Sunday, February 6, 2011

Tournament Structure Deep Dive: Part II - Voluntary Comp?

I noticed a very interesting thing in the results of the tournament yesterday which I would like to share.

The tournament used the NOVA system for sportsmanship scoring: Rank your opponents in order of enjoyment. Most people will get average Sports scores under this system. It minimizes sportsmanship drama and "gaming" the system (though it's possible to a degree). But you really only benefit if ALL of your opponents had a particularly high opinion of you, and you are really only punished if ALL of your opponents had a particularly low opinion of you, especially in a three round format. I like it a lot.

Well I noticed that when I looked at the Sportsmanship scores, the "average" scores were about the same for the first and second rounds, but quite a bit higher during the third round. I also noted that when you look at the number of "3's" that were scored (meaning that that game was the best for your opponent out of the three rounds) there were TWICE as many scored in the third round as in the first or second.



What can we tell from this? Well, nothing for sure. It's possible that there is some temporal bias at work here and that people are just remembering their most recent game the most fondly. I don't believe that's the case though. I think that what this represents is that as the tournament moves forward there is a nice "filtering" effect which is separating players by skill level and army strength. And generally speaking, I have found, the more closely matched you are with your opponent the more BOTH of you enjoy the game. You are on the same wavelength and usually have the same mindset. Even with top versus bottom pairings within brackets we were finding that as the day went on people were playing people that they enjoyed more, because by and large they were playing like minded players.

After the tournament, an interesting suggestion was made to me that I need to spend some more time thinking about - Seeding the tournament. It was suggested that I match the best players against the best players, based upon my own judgement. Now, I am not convinced this is the right way to go for several reasons. For one, this makes it more likely that good players will get knocked out early. This is one of the reasons that we use top versus bottom pairings - we want to ensure that the best players move on and those that got lucky pairings get put back down in the brackets they belong in.

But then the suggestion was made to allow players to simply "elect" to be placed into a non-competitive bracket. That they essentially take a "ghost loss" before the tournament even starts to ensure that they are taken out of the shark tank and put in with more like minded players. This is an interesting concept. I know that a lot of tourneys have tried to achieve this with round one comp based pairings. But the opportunity for bias there is so great that I don't think that's necessarily the way to go. And again if I was to do intentional pairings I would be pairing the worst baby seals against the meanest sharks - INTENTIONALLY - to ensure that people got separated out into their proper brackets. Trying to do the opposite - separate out the non-competitive folks into their own bracket, would likewise be very unfair if the TO did it. Who am I to say that the player with the three max squads of Ogryn doesn't WANT to be in the competitive bracket, trying for Tournament Ace or Best Overall? It's certainly not fair to simply ASSUME that he doesn't and prevent him from even trying.

But if that player voluntarily takes a "ghost loss" before the tourney begins, because they don't WANT to play against the sharks, with the FULL knowledge that this is going to take them "out of the money" (but with their own, relatively small prize pool mostly based upon certificates and recognition to prevent sharks from simply opting into the hobbyist pool to gorge) would that work? Could that work?

I'm not sure. I love what Mike Brandt has to say about "incidental comp." I think that a GT where you have two days to first separate players into brackets by skill level and then allow them to compete within their brackets for prizes/recognition is brilliant. But with a three round tourney we simply don't have the time, and the pairings of like-minded players really only hit their peak during the last round.

Can we do better? I'm not sure if we can. I'm trying to think through the possible repercussions of such a thing. I would love to hear your thoughts.