"In its current state, it surprisingly predicts a close win for India if it bats first and a comfortable win if it bats second."
Is that a great prediction or what! For the first time yesterday we decided to predict the win/loss chances. And we got it bang on target immediately!
We did a very ordinary job with Sri Lanka's scores this time. We also got all the half century makers wrong. We are pleased we got Sehwag's score and quite a few Lanka batsmen's ranges pretty correct though. The comparision table:
1 comment:
Hello Prashant,
Its a good effort; congratulations :-)
Just wondering: how big a feature set (the predictor variables i.e.) are we talking here ? I believe that in sport/stock prediction there are a number of subjective factors right from the weather conditions to economic, political factors and what not! isn't it? From where did you get the training data to train your model? I am sure you might know about this but if you don't then the following might be useful:
www.howstat.com
Do keep us updated about your progress. Quite interesting!
Thanks,
Anurag
Post a Comment