Balancing multiplayer RTS games -- a herculean task?

It’s fascinating to me what an incredibly difficult process it is to balance multiplayer RTS games.

What does “balance” mean? For the purpose of this discussion, balance means every unit is useful in the right circumstances. Balance means there is no one perfect strategy, or one perfect unit mix, that works all the time.

I know Warcraft 3 like the back of my hand, since I’ve played it so much, and even after five patches-- we’re at 1.05 now-- the balance is still being significantly tweaked to address the numerous min-max exploits that 50,000+ players beating on the game non-stop have slowly but surely developed.

Based on the real world experiences of the RTS games I’ve played extensively online, balancing your game is evidently a herculean task. I mean a task that can take OVER A YEAR after the game is released! Holy… crap. That puts “when it’s done” in perspective, doesn’t it?

It’s gameplay darwinism-- you tend to use (and abuse) strategies that other players use to defeat you. Over time, this process guarantees “ideal” unit mixes which cover the majority of the games you’ll play. This runs counter to the design directives of any RTS game. Why have units in the game that nobody uses? Why have “uber” units that guarantee victory 90% of the time? Clearly no game designer wants this. Heck, as a player, I don’t want this!

I don’t want this to turn into a WC3 specific thread; I can cite examples of this from Total Annihilation as well. The big difference in that case is that Cavedog didn’t do a very good job of balancing their game-- they provided some cursory, lackadaisacal balance “patches” in the form of new units, for example the flakker anti-air cannon. But that was about it. Whereas Blizzard has a sort of implicit company policy that they will do whatever it takes to balance their games. Sometimes this means sweeping fundamental changes or “nerfs”.

Anyway, I encourage any other RTS online players with significant experience (read: 50+ games under your belt) to chime in on this topic.

Here’s the specifics on the WC3 changes coming in the expansion. Current balance problems with WC3 involve massed spellcasters, which are WAY too effective. And they eat level 3 melee units for breakfast courtesy of focused fire and pierce attack bonuses.

Looks like there are some MASSIVE changes coming, which almost rewrite the rules for how the game is played at even the casual level:


Most spellcasters (Sorceress, Priest, Shaman, Witch Doctor, Necromancer, Banshee, Druid of the Talon) have been re-balanced and given new armor and weapon types. All of these casters, as well as Dryads, now have “Unarmored” type armor, which means they take additional damage from Piercing and Siege attacks. Additionally, all of these casters (but not Dryads) deal Magic damage, which works much like Piercing, except it does additional damage against Medium armor instead of Heavy armor.

Most advanced technology structures have had their build times reduced. This reduction ranges from 20 seconds for especially underused structures, to 10 seconds for less underused structures. For instance, the Slaughterhouse builds 20 seconds faster now, while the Arcane Sanctum build time was only reduced by 10 seconds.

Units can become ethereal in certain situations. Ethereal units can only be damaged by Magic attacks (from casters such as Sorceresses, Banshees, etc) and by spells. Ethereal units cannot attack, but can still cast spells.

Magic-immune units cannot be attacked by units with the Magic damage type, such as Sorceresses, Banshees, etc.

Gold bounties given for killing creeps have been SIGNIFICANTLY reduced, especially on lower-level creeps.

The gold costs of all units and buildings have been reduced by roughly 15%.

Auto-cast buffs have been rebalanced with new mana costs, and in some cases such as Curse, have been slightly upgraded.

Night Elf Ancients now have Heavy armor when uprooted.

Thunderclap slow duration and strength have been tweaked so that the slow effect lasts 5 seconds at all levels, and decreases attack speed and movement speed by 50% at all levels.

Entangling roots, ensnare, and other immobilizers no longer prevent units from town portaling.

Many more minor tweaks to units and abilities…

I suppose it’s difficult to balance any game where thousands or millions of people will be playing it and trying to find an edge. I wonder what the history of older real-life games like chess is, and to what extent they have been tweaked over the centuries in order to eliminate exploits, etc.

Isn’t it almost a recursive process of metagaming? I.e., the guys who really want to win play to the rules, and sometimes the rules must be changed, or sometimes the nature of the game changes, and what was once “metagaming” is now simply the way the game is played. Take tennis. Back when it started in the 1870s or whenever, people “served” in order to put the ball into play, just hitting it in underhanded to begin the point. Later some players developed the overhand serve and it became a “weapon.” Some decried overhand serving as against the spirit of tennis, but people stuck with it and it became an intrinsic part of the game. Now you have people who spend thousands of hours learning how to bomb 120 mph serves so they can get lots of free points. It’s just par for the course, though some people complain (for this and other reasons) that the game of tennis has become “broken” and yearn for the days of wooden rackets.

To be honest I dont even really get excited about RTS titles anymore. There WAS a time where Red Alert and Starcraft dominated my PC gaming like none other, but its quite obvious that time has passed.

A lot of it has to do with the simple idea that I have a finite bank of time. Yet if I am to play online, I challenge people, often kids, that have nothing better to do than find the most efficient build order and crank out the dominating unit and murder me.

This sort of exercise, while I am perfectly open to participate in, I feel destroys the soul of an RTS. It makes it a tool to achieve victory over someone, rather than a fun game. Winning is a lot of what games are about, but there is something to be said for just taking part in a ‘good game’.

I currently only play Kohan. I do play the faction with the weakest Econ, so I dig my own hole in most situations. But what I love about Kohan is that you can either play an Econ game, or a Field game. Royalist Econ is pretty piss poor, so I place 90% of my attention on the field. While other players (mostly council) use their insanely effective econ to produce gobs of goons and hurl them at me in giant clumps. I am able to use a smaller number of companies that can often prove to be more effective simply because I am taking every advantage of my environment, and where my units are in the field. There are many games where my flanking company will reach veteran status simply because my opponent isn’t playing a Field game, but rather an Econ game, which requires its own amount of attention.

Further more, there are enough ‘safeguards’ in kohan to keep the effects of rushing to a minimum. While a council player can out produce a royalist player in terms of number o units, a good royalist player can his company bonus and some proper field tactics and solid company mixes to combat this.

Not to say all Council players go for the Econ based Super Expendable Goon Clump, but its something I run into often.

That said it will be very hard for any game to compare to kohan, to me anyway. Most ‘popular’ rts offerings this generations have been extremely pretty, but have not fixed what I find so terribly wrong with them. Not to say I expect them to fix it. Most people who play RTS titles online competitive ENJOY loading up a game and winning within 15 minutes. Winning is their goal. Every game of Kohan I play I learn something new, and plan to utilize it sometime in the future.

God bless Blizzard.

They’re finally nerfing the damn casters. Took them long enough. (personally though, I think any caster with autocast should not have an attack…)

We have about 7 people (6 testers, 1 designer) assigned full time to balance in our games. It’s a major focus, and even with that there are always still issues we’ve decided to patch. Obviously there will always be SOME dominant strategy, and oftimes what the fan community is up in arms about is perfectly counterable… but sometimes not.

It’s definitely a huge task, and only getting huger as games get more complex and less symmetric (ie, balancing Norse ox-carts vs Egyptian pharaoh empower).

Most people who play RTS titles online competitive ENJOY loading up a game and winning within 15 minutes. Winning is their goal

Uh, yeah. You play games to win. It is possible to design games where the goal isn’t winning, of course, but the majority of games are only “fun” if you “win” at least half the time. I don’t care how good of a sport you are, nobody enjoys getting slapped into the dirt time and time again. Wasn’t this a significant factor in the closing of Motor City Online, as I recall? Heck, forget MCO, we need look no further than these boards to see proof of this. I’ve seen dozens of complaints right here about how WC3 is too difficult online.

That’s why auto matching systems are so critically important. You MUST FORCE people of similar skill levels to play each other., which otherwise does a spectacular job of auto-matching, has a critical flaw in that it lets people create unlimited level 1 accounts with a single cd-key. I fear that left to their own devices, a lot of people will get their rocks off beating up newbies because they get to “win” all the time. Debatable, I suppose, but we have to protect everyone from the 5% of jackasses lest they ruin it for everyone else.

It’s definitely a huge task, and only getting huger as games get more complex and less symmetric (ie, balancing Norse ox-carts vs Egyptian pharaoh empower).

Yep, I suspect AoM has even more severe balance problems than WC3, if for no other reason than the cubic assload of units and upgrades in that game. In other words, you can’t play as the Undead in chess.

It’s really too bad that there is this crazy expectation of “different sides” asymmetry in every goddamn RTS these days. That’s a total victory of marketing over gameplay, because this has NOTHING to do with how fun a game is, strategically speaking.

And now I vet everyman Mark Asher to patiently explain why, really, different sides are more fun. You know, for the people.

I’ve lost several times to people in games and still felt great about the experience. I’d take a great match over an easy cheese win any day. Losing should be a learning process, not a quick trip to gamefaqs for the ultimate build order.

I simply can’t stand RTS games where the super clump is so damned effective. In a game like Kohan i’ve come back from having only two towns by raiding back country and producing a front line force and one flank force, that poo’ed all over a stupid super clump.

There’s a reason I refuse to play Council in Kohan ;)

Obviously there will always be SOME dominant strategy, and oftimes what the fan community is up in arms about is perfectly counterable… but sometimes not.

I’m not sure the logic adds up here. You have to look at the statistics of online gameplay, something which Blizzard does extensively. If a particular unit is only built in 0.5% of online games, THERE IS A PROBLEM. As long as your sample size is large enough-- and it’s clearly massive in the case of it doesn’t matter what the user community “says”, just look at the numbers.

I wish the all knowing Blizzard would fix the fucking Necromancers. I used to love playing that class, then they made them completely useless in competitive and high difficulty play.

Absolutely. I agree 100 %. I think I’d have more fun with WC3 if I didn’t consistenly get my ass handed to me severely by the gimped AutoMatch. We’re talking in terms of Operant Conditioning/reinforcement schedules. I receive nothing but punishment from WC3. Consistently. I really want to like that game, but it’s just too damn frustrating.

Chess is turn based. :). RTS chess, those queens, bishops and castles would be nerfed real quick, players would complain of being “pawn rushed” and no one would ever use knights because they are just too darn hard to micromanage, and developers, if you’re reading this, they should have an “auto-move” function anyway.

Different sides make the single campaign (or even LAN games) more interesting. I tend to play against people I know, and who have bought the game about the same time as me so our respective skill levels are the same. The different sides add a bit of spice, for example knowing that your melee units are significantly stronger, but you have an ground to air weakness.

For me it’s not the different batches of units that are the problem in game balancing (though I am bloody glad they are looking at that mage issue), it’s different game concepts. Say we have a game with two sides that goes side one = death dealing, side two = healing. Generally, it’s been the side that deals more damage dominates, as a damaging unit just does, whereas healing units need mana, have to be close, might not have auto target etc etc etc.

And reading over those changes, my first thought was “Starcraft?”. Borrowing units? Exploding flyers? Oh my!

Not sure why video games should be different than any type of game or sport. Professional sports leagues are continually “balancing” their games, and they’ve all been around way longer than Warcraft 3 for example. The NBA is continually futzing with the lane size, 3-point line, various rules, etc. In hockey they’re in the middle of a balancing period to adjust the level of offense vs. defense (that came about from previous rule-sets).

It’s possible that there’s no such thing as a game that has been completely balanced. People change over time and the more they specialize the better they get at any given thing, mostly in ways the game’s designers didn’t think about.

Glad someone brought up the sports analogy. If I can drift this thread slightly, I think it’s HIGH TIME the NCAA moved the men’s division I basketball three point line back to the NBA or at least the international FIBA distance. 19’9" is ridiculously short; if a chump like me can hit three, four, or five treys in a pick up game, then for damned sure it’s too easy. In today’s world, high school athletes are bigger, stronger, and more skilled than ever before. They get to college and beef up even more. By the time their junior year rolls around (if they stay in that long), a 20’ jumper feels like a layup.

As for balance in RTS there are no real easy answers out there. I think it’s in the developers’ best interest to always have a multiplayer balance test so the game is reasonably playable by the time it reaches retail, but from what we’ve seen in recent years, more tweaks are inevitable. For those who stuck with Starcraft and Brood War over the years, they did eventually reach an admirable level of balance in that game. It’s also worth noting that the more popular a game becomes, the more difficult it is to balance it, simply b/c there are more minds out there hacking out the one-true-path to victory.


Glad someone brought up the sports analogy.

cough secondpostinthefriggin’threadcough :)

"I’ve lost several times to people in games and still felt great about the experience. "

Same here. I do get into the game and want to win, but a fun knockdown drag out game is a blast to play for me. When I first started plaing RTS I got my ass handed to me for a whole summer in WC 2. But it was one of those I was going to get better no matter what to beat these guys. I watched them play and gt at it. Eventually I got to the point of being one of the better players in out group. Of course living in the Apt. that was usually the place for any gaming helped since I got to play allot. Wasn’t to good for the grades though. :)

Yeah, those Kilrathi pilots can be a bitch if you don’t pick the right wingman.

I watched them play and gt at it. Eventually I got to the point of being one of the better players in out group.

Right-- multiplayer darwinism. You emulate the strategies that others use against you, until you can win. Which puts the game balance on a razor’s edge; as the number of players goes up, the faster they will realize the min-max solution.

i don’t pretend to understand wc3’s leveling or match making system but i did run across some really interesting statistics. just to point out one…

solo level ladder disturbution goes from level 1(71k players or 30%) which is the huge red spike on the left, to level 33(2 players) on the far right.

it seems a tad weird to have a graph like this where ~150k(65%) of the players are level 3 or below while the highest current level is 33. there must be tons of disparity among the lower levels just given the sheer number of players in them(which might explain why some people seem to have a hard time getting started).

more info -

They just increased the komi in Go to 6.5 from 5.5 in Japan, as they have previously done in Korea, and that game has been around for a few thousand years. In China the komi is 7.5. This to reduce the advantage of having black.

Can you explain this a little? I don’t know a thing about Go so I don’t know what the komi is. Also, when this was increased in Japan, was it by some sort of Japanese Go federation body or something like that? And how did they let people know?