Thursday, March 31, 2016

Intentional Concessions in Magic: the Gathering

There's an interesting controversy in tournament-level Magic these days around the concept of "intentional concessions".

Background

Tournaments in Magic follow a standard format: X rounds of Swiss, followed by a Top 8 playoff. In Swiss matching, players with the same record in the tournament so far are matched with each other. For example, for Round 4, the players who are 3-0 play each other, the 2-1's play each other, etc. So you're always playing an opponent who's done roughly as well as you are. The top 8 after the Swiss portion go to the playoffs.

In the past, in the last rounds of the Swiss portion, the players who are at the top often "intentionally draw". With the draw, both players' records are good enough to qualify for the Top 8. And they get to save some time and the possibility that a loss might knock them out of the Top 8. Though getting knocked out is pretty unlikely, due to the way tie-breakers work.

The intentional draw is regarded as a fact of life in tournament Magic. It may not be the best scenario, but it's equal on both sides and has a fairly neutral effect on the tournament on the whole.

Current Controversy

At the pro level, some players are now asking for intentional concessions in certain situations, where one player deliberately takes a loss, not just a draw. For example, say Alice is 8-0 coming into the final Swiss round, and she gets paired down with Bob, who is 7-1. A player will need at least an 8-1 record to get into the Top 8. So regardless of whether she wins or loses, Alice is a lock for Top 8.

Bob, on the other hand, must win this round. Even a draw, which will bring him to 7-1-1, won't be enough. Bob asks Alice to intentionally concede the match, since it makes no difference to her.

Now, the non-Pro community is strongly against Alice deliberately losing. It is completely against the spirit of competition. It screws over Carl, who ends up in 9th because Bob got a "free" win.

However, the Pro community is a bit more torn on the issue. To see why, you can think of them as being in an iterative Prisoner's Dilemma where they can co-operate or defect. The optimum strategy in these types of situations is to cooperate. After all, in next tournament, maybe Alice will want to ask someone to intentionally concede. If she defects first instead of cooperating, she can expect future partners to defect to punish her. And hardcore gamers are the type of people who will very strongly flock to the optimal strategy in a Prisoner's Dilemma game.

It's also really hard to outlaw intentional concessions. For example, there are good reasons to concede. Maybe you need to leave, so you concede the current match. And it's really hard to tell when someone is deliberately playing badly. Gamers are very good at obeying the letter of the law and completely evading the spirit.

Solutions

In my view, the main reason this is a problem is because seeding in the Top 8 does not really matter. It's really hard to predict who your first opponent in the Top 8 will be. Plus the 8th person's deck is only slightly worse than the 1st person. That difference would be swamped by the variance in the game of Magic itself.

If seeding mattered, Alice would be hurt, perhaps significantly, by taking a deliberate loss.

Right now, the Top 8 playoffs are a best-of-5 match. My suggestion would be to give the higher-seeded player a game in hand. So the higher seed only needs to win 2 games, but the lower seed would need to win 3 games to take the match and advance.

Of course, this significantly slants the matches in favor of the higher seed. But that in turn makes it vital to get as high a seed as you can in the Swiss portion of the tournament. Perhaps this solution could be toned down to only apply to the first round of the playoffs, essentially giving the Top 4 an advantage over the 5th-8th place.

8 comments:

  1. This reminds me of a similar issue I read about a while ago wherein the competitors in fighting game tournaments would often agree to split the winnings in the final round. So the last two would just essentially throw the final match and share the prize money. Perhaps risky, but it did happen. Partly that was because at the time tourneys were pretty much an all-or-nothing deal which is hard to make a professional career off of. Still, it hurts the game as an esport because people aren't going to watch games if they know the "best" won't actually play at their best in the find round.

    I like your idea though; you basically need to give motivation to always be wanting to win, even if your spot appears locked.

    ReplyDelete
    Replies
    1. Splitting the winnings is pretty common in Magic as well, and the Top 2 just play for the trophy and Pro points. But they still try to win the last match.

      Delete
  2. The old Romans had a solution for that. If the audience didn't like the effort the losing gladiator put into his fight, he got eaten by a lion.

    ReplyDelete
    Replies
    1. Hah. But if the gladiators are good enough actors, I'm sure they could "sell" the fight to the audience.

      Delete
    2. But if they're good enough actors, people might not even notice that there's an intentional concession happening.

      Delete
  3. Seeding decides who is on the play and who is on the draw in the top 8. It's not as significant as your solution, but seeding does matter somewhat, especially in formats with very fast games where play/draw matters significantly.

    ReplyDelete
    Replies
    1. It does matter a bit, but only for the first game of the match. And it doesn't guarantee the high seed victory.

      I think it's clear that getting to choose play or draw is not a strong enough advantage to encourage players to avoid intentional draws or concessions.

      Delete
  4. Given how high variance the games are, and the fact that some decks are much stronger post-board, giving away a free win to the higher seed is far too high a bonus. It would essentially make the first seed a near-lock to win the entire tournament, at which point why bother with the top 8?

    I think you're trying to 'solve' what is essentially a gentleman's agreement between professional players. The odds are that #9, Carl, is also a pro-player, and has been the beneficiary of the same intentional concession in the past. Being able to intentionally concede in a situation where a win has no benefit for you is much better than the alternative of trying to clearly throw the game (I'll mulligan to 2, why not) or other forms of under the table collusion.

    ReplyDelete