importance of the Intel mechanics - definitely a step in the right direction in an area heretofore neglected.

So far that sums up the matrix in the quest for solutions.

Does that exhaust the systemic mix, or all the factors involved, in the 14 year quest for that holy grail - MP balance...?

No. It doesn't. There is at least one more dynamic that would make a positive difference and it's been entirely neglected in all balance discussions. We'll get to it in due course.

Right now the biggest handicap to any re-balance effort is organized MP Playtesting and associated data harvesting. There really isn't any. Without that you will only ever have untested MP hypothesis about better balance alongside the thoroughly inadequate SP testing against BOTS.

So another question would be, what's the point...?

Since balanced MP modifications cannot be adequately tested then perhaps such changes must be relegated to expanding the SP experience vs. more interesting BOT play until such a time as organized MP Playtesting and associated data harvesting become a reality....?

Much to consider.

Let's begin with the recent exchange mentioned at the top of this post.

============>

Hello all,

Sorry shadow wolf, but I'll repeat what I expect so they can be aware of that earlier...

Even if the mod was perfectly balanced (which is impossible due to complexity), there would still be "timing" issues.

One weapons will be accessible straight at the same time than an interesting body allowing the player to combine them in design... A ROF upgrade will appear just after you end your power upgrades allowing you to chain you labs loss less etc...

Whatever how hard Shadow Wolf try to balance it, some lines will be +-20% stronger than the others.

In the standard game, the game theory would come into play. One would be aware that playing cannon give a better payoff and will tend to play it more often, but then, cyborgs that is the counter to cannons would become likewise interesting and would be played just as much as the cannons is played more. That would balance both lines in a prisoner's dilemma equilibrium.

At the end, if the cannon are 20% better, they will be played 20% more often. 24% cannons, 24% borgs, 18% everything else... where we aim 20% for all. it would be barely noticeable and every weapons lines would still be played.

However in contingency, shadow wolf gave an anti-borg weapons to the cannons. So the cyborgs strategy will not counteract the cannon line as much. As a result, if the cannon is 20% stronger than the other lines, you will have a fully 20% more chance to win if you play it and nothing to counter you...

The proportion of players that will play cannons will grow until there is only that strategy played. Every other lines would be useless. We'd get 100% cannon+borgs and nothing else. Every other strat. will be decorations...

I'm telling this so everybody can have an eye on the phenomenon -

Once players will start to discover what line yield better result, all other will progressively flow into the same strat. and the research and design layer will be dead.

========>

And I'll repeat what I expect of your rant here in a much more shorter, simpler sentence:

we need more points of view that confirm that your data is valid.

They must know what they are looking for for that to happen.

===========>

@Iluvalar: balance issues apeears only when you have more than 50-100 players and at least 5 pro-players

(that digits just because... i hope you undestand idea)

That's why I warn asap, so players can keep an eye on the problem and identify it asap and confirm the inevitable faith of that mod. Meanwhile, ignorant players might pretend the game is "fun". And Shadow Wolf might been dragged into wasting more time on that idea.

Again, I fully respect Shadow Wolf, I see the very good work he done since now, it's just a shame he built that tower on mud. If it was on sturdy ground, that mod would have beaten the original game...

============>

I'm talking about the core ideas behind the mod, not the execution of it. The execution is very nice. Shadow Wolf is a very good modder.

Many people disregard the game mechanic, the devs do the same BTW... but the mechanic abide to some rules, and it's not always self-evident. I believe I explained enough...

=========>

[NoQ]

This community needs more players (:

[Rman Virgil]

This community needs more players (:

For sure. It's a "Catch-22" too.
.

[Reg312]

balance = (?)
1) when player lose game - he can say "no balance"
i mean its hard to determine balance
2) pure balancing follows to ugly gameplay
as example: light body was stronger than medium body in NRS Mod

3) players have uneven skills and most games are unbalanced only due to player skills..

4) ....

my conclusion: pure balance discussions never can be finished with effective result
we have new balance patch by NoQ.. when we start playing it then we will find new balance issues.
that process is unending


*when we get better mod-support in Warzone then it will be easier to test balance changes
**we need more compicated unit design screen in game.. with more unit parameters and calculations

[Terminator]

warzone needs real statistics of common used templates. For example every game client (im MP only) collect data about used units(designs & produced), lose units & killed, & so on... And in the end of the game (win/lost state) client sends data to master server where it could be analyzed in any way. I know not much game lasts so long to see a win FMVsequence, but its better than nothing.

[NoQ]

when player lose game - he can say "no balance"

Then the winner publishes the strategy, and the looser tries to win with this strategy, while the winner is doing his best to counter it with something else. If the looser wins the second game, then it's not a balance problem yet, but it's already a nice starting point for investigation.

[Reg312]

when player lose game - he can say "no balance"

Then the winner publishes the strategy, and the looser tries to win with this strategy, while the winner is doing his best to counter it with something else. If the looser wins the second game, then it's not a balance problem yet, but it's already a nice starting point for investigation.

its just balance in duel games.. such games is not best games for me
each strategy requires many probes and adjustments..
for example, rocket hit&run is not only strategy - you need pactice it
version 3.1 can be played very differently depending on game ping.. e.g. in games with high ping flamers are better a bit

another problem - even pro old players dont know all possibilities in game
for example i see tracked propulsion underpowered in 2.3.9. while iluvalar said it is underpowered just because no one knows how use it

so adjust games balance only to duel games - is good and simplier method.. but we get unexpected and possibly wrong result in team games


balance depends on human factor..
another example: some player likes rockets,MRAs and attack tactic.. so when he lose games he can say "flamers are OP" or "towers are OP" or "high oil games - sh*t"

[aubergine]

What if a game went through some evolutionary stages?

So you start out at T0 with scav tech. Scav buildings would initially be difficult to destroy, because you'd be starting with trikes and mgs. Thus, a rushed player couldn't be wiped off the map, their buildings would stand for some time, giving them chance to research something that can pummel trikes and get a few units out to get themselves back in the game.

The scav tech slowly advances to a point where scav buildings can more easily be destroyed by scav units. Then we enter T1, era of the cyborgs and some of the more modern buildings.

Again, cyborgs, initially can't destroy modern buildings, but are effective against scavs and the older buildings. Thus, as players start to get their modern buildings they again get the benefit of being able to survive base attacks for longer.

Then T2 would usher in tanks, etc.

T3 would usher in VTOLs.

The aim would be to try and minimise the adverse effects of imbalances, rather than try to achieve perfect balance (which will never happen).

As has been pointed out in posts above, its not just balancing the research tree and stats, it's player skills as well and for 90% (a rough guess, but I imagine fairly accurate) of players it's their skill levels that are leading to imbalance.

[Rman Virgil]

.

I aggregate here the continuation of this discussion in Shadow's "Possibility of a Contingency Campaign".

viewtopic.php?f=5&t=9523&start=15


==========>

Variety and fun were also WZ Creators goals and priorities when they created the game. The implementation of "strategy by design" added an other layer of complexity to the already complex matrix of component stats and tech tree research advance, etc. All these layers of complexity was a strategy in itself as far as ever having to achieve optimal balance and players figuring out what to focus on and what to ignore. This all breaks down in the process of achieving pro-level competency. Not many get to that level and when they do, any major change is not exactly welcome. So the fun and variety of changes to the game will work for a majority of players who are not pro and have no interest in that side of the game.

WZ complexity is a White Box that may as well be a Black Box when it comes to the usual methods of approaching balance adjustments. Those balance methods themselves have become a mental box that has excluded a whole other realm of fun and variety that would also attack balance issues from a more profitable angle that works with all those layers of complexity in such a way as to render RPS balance far less critical - for the casual or pro player or any level of skill in between.

===========>

I see your 2 different positions (Iluvalar & Shadow Wolf) more a matter of emphasis / focus than conflict. In the context of fun, both are valid and both will serve their respective audiences well (very different audiences for the same game, are a reality mostly ignored & that's a grave mistake right up-front).

Both represent pieces of the puzzle to advance GP effectively on all fronts, seperately and in tandem. There's just more to that puzzle than these 2 positions. Ironically, you both have common ground in that other over-looked area of the missing puzzle piece. Unfortunately, that area of dev has been stymied by insufficient play-testing support.

============>>>

........

@Rman : I understand your point about the distinction between pro and casual players. But:

A) The learning curve in a dominating strategy game is absurdly fast. Distinguishing both in such scenario is a pit trap.

B) I refute the fact that casual players, at least a majority, don't enjoy the capacity to improve themselves and gain skill while playing and become better.

Therefore, I fully understand the exploration-type of fun of discovering the new universe and new possibilities, but unless Shadow Wolf provide a constant supply of new changes periodically that fun will stop soon after the last release.

Is that what you aim for Shadow Wolf ?

That was exactly what Pumpkin was doing through the 10 patches to the retail game. It was their on-going, specifically stated strategy, till cut short by Eidos closing them down....

That said.... the question becomes, would Pumpkin have followed that same strategy of dev for WZ 2120?

I think not. I think they would have went back to deving a component of WZ 2100 that they had to scale back on, and almost abort, to meet the release deadline set by Eidos. That component would have made a shift in GPMs to place a much higher value on the effective use of complex, multiple combat group maneuver tactics and combined arms.


========>

@Rman : I understand your point about the distinction between pro and casual players. But A) The learning curve in a dominating strategy game is absurdly fast. Distinguishing both in such scenario is a pit trap. B) I refute the fact that casual players, at least a majority, don't enjoy the capacity to improve themselves and gain skill while playing and become better.

I agree... but with a caveat.

That fast learning of the pattern, what to focus on and what to ignore to win, takes place if you choose to play a lot of MP with the goal to learn the patterns. In ain't gonna happen in SP vs BOTs and it ain't gonna happen without that dedicated effort in MP vs good players.

I think the majority of the game's audience is not gonna make that dedicated effort in MP and will contentedly remain decidedly non-pro. So "Contingency" is safe with that audience.

And if Shadow Wolf follows Pumpkin's strategy, keeps changing the matrix with new additions and rebalances, then "Contingency" is safe with more pro-level players.

That leaves the entirely unexplored "Third Way" approach mentioned above in the context of WZ 2120.

===========>

That leaves the entirely unexplored "Third Way" approach mentioned above in the context of WZ 2120.

And what could that "3rd way" be?

..........

That was exactly what Pumpkin was doing through the 10 patches to the retail game. It was their on-going, specifically stated strategy, till cut short by Eidos closing them down....

That said.... the question becomes, would Pumpkin have followed that same strategy of dev for WZ 2120?

I think not. I think they would have went back to deving a component of WZ 2100 that they had to scale back on, and almost abort, to meet the release deadline set by Eidos. That component would have made a shift in GPMs to place a much higher value on the effective use of complex, multiple combat group maneuver tactics and combined arms.

.

IOWs......

Because of the extremely rudimentary state of Command and Control of multiple combat groups, maneuver is restricted to bumbling babysitting via monolithic, symmetric vectors. This makes in theater conflict possibilities super simplistic thus the over-focus on the predictable, pattern identification, of pre-conflict, non-combat GPMs.

===========>

.

I think the majority of the game's audience is not gonna make that dedicated effort in MP and will contentedly remain decidedly non-pro. So "Contingency" is safe with that audience.

So in definitive, this is a single player mod ? That have absolutely no ambition to be, one day, competitive in multiplayer games ?

@Shadow :

No, I don't mean simple addition, or pathetic tries to do the impossible balance, but more changing in the stats just for the fun of changing the stats each week or so. The "fun" would in metagaming where players try to find out what is overpowerful this week. That would keep your mod playable for some audience.

.

===========>

@ Iluvalar: When I had access to statistics for several years that was the trend. Perhaps it would be worth polling now.


As for your suggestion to Shadow - thats one way.

I still think he can follow the Pumpkin pattern and introduce new stuff in a more staggered, spaced apart, timeline (along with associated balance changes). That release schedule model seemed to work out well for the tens of thousands of fans of the original retail, MPlayer, WZ community. It was like experiencing X-mas every few weeks.

===========>

[Rman Virgil]

I still think he can follow the Pumpkin pattern and introduce new stuff in a more staggered, spaced apart, timeline (along with associated balance changes). That release schedule model seemed to work out well for the tens of thousands of fans of the original retail, MPlayer, WZ community. It was like experiencing X-mas every few weeks.

I don't believe this situation can hold the comparison. Yes, Pumpkin did released a work in progress and improved it a lot afterwards... But obviously, at least one of the knew what they were needing for a balance to be possible. I still find once in a while curious patterns I didn't never realised before. One of the was a game designer and he knew what the constraints he needed/wanted to follow in order to achieve the game.

Shadow wolf, however, is nuking the pillars in the name of the "variety". But they are what sustain the variety we already have and he dont realise it at all...

=========>

@ Iluvalar: You could be right. It could turn-out that way in the end. But I think letting it play-out in actual gaming over time will be the true test and will also, possibly, yield worthwhile object lessons and even useful data. Actual gaming over time has been known to overturn even the most robust theoretical predictions with totally suprising, and satisfying, results.

============>

+1 I'll agree with that. Here is a quote from my first post :

In any ways, I strongly suggest that you concentrate on failing to balance what you already have in contingency. The quicker you realize it's impossible, the quicker we gona have that next mod.

I consider starting a new campaign project now wouldn't be a valuable effort.

Schedule :

1- Fail Contingency
2- Learn from error and rebuild another mod
3- Talk about campaign.

=========>

Iluvalar, there are some parallels here to the Science in Economics, don't ya think?:

Is Economics a Science? - It would be, if it weren’t for the people.

========>


.

Let me posit this reiteration with elaboration.

1.) There are 2 distinct audiences here: Continginency for MP and Contigency for Campaign.

2.) The overlap of audiences is practically nil.

3.) Retail WZ introduced the Campaign Matrix FIRST.

4.) Over the course of 10 patches the MP Matrix was introduced gradually, altered along the way and significantly diverged from the Retail Campaign Matrix.

5.) Some components of the Retail Campaign Matrix NEVER factored into the MP Matrix - or even made it into the MP Matrix..!!!

WZ Creators followed a different schedule with a different agenda. SEE: "Is Economics A Science? - It would be if it wasn't for the people."

.

[Iluvalar]

.
No. It doesn't. There is at least one more dynamic that would make a positive difference and it's been entirely neglected in all balance discussions. We'll get to it in due course.

I'm waiting...

2) pure balancing follows to ugly gameplay
as example: light body was stronger than medium body in NRS Mod

That's not true. It's true that some weapons perform better on light bodies, but not all lines. Light and heavy bodies should have same overall strength in recent versions.

another problem - even pro old players dont know all possibilities in game
for example i see tracked propulsion underpowered in 2.3.9. while iluvalar said it is underpowered just because no one knows how use it

I'll refine my opinion : They are actually slightly weaker because of the python domination. Python rule, but pythons become too expensive on tracks. Mounting tracks on cobra or scorpion should do the trick, but the python body being so OP, it actually kill the benefice.

However, it's still inside of the weapon modifiers margin (for the mods that have one ). And such, they are still a clever choice to face mortars, flamers or mg and they are also nice for defence or slow build-up of armies when you don't need speed.

[zydonk]

Strange. I've been playing two recent mods in sk mode and I thought it has to be time to make a number of points about the (relative) plethora of rebalances. So let me add my pennyworth here.

If you want an effective mp balance, just check out existing mp games and copy the best. That's it. Don't waste your time fiddling with WZ. There's nothing special about WZ if all you want is a quickie.

Next, who said starting a rebalance from scratch in WZ is a sensible idea? Yet everyone does exactly that. Be sure that a rebalance offering will try from v.001 to balance everything, so that we end up each time with single dimension mush that starts you with a rush of anxiety and ends with a long drawn out whimper. If an accountant's rebalance is the great idea why don't we already have The Rebalance?

OK. That said, let me propose two conditions for rebalancers.

One: that they have actually played WZ as it was originally designed to be played. Ideally this would be The CD game upgraded to v1.10 (239 is a little soft in places). They should have played this CD version at difficult level for a few years over a wide range of maps, Pumpkin and custom. In other words, you should know what it is you are rebalancing.

Two: the game should be rebalanced by degrees. Modify first the obvious imbalances (yes, there will be arguments as what these are) and see how these play out. This is what Pumpkin were doing. You should not need extensive play-testing. An experienced player will register the differences very quickly.

In fact, every player of any experience should have a go at rebalancing. I would do it myself if I knew how...

[Rman Virgil]

.

.
No. It doesn't. There is at least one more dynamic that would make a positive difference and it's been entirely neglected in all balance discussions. We'll get to it in due course.

I'm waiting...

lol.... I did. IOWs (above, a re-statement, summation, of several prior)

.

[Rman Virgil]

.

Shadow Wolf's summation of his position makes key points that apply, IMO, outside his Contingency creation...

@Iluvalar: Let me get one thing straight: I, 1st and foremost, created the Contingency mod as a primarily multiplayer-oriented mod, though I COULD (meaning I won't guarantee) one day (which could be anywhere from a couple months to a couple years) take one set of multiplayer balance data from it, tweak it, and adapt it for a singleplayer experience (regardless of how balanced it is). That's frequently what I see in games like Command & Conquer 3: Tiberium Wars and StarCraft 2, as the 2 sets of data, singleplayer and multiplayer, might've shared a common beginning, but often tended to drift apart as time went by (as was especially apparent with StarCraft 2). Also, it's very unlikely for a singleplayer experience to need balance changes on as frequent a basis as a multiplayer experience. (Again, for example, while the multiplayer experience in StarCraft 2 continues to change to this very day, the singleplayer experience has remained virtually intact throughout all these couple years.)

Personally, me working on the Contingency mod was just me practicing for when I'd eventually want to develop my own rts game (which would've likely played differently from Warzone 2100 altogether, let alone the Contingency mod, despite possibly being somewhat of a spiritual sequel). Regardless of whether or not the balance fails, I'm unlikely to develop another multiplayer-oriented mod for Warzone 2100 in the future, though I could still learn something useful from developing my Contingency mod.

Once I'm ready to move onto other things, I may want to let others do the rebalancing for me. I'm aware that trying to keep the gameplay balance fresh for years would require that much time to support, and I'm personally not willing to devote that much time on maintaining such a thing. I'd burn myself out within only a couple months. (I'm already feeling fatigued when it comes to developing the Contingency mod.)

.

[zydonk]

Well observed by Shadow Wolf. He crystallizes the difference between mp and sk balancing very well indeed. The effect for me is to point up the - now obvious - fact that the WZ dev project is concerned by and large with the mp aspect of the game. Fair enough - that's where the pressure for change has been strongest.

My only regret from this perspective is that the vanilla game has been so weakened by successive rebalancing. All rebalancing should have been restricted to mods and the game itself kept to its original form. I realise this would not be possible for the master or v3.1 onwards - but they are mp games pure and simple, where single players will have to make do with what is left to them.

OK. That said, what I don't understand is why the real problem facing mp rebalancing has not been addressed directly. Two points.

One. Shadow Wolf can see that in effect each mp player requires a bespoke balancing of the game, if only because each player develops a unique playing style. The problem underlying the repeated calls for rebalancing is the fact that one player's play-style might be "better" than another, ie competitively advantaged simply by the kind of personality and character one individual player has in contrast to another player. No amount of rebalancing can resolve this kind of existential fact.

Two. The rebalance mods as they are available (played NRS and Contingency most recently) are ludicrously complex for what in effect will be 40 to 60 minutes of panic response. What happens, of course, during play is that individual players select from the research menu those assets that best suit their styles of play, which are then upgraded to a level determined by resources and time. The aim of the modifier in this case is in effect to level off the various weapon and vehicle offerings to some common denominator - which will (if only because it is easier to achieve and no doubt in keeping with the brevity of the mp bout) tend to a/the low(est) common denominator.

There is a third element here that is more directly related to the WZ game itself. Many of the more recent rebalances have been undertaken as little more than intellectual exercises - to practice IT skills or as school projects or (certainly not!) out of vanity. It explains why the rebalances are (a) so comprehensive and (b) unrelated to actual WZ gameplay (ie map restrictions and the ranges and speeds practicable as a result - WZ is a "slow" game compared with C&C or TA, for instance). This fact has always underlain my repeated suggestion that rebalancing be restricted to mods, where they can have no permanent effect on the game itself.

Is that it? Enough said for now...