[pmacdaddy]

After reading this thread and thinking I understood the consensus, I took a look at the results and got promptly confused.
I believe it is because win $ are factored in.

If the Win wager is removed (as it is equal between scenarios and therefore a wash) the results will more clearly show when the place wager paid and the exacts wager "failed".

When our horse runs 1st or 2nd and the exacta with fav is not hit, the Place player will have earnings and exacta player will have a zero for that race.

Under the current method, you need to back out Win payouts to see that the exacta failed in the cases where our horse ran 1st. Without the win $ in, these will stick right out.

In any case this is very interesting.

Next I vote for the Triple Partial Wheel versus Exacta Box study...