Quote:
|
Originally Posted by Bill Dozer
Right. Maybe our resident mathematician can give us the odds of a tie on both the record and best bets. Ganch?
-edited for political correctness- 
|
This is actually a nontrivial problem for several reasons, not least of which is that there exists correlation between players' results.
Anyway,
completing discounting that correlation and
completely ignoring pushes ATS, I ran a small Monte Carlo simulation over woefully too few trials. This implementation is extremely slow but if anyone cares for the source code just PM me.
The results assume 175 players each making 4 picks including 1 "Best Bet" over 16 weeks:
Code:
N 65,536
Mean Winning Point Total 27.34
Mean Winning Best Bets 11.50
Std Dev WPT 3.98
Std Dev WBBs 1.22
Any Tie Likelihood 6.17%
2-Way Tie Likelihood 5.68%
3-Way Tie Likelihood 0.46%
4-or-More-Way Tie Likelihood 0.03%