Scrabble Cubes Database

[rahnnyboy]

All,
I lost most of my scrabble cubes scoring data...old comp died.

Looking to repopulate long term trends of expert scrabble cubes scores.

If any of the top guys want to contribute, send me screenshots of ur filled tourneys or the following raw data:

tourney ID (only for indexing)
entry fee
top prize
players and scores

This is open source, so I will share all the data openly.

NOTE: This ONLY works if you give me all the data, not just games you won

Stats can be compiled on request.

send data to rahn_the_cheme@hotmail.com

Thanks.

[Raptor]

Hey rahnny, do you still have your hard drive? Was it making noises, any more details? If you're able to ship it to me, I might be able to recover data for you, that is assuming you were just going to toss the drive anyway and even if it doesn't work you're okay with it?

[rahnnyboy]

sorry man, traded in for parts for discount on new comp.

I have collated a weeks worth of my own games, but don't have microsofts version of excel. so still working out how to do the formulae. I will likely run PCA on it eventually (once i have enough data), so I can ignore excel and use a better piece of visualization software.

No one else has sent anything .

just FYI, for multiple entry tourneys, I collect ONLY the top 2 scores the person put up. Any scores below 500 are ignored as likely "punts". and I do not record the daily delights.

Keeping the collection to "expert" players for now just to reduce the amount of keying I have to do.

[Raptor]

oh that sucks, it might have been retrievable.

So are you using OpenOffice.org or something? If not, that might do what you need.

So what all does this data tell you anyway?

[rahnnyboy]

openoffice is what i'm using. I can do a lot of array math with it to get averages and stuff, but i'm too lazy to do PCA in excel.

what does the data tell you?
It points out statistical anomalies...and helps me present a case to WW if I feel something is unfair based on the following:

1.) Description of how long-standing experts scores are distributed
2.) Comparison of player X's distribution of scores to the distribution of expert scores
3.) Detection whether or not Player X's distribution belongs to Expert distribution using statistical tests
4.) Prediction with confidence intervals whether or not Player X is an "Expert" player
5.) If Player X is data shows that they are a "better" player than Expert pool...better being defined by raw average, consistency, or other causal relationships (performance in high money competitions), then player X is anomalous
6.) Anomalous players can be assumed to be cheaters unless they can prove otherwise.

[bmy1]

all i have are some top scores from specials from earlier this year, and whatever I've played in the past 45 days.

[Raptor]

Now if only WW's fairness department had the commitment you do. That's great work.