Arimaa Forum - Print Page


    
      
        Arimaa Forum
        (http://arimaa.com/arimaa/forum/cgi/YaBB.cgi)
      

        Arimaa >> Events >> Arimaa WC Final Player Rankings
        
(Message started by: mistre on Mar 19^th, 2014, 3:16pm)

Title: Arimaa WC Final Player Rankings
Post by mistre on Mar 19^th, 2014, 3:16pm

Should Fritzlein and 99of9 both lose in this round - according to the rules - Fritzlein will get 3rd place due to his 7 wins vs 99of9's 6 wins. (If I am reading the rules correctly). If they had been tied in wins then there would have been a playoff. Because Fritzlein lost head to head to 99of9 during the tournament, it is too bad there wouldn't be a rematch for 3rd place.

Now if one or both of them can pull an upset, then things will really get interesting.

Title: Re: Arimaa WC Final Player Rankings
Post by Fritzlein on Mar 19^th, 2014, 5:47pm

It's too bad that I will place ahead of 99of9 in this scenario, because he will have had arguably a tougher schedule despite the fact that I played an extra game. Removing our game against each other and removing common opponents chessandgo, browni3141, and Hippo, our schedules line up as

99of9: chessandgo, Tuks, Thiagor, odin73, aurelian
Fritzlein: Adanac, supersamu, clyring, RmznA, SilverMitt, Centauro

This is a case where performance rating would be a better judge than number of wins; 99of9 being forced to play chessandgo twice is just brutal. But regardless of what happens to me and 99of9 this round, Adanac has clearly been robbed by the current rules to get 7th place instead of 5th place. Comparing Adanac's schedule to supersamu's, removing common opponents browni3141, Fritzlein, rabbits, and harvestsnow, the schedules line up as

Adanac: chessandgo, Tuks, odin73, Arseface
supersamu: Alfons, omar, BrendanM, ikalyoncu, Centauro

Further removing roughly equal opponents leaves

Adanac: chessandgo
supersamu: Alfons, ikalyoncu

I submit that for players at the level of supersamu and Adanac, there are more expected losses from one game against chessandgo than from a game against each of Alfons and ikalyoncu.

It was my idea in the first place that byes shouldn't count as wins, but in three years of experience with this format it has become clear that I was wrong. See also omar/browni3141 in 2013. Balanced against these three cases where not counting the bye as a win was unjust, there have been no cases where counting a bye as a win would have been an unfair benefit by promoting a player with a weak strength of schedule. Therefore at the very least we should count byes as wins starting in 2015.

If the reverse case ever arises, i.e. where the counting a bye as a win is an injustice in favor of the player receiving a bye, then we could consider scrapping the ranking by wins entirely, and move straight to performance rating for all places below the champion. Another reason in support of straight performance ratings is that it isn't only byes that mess up strength of schedule. Consider Fritzlein/Nombril in 2012 where we each had a bye but I got eliminated a round earlier. Removing our four common opponents hanzack, chessandgo, Adanac, and rabbits, our opponents were

Fritzlein: chessandgo, Harren
Nombril: Tuks, ocmiente, woh

Other than this one issue, though, the tournament format has worked fabulously well. I don't know any other way in which I would tinker with it.

Title: Re: Arimaa WC Final Player Rankings
Post by 99of9 on Mar 26^th, 2014, 7:40pm

on 03/19/14 at 17:47:30, Fritzlein wrote:

It's too bad that I will place ahead of 99of9 in this scenario, because he will have had arguably a tougher schedule despite the fact that I played an extra game.

Ah well, fourth is still my equal best performance, and I'm pretty happy with how I played this year, so it's no big deal. And anyway, I've held on to my lead in the *Most ever WC losses* category :D

Title: Re: Arimaa WC Final Player Rankings
Post by aaaa on Mar 27^th, 2014, 1:45pm

No way a bye should count as a win. What if there are three players left, each having received an equal number of byes and being down to their last life? Would it then be fair that whoever is excused for the next round to be guaranteed at least second place?

Title: Re: Arimaa WC Final Player Rankings
Post by Fritzlein on Mar 28^th, 2014, 12:18am

on 03/27/14 at 13:45:49, aaaa wrote:

Potentially yes, if that person had played the toughest schedule to get to that point.

I do get that it is an unfair scenario if all three of the remaining players had an equally tough path to that point, and exactly that fear motivated me to insist that byes not count as wins in the first place. Now, however, I see how it is quite possible that after, say, a dozen rounds, one of the three players who all have nine wins, one bye, and two losses has faced much tougher competition than the others and actually deserves a bye into the finals while the other two duke it out for third place.

In any case, if you don't like counting byes as wins, how would you suggest we address the unfairness that 99of9 and Adanac experienced in this tournament, whereby getting a bye punished them both?

Title: Re: Arimaa WC Final Player Rankings
Post by 99of9 on Mar 28^th, 2014, 1:13am

on 03/28/14 at 00:18:15, Fritzlein wrote:

In any case, if you don't like counting byes as wins, how would you suggest we address the unfairness that 99of9 and Adanac experienced in this tournament, whereby getting a bye punished them both?

One option would be to use one of the performance ratings. Preferably one without any dependence on the player's gameroom/WHR rating.