Who Are You Balancing For?

Extraction (née Dirty Bomb) is the new team-based FPS Splash Damage is developing. I am not going to rehash Brink, since even Splash Damage is doing their best now to pretend that game never happened. Suffice it to say that I had not bought a multiplayer FPS going on 9 years before Brink, and after that experience it will probably be just as long before I’ll be tricked into buying another.

What I find interesting about Extraction–besides that anyone would pay Splash Damage over $100 (or any amount of money) for access to their closed alpha–is their new balancing system, which they are calling Echo.

This started throwing up all the same red flags as when I heard some fighting game developer talk about how they collected online data about characters and moves to help them balance their own games. It is not even just that a variation of Sturgeon’s Law is always in effect when discussing any competitive player base, but that people are lazy and evolution takes time. So who are you balancing for, the 90% of players who are bad at the game, or the 10% of players who are good at the game? This is especially problematic for games that are heavily based on teamwork, where no amount of forced cooperation in a pub environment can ever come close to an actual team of coordinate competitive players.

So, your fancy stat tracking program decides that most players are incapable of actually hitting anything with more than 1 in 4 shots from their spray and pray rifle. Do they need more ammo, or do they need to get better aim? Does it matter that most players can’t hit 30% of their railgun shots when pro players expect to hit at least 50%? Does the railgun need to be buffed, or nerfed?

There is a reason nobody pays attention to tier lists in fighting games until after sufficient tournament play. Top players need enough time to demonstrate the practical application of theory fighting. As the idea of fighting games competitive fighting game play has become more excepted, even normal, the language used in reviews has also changed. Seldom do modern reviewers have the temerity to make declarations about the balance of certain characters and mechanics based on the 48 hours they spent with a review copy of the game. The word “cheap” has also made an exit. That’s because every statement of the type made in the past was as wrong as when the Starcraft strategy guide I had back in the day said that Firebats were effective against buildings.

Everyone knows about the skill ceiling–the maximum level of mechanical skill for a weapon, hero, race, or game–but there is also a skill floor which develops at the same time. When Quake 2 (where the railgun originated) was new, 56k modems were still a dream. A player could expect 200+ ping to their nearest server on a good day. A hitscan weapon that did 100 damage and had a 1.5 second reload time, on top of the slow weapon switching speeds in Quake 2, seemed reasonable enough at the time. Even with Quake 3’s release, most players were expecting 80+ ping to a server on top of poorly optimized net code settings. But it didn’t matter. Pings got lower, and players got better, and the railgun, already considered a very good weapon in Quake 2, started to take over. Later mods like CPMA would nerf it hard, and even Quake Live lowered the damage. The fact is, a good Quake 3 player today is just vastly better (in terms of acquired and applicable skill, not necessarily talent) than a good Quake 2 player from 1998, even on a LAN server. It’s the same way a professional Brood War player from 2008 was better than a professional Brood War player from 2003. That’s not a knock against the earlier players–standing on the shoulders of giants and all that–it only demonstrates that the requirements for professional play changed over the years as players learned more about the games.

One could say the same of every competitive game, even sports. Go back in time and there will still be great players and great games, but as the game itself is played, everyone in it has to evolve. The develop new skills, and old strategies die out or become superseded. It’s not a bad thing, either. It is the sign of a good game that over time the players have to get better and better to compete, because the skill floor raises in proportion to the skill ceiling. A bad game would simply be solved by the good players, leaving no room for further evolution, and everyone would move on.

It may seem unfair to do a 1:1 comparison between the most iconic hitscan weapon in FPS history and an offhand remark about an SMG in a promo video for a game that doesn’t yet have a release date, and it probably is unfair. But I think the point stands, especially for a game that wants to be taken seriously as competitive (and don’t they all these days?).

The lesson here is that pub stats correspond more with gimmicks than with a balanced competitive environment. Another game with a lot of stats available is DotA 2, where sites keep track of the win rates for both pub games and professional games. According to dotabuff.com Zeus has the 3rd highest win rate of any hero since DotA 2 entered closed beta (at this article’s publication date), while Chen is both the 2nd least played hero in pub games and also has the current lowest win rate. According to datdota.com Chen has the 4th highest all-time win rate in competitive games, and is also the 11th most picked hero, while Zeus has the 3rd lowest win rate and is the 11th least picked hero. So which of these heroes needs a buff, and which needs a nerf? Is Zeus really that strong and competitive players are ignoring him, or are his perceived strengths merely gimmicks? Is Chen an awful hero, or do pub players really suck with him? Why does Chen keep getting nerfs while Zeus is getting nothing but buffs?

Not to say that pubs should be ignored, or that stats should be ignored. Both are important for balancing games, but even more important is who is doing the balancing, and how and why. That is why I have serious doubts about any automated system trying to do the job, even if it is only being used as filters for programmers. Also, Splash Damage, but I already said that.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s