Backhand of Justice

Gamasutra’s Chris Beault claims that Starcraft II’s ladders are its biggest flaw. Is this unfair criticism, or does he have a valid point?

For those unfamiliar with the system, the game divides you into one of five tiers: platinum, gold, silver, bronze and copper. You are placed into one of those tiers based on several placement matches, and within each tier there are many 100-player divisions.

Chris claims the lack of global tiers are deceptive. He claims the opacity of the system which promotes or demotes you between tiers is harmful to the game. He also claims the way the game decides whether you’re favored or not compared to another player is confusing.

Flaws with the ladder?

Let’s quickly review the reasons he claims the ladder is flawed.

The first aspect criticized is the lack of global tiers. Let’s just assume that the game sells about as well as the original and there are over 10 million copies sold. Instead of thousands of 100-person divisions making up each of the five tiers, Chris believes there should be global tiers.

Because each division is made up of 100 players, if you’re one of the better players in the silver league you might be ranked 25th in your division out of 100. If there were global tiers, on the other hand, the equivalent position would be 500,000th out of 2 million silver tier players.

Chris claims that the divisions are deceptive because if you’re 1st in your silver division, you’re not really the ‘best player’ in the silver league because there are thousands of other divisions; this artificially inflates the ego of gamers and is wrong.

Here is the Blizzard Starcraft II team in a design meeting (with Jace Hall). This is why the matchmaking system has come apart at the seams; too much drinking

The second criticism of the ladder is the way in which it occasionally matches you with players in different tiers. The point of the ladder according to Chris is to match you up against players of similar skill levels, but what’s the point if it’s going to occasionally match you up with players in the tier above or below?

The final criticism is that the system that promotes or demotes you to the tier above or below is opaque (and intentionally so), and this is harmful for gamers.

Does Chris Beault have a point, or is he trolling for hits?

Now let’s analyze those points and provide a counter-argument. First is the concept that separating out each tier (league) into separate divisions is somehow deceptive.

It’s difficult for players to really appreciate progress either up or down the ladders if there are 2 million players per tier. Let’s say you win a game and go from 40th place to 30th place on your 100 person ladder. That’s real progress. If you move from 550,000th place to 500,000th place though, are you really making progress? Technically, sure, but it probably doesn’t feel like it. There are half a million players in your league superior to you.

Matching you against opponents of a comparable skill level is the most important aspect of the matchmaking system

While Chris Beault says this is artificially inflating the ego of gamers, I’d say that one of the primary purposes of videogames is to offer concrete accomplishments where effort directly translates into reward. Outside of videogames, it’s rarely as concrete.

Is it really deceptive to players to divide them into divisions within each league? Does Chris believe there are players out there who see 5 tiers with 100 people in each tier and jump to the conclusion that every player believes there are only 500 people playing the game?

There is even precedent for this type of separation: World of Warcraft has multiple servers, each of which contain a small fraction of the total player base. While this is partially because of server limitations, it’s also because it’s difficult for players to come to terms with numbers in the millions.

You might not notice man, but your buildings are on fire. Might want to check that out

Chris claims it’s deceptive; I say players can easily deal with 100 player divisions. They understand if they’re 30th in their division out of 100 players, that’s equivalent to a percentile. They’re better than about 70% of the players in their league.

The second criticism is that the leagues hold no purpose because you’re occasionally pitted against players in the league above or below yours. I disagree with this conclusion. When you are approaching the top or bottom echelon of your current league, the game pits you against players in the league above or below yours respectively to see if you should be promoted or demoted.

If you’re in the top 5 in your division of 100 in the silver league, the game is going to pit you against players in the gold league frequently. If you lose most of your games against them, then you probably deserve to stay where you are. If, however, you still win the majority of your games, then eventually you’re going to be promoted to the gold league. The same is true if you’re at the bottom of your league; if you still lose to players in the league below you on a consistent basis, then you’re going to be demoted.

It's kind of unfair that this guy's base is surrounded by lava. Usually people can just run right into mine and kill me

The point of the ladders is to match you with players against whom you will win half the time. The pace of games and the strategies used between leagues are significantly different. Just because you win most of the time in your league doesn’t necessarily mean you’re going to cut it in the league above yours, or vice versa.

Chris’ final criticism is that there needs to be more transparency in the system that promotes or demotes you. He also wonders why it’s not based solely on wins; after all isn’t that what’s important?

For those that haven’t played the game, after a match ends there is a scoring summary based on resource usage, units created and destroyed and structures destroyed. The exact breakdown of points isn’t revealed, but generally it lines up with the player who performs the best overall having the highest score. This isn’t always reflective of who is the best player; in some 2v2 games for example, a team with both attack and cripple one ally, reducing his ability to perform for the rest of the game. Even if he’s the best player he probably won’t end up with the highest score.

Placement and promotion / demotion is likely partially determined by this score as well as the win/loss record. This is completely reasonable to me as the game only uses 5 placement matches to put you into a league. What if in every match you barely squeak out a win against potentially terrible players, are you really deserving of a spot in the platinum league? You’ll just be crushed until you’re demoted down to an appropriate level (if you even stick with a game where you constantly lose for that long).

If, on the other hand, you crush all your opponents and end the game with 2-3x their score, then you’re likely a more proficient player. It’s not an exact science, but it’s good enough for me. If Blizzard broke down exactly how the game rated you, then people would start to game the system.

Plug and play

Blizzard is trying something incredible with Starcraft II: perfect matchmaking. The long term viability of their game is going to hinge on its success, but I encourage the attempt whatever the result. Being able to press a button that says ‘find game’ and having it swiftly match you with another player at your skill level is incredible.

What Chris Beault wants is the status quo of a decade ago: gamers would create a server and find players of their own skill level on their own. That’s like installing drivers for your new mouse before it would work, or entering the IP address of the other computers on your network. Everything is plug and play now. You just log on and you’re good to go. Why shouldn’t game matchmaking function in a similar manner?

Gamers criticize Starcraft for being too elitist, or inaccessible, or having regressive design. The game is the highest selling RTS of all time though, and one of the best selling games ever. Whatever Blizzard is doing is resonating with gamers, and the interest across the internet for the sequel is sky high.

Not only is the game still in beta and therefore exempt from criticism, but its level of polish is incredible. The matchmaking system could use minor tweaking but at its foundation it is certainly not a flaw.