Puzzled by the maths

I've been looking over my figures, and I'm wondering if they do in fact give me an edge at all!

I said that there's an 87% strike rate with these horses, and that the average price is 8.63, and that winners' prices average 6.42. By my arithmetic, that means that the losers' average price is 8.96.

If I lay 100 horses each to a liability of £100, I expect 87 to lose. I calculate this to provide profit of (100 / (8.96 - 1)) x 87 = 1092.96 less commission, say £1038.32. The 13 winners would cost me £1300. That's a net loss of £261.68. Not much of an edge?!

So why did I think it was? Need to check all my figures again, or maybe I'm missing something.

On a possibly more positive note, I checked back the last five years data, and the strike rate is consistently between 87% and 90%. I used an Industry SP of 12.0 (11/1) as the maximum price, because that was roughly what the average 16.0 BSP equates to.

Also, and this will need checking too, it's possible that these horses are worth laying to place as well. By adding a second laying system, it would help with another area that would increase my edge (if there is one).

Have a low commission rate
The profits I quoted for Maria's Laying System are probably underestimated by a fair bit. She deducted 5% from her profits even though she was probably on a rate of 2% commission. The difference could turn a losing system into a profitable one.

----------------------------------------------------------------------------------
I've sussed why the figures don't tally. Average prices can't be used in that way.
Pheww!