Dominion Strategy Forum

Please login or register.

Login with username, password and session length
Pages: [1]

Author Topic: Tiebreakers and first-player advantage  (Read 3306 times)

0 Members and 1 Guest are viewing this topic.

theory

  • Administrator
  • *****
  • Offline Offline
  • Posts: 3017
  • Isotropic username: theory
  • Respect: +4258
    • View Profile
    • Dominion Strategy
Tiebreakers and first-player advantage
« on: June 17, 2011, 08:55:43 pm »
0

Source


Quote from: Donald X.
You don't actually need a tiebreaker for most or all of a tournament, for most games. For example Magic has no tiebreaker until the cut to the top 8. A draw is a draw. They do need a tiebreaker to get a specific number of players for the single elimination rounds, but that's because the matches are strictly two players; if your game handles different numbers of players, you can cut to wherever the ties stop. See http://en.wikipedia.org/wiki/Swiss_system_tournament
We did not provide tournament rules, because Jay doesn't like including rules variants, and he wasn't ever going to run a tournament. Hopefully he's not mad I linked you to this wikipedia article. I am not specifically endorsing a completely usable tiebreakerless Swiss system for Dominion tournaments, I am just pointing out this interesting method used for other games. Isn't it interesting!

Outside of tournaments it is totally fine to have ties; I felt no compulsion to determine a winner at all costs. Furthermore I feel that there is no advantage to going first, other than the potential to get an extra turn. Getting to play an attack card first is really very minor, and similarly going later means you know what the other players buy on their first turns before deciding what to buy yourself. And not vice-versa! I have had games where an unusual first-turn buy from me resulted in people scanning the table to see what they'd missed, before they'd had a chance to buy something they'd regret. If I knew I was going to get as many turns as you, I would choose to go last.

Getting an extra turn is definitely an advantage. But going first is only an advantage if you get that turn. If you go first but don't get an extra turn, a tiebreaker going to the other guy amounts to just randomly stealing a deserved tie from you. You didn't play worse, you just randomly went first. A tiebreaker there doesn't solve the problem of an advantage due to going first (and does nothing for cases with no tie); it just randomly awards victories in some cases where there was no such advantage. It's the same as flipping a coin.

I know some people will find this hard to see. "But he had the advantage of going first!" Going first is only potentially an advantage. If you don't get the turn, it's not an advantage. At the start of the game, you want to go first, because who knows, you could get that extra turn, but at the end, we know whether you got it or not, and there's no prize for not getting it. Say we're walking to a bar when you spy a lottery ticket on the pavement. Huzzah! We get to the bar and each buy a round of drinks. There's a TV at the bar and they show the winning numbers. Aw, your ticket was a loser. Well, one last round of drinks before we leave. Someone says, you should buy this round, because you had that losing lottery ticket. Does that seem fair? That is not what fairness is.

The existing tiebreaker does not break most ties. It does however break all ties that can be broken fairly, without warping strategy. If you require a tiebreaker at all costs - i.e. even when it's unfair or warps the game - then that's up to you. There's nothing I like there.


Where going first is really significant is where it gets you an extra turn that you win the game with. It doesn't matter what tiebreaker rule you have there; it isn't invoked. The rules address this case by having the winner go last next game. My friends and I always play a bunch of times in a row, so that's always seemed fine by us. The bias we've noticed is, the better player wins more often. Despite going last! It's perplexing.
Logged
Pages: [1]
 

Page created in 0.112 seconds with 20 queries.