Sunday, September 6, 2009

Jeff Ashton explains his GOTW judging methods



I'm too indecisive, and this took too long, but I think I have a good judging system down that IS PROVEN BY SCIENCE!

Overall I was looking for two things:

(a) Makes people like chess more. i.e. entertaining attack or creative play
(b) Good quality game (not too idiotic). Basically provides a cool "luckometer" trend analysis/Rybka infinite analysis tests.


Step 1

Intuitive Rough Draft: Made a quick "feeling" order. Ranked them without too much thought. Malcolm Gladwell "Blink" style.


Step 2

Used my super scientific simulations (PROVEN BY SCIENCE) and basically paired off games against other games using Swiss System. If they were roughly equally interesting, I'd give it a draw. Note, this is how I pick where I go to vacation and stuff. Usually this ends up matching my intuitive rough draft somewhat but kind of ends up in a tie in third place usually.

Note: I do use ONE "Wild-card" trick that makes this simulation slightly more randomized. I can explain more later if necessary. This is so all the people who are trying to CRACK MY CODE (scientifically proven formula) will not be able to.


Step 3

Run them with Rybka 3 and look at the cool "luckometer" evaluation profile graphs. Pretty graphs = cool


Step 4

Go back and think about my criteria again. Was it fun? Do I like chess more? What would Joe McDumbDumb think about this game? What would Boris McGoodPlayer think?


Step 5

Usually end up going with pretty much my intuitive feeling and realizing some games that I thought were good, were actually kind of boring and lame.


Step 6

Mess with the "sort" order of the database so it matches my rankings somewhat.



Luckometer graphs for my top six games from Week 1:





Ippolito vs Charbonneau
























Zaremba vs Esserman
























Perelshteyn vs Vovsha
























Ramirez vs Mitkov
























Kudrin vs Shabalov
























Becerra vs Bartholomew

























Please give me your feedback on my methods!


6 comments:

Elizabeth Vicary said...

Your graphs are v funny!
Do you share my outrage that Greg refuses to link to my last two blog posts?
I was/am very demoralized.
www.lizzyknowsall.blogspot.com

Anonymous said...

Whoever approved Jeff Ashton as a GOTW judge should probably not be allowed to appoint any judges in the future. His banter is not witty and really it just disgusts me and makes our league look juvenile. Grow up, Jeff...you are not a teenager anymore!

Anonymous said...

you have very poor taste, stupidface 11:45. grow up yourself, learn to laugh a little.

Craig said...

Maybe the game of the week contest should be done without knowing who played the game? Find a couple strong players who don't watch the games being played then they are submitted anonymously. My guess who plays a game weighs into peoples decision making?

Elizabeth Vicary said...

probably it would be hard to find strong players who care enough about the USCL to put hours each week into judging the games, but who don’t want to watch?

Jeffrey Ashton said...

I have learned two things from this USCL judging experience:

1) USCL fans really care a lot about game of the week! Don't forget: Fan is short for fanatic.

2) This judging process takes a lot of time and energy. If a judge isn't into regularly studying chess games, then this job is hard. Luckily I am always looking for new material to share with scholastic players so this is an enjoyable experience.