Thoughts on a new tournament class prize system

[aibretty]

I am hosting an unrated tournament on November 1st in Chicago (CUSS WORDS!) and have been working with Dave Turissini on a class system that would reflect a player's performance relative to their expectation for any given game.  The system uses all of the score data available since the new dictionary and calculates an expected win value relative to

Comments

[porosus]

It's a really small point, but the program will be written in perl. I may call some R scripts to generate some plots or histograms for one of the monitors in the room.

[magratheazaphod]

supercool. keep us posted on how it's going.

[quinquennia]

Just pointing out that most people will aim to win a game rather than get excessive amounts of points, but I'm okay with a prize for exceeding expected perfomance.

David Boys did a handicapping tourney in Montreal, and he figured that each +100 rating points pretty much translated to +10 Scrabble points, and handicapped players accordingly, starting the expected loser with enough of a head start that everyone would be as close as possible to a 50/50 chance of winning. That was with the original TWL pre-QI and ZA lexicon. I'm curious to see points/ratings ratios now. Thanks for posting. I'll check them out in a second.

[quinquennia]

Hmm. I can't get to your document via your tinyurl address.
Document does not exist page.

[porosus]

I don't expect people to change their playing style much in pursuit of a class prize. Also, people were playing to win in the historical games so the comparisons should be justified.

10 score points per every 100 rating points isn't very accurate (it should be more than 10 points) once the rating difference is greater than 100. I also don't like the idea of changing the winner based on score handicaps. In this system the score ranges from 0-100 as opposed to awarding either 0 or 100 points. So you're able to record more information from each game and get a more accurate assessment of everyone's performance.