Trust the physicists

Geekman's department is running a pool for the cricket World Cup. These are the betting rules, as sent around to everyone by email:

You email me your prediction of the net run rate (NRR) for the match before 11pm each night that there is a match.

The NRR of the match between team A and team B is calculated by RR(A) - RR(B), where RR(A) = runs(A)/overs(A), where overs(A) is the total number of overs that team A is required to bat. For example, if it is an N over match, then overs = N, unless they are batting 2nd, and win the match in less than N overs. In the case of rain and each team gets a different number of overs, the run rate is the ratio of the actual runrate to the target RR (or score) calculated via the duckworth-lewis system, multiplied by the run rate of the other team.

If you correctly predict the outcome of the game (who wins), you get 1 point (if it is a tie, and you predict a win, you get 0.5). If you predict a tie, and it is a tie, you get 1 point. If you predict a tie, and it isn't tie, no points for you, loser!

You also get up to 1 additional point for predicting the NRR correctly. Specicfically, you get P points, where:

if |diff| < 2
P = 1 - |diff|/2
else
P = 0
end.

where diff is the difference between your predicted NRR and the actual NRR.

The points are doubled for the semis, and tripled for the final.

If you fail to bet, no points for you.

Each player puts in $5. The winner at the end of the world cup gets a prize (yet to be determined) valued at 75% of the total input. 2nd place gets 25% of the total input.

Now will you join me in laughing at them?