In the summer of 2007 I played my last game of online StarCraft. The people I was subletting from had a copy on their bookshelf, and one night I gave in to nostalgia. After tinkering around in a few skirmishes, I decided to dip my toes into unranked non-ladder online play. How bad could it be?
It turned out it could be pretty bad, because I was pretty bad. In the nine years since Starcraft had launched the general level of play had advanced while my skills had atrophied. I died to my opponent’s first push. I then joined a random 4v4 game on Big Game Hunters, a game mode and map specifically designed for people who don’t want to deal with all the fuddy-duddy details of being good at Starcraft. Within 90 seconds, two of my allies cursed me out and quit. The third helpfully stuck around for another 30 seconds to yell at me for having a non-optimized build.
After the jump, sink or swim
World of Tanks seeds battles in the same fashion. Clicking “random battle” throws you into a truly random battle. While matches are tier-limited, the game puts new players opposite players with over 10,000 games played. If you’re driving around in a tier 1 tank people should accept that you might be new. This being the internet, they don’t. Sometimes your allies will swear at you, claiming that you have lost them the game.
The good news here is that they’re being hyperbolic. It’s nearly impossible to screw up a World of Tanks match so badly that you doom your team. 15-player teams mean that one bad player makes less of an impact than in a StarCraft match. The bad news is that in a battle with a few new players and a few really good players it is entirely possible that the good players will all be on the same team, staring down the poor doomed newbies across Malinovka’s field. The matchmaker does not care about balanced matches, and has no interest in helping you learn to play, or ensuring that your match is fun or interesting. Because of this, new players negatively impact the teams they are on. If you want to get into World of Tanks, accept that you’ll be a net negative while you get the hang of things. We all were there once!
This is a remarkably archaic approach to seeding online games. In StarCraft II, new players are put through a training league and a series of placement matches which are used to place them in a league. Bad players are placed into Bronze League, and as they improve they graduate into higher leagues named after more valuable metals or carbons. Sometimes you’ll fight players in different leagues (especially in team games) but your league placement serves as a measure of how good Blizzard thinks you are at their game. Cleverly, Gold League is actually the middle of StarCraft II’s five leagues, so I can tell people that I’m a gold level StarCraft II player. This sounds a lot cooler than telling them I’m average.
It’s impossible to so easily quantify my World of Tanks skill level. The game tracks a huge number of statistics and achievements, but makes no attempt to actually track any meaningful ones. For example, there is no differentiation made between a player’s win rate in solo random battles and a player’s win rate in platoons, companies, and clan battles. This means it is possible to pad your statistics by only playing with people who are better than you are. Metrics like your efficiency score try to solve this problem by bundling together other statistics (accuracy, damage per game, etc.) to make an estimation of a player’s value. Unfortunately, people respond to incentives, so prioritizing things other than winning tends to cause “serious” players to behave in ways that make them win less.
This may explain why Wargaming.net’s staff exclusively focuses on win rate when talking about player skill. It may not be a perfect measurement, but focusing on it does not cause people to play poorly. In interviews, employees stress their continued opposition to any form of skill-based matchmaker, referring to the concept as being akin to separating players into a low-skill ghetto and noting that a “skill-based matchmaker for standard battles is not a panacea”. This attitude ignores decades of advancement in online gaming that supports the idea that seeding more balanced matches improves the game.
At this point I fear that the biggest hurdle to the implementation of a better matchmaker is the players, many of whom have internalized the idea that the true mark of talent is constantly beating up on bad players. Some of the most notorious (and noxious) posters on the official forum have turned win rate into an obsession, playing overpowered tanks in platoons so as to win as often as possible. It’s as if the Yankees decided that the best thing they could do to improve as a baseball franchise would be to drop down to the minor leagues, where they could win even more often.
Wargaming’s refusal to address the matchmaker is all the more puzzling in light of their quest to improve the new player experience. The decision to start new players with 100% crews was intended in part to combat “seal clubbing”, Wargaming’s unfortunate nickname for the phenomenon of seeing skilled players mucking around low-tiered battles in tooled out tanks. Wargaming seems to think this phenomenon is attributable to a desire to beat up on new players, and I’m sure that is the motivation for some. But there’s an alternate (and far more innocent) explanation. Low tier tanks are fun to play, and having fun is why many people play internet tank videogames. I want to be able to play my awesome tier 2 Tetrarch without stomping all over some poor player who is still figuring out how to aim. Taking player skill into account would improve the game for new players and veterans alike.
There’s a reason Blizzard no longer throws new players into the deep end. Kiddie pools are not ghettos. They are how we train tomorrow’s swimmers.
(Click here for the previous World of Tanks entry.)
David Lydon has been playing games since he was very young, and hopes to still be playing games when he is very old. David posts on the Quarter to Three forums as Dave47. His previous World of Tanks diary begins here.