I had a 1.03 beaten the other day, and the winner traded at 1000, but alas I didn't have that one.
I actually did a bit of research on the in-play situation a while back.
I wanted to see just how the market reacts and just how viable it was.
I imagined, you wouldn't get matched on the fallers, the unseated riders and those that came out sideways etc, so would laying below Betfair SP be profitable?
I considered all horses at first and ran it over the data.
The longer the price of the horse, the worse it was, the more you were laying winners.
However, one thing really stood out.
Horses that start odds on, including fallers, no hopers, and those that just didn't get a clean break (this percentage was negligible) was still profitable to lay below Betfair SP, providing you don't get greedy.
There were 820 horses that ended up odds on at Ladbrokes before the jump in the sample, that also traded at or below the final Ladbrokes price
460 of them won.
So a 56% win strike rate ( higher because of the UK jumps).
That's a 44% lay success on odds on favourites without including the ones that never traded at or below Ladbrokes final price.
The average dividend was $1.61 based on Ladbrokes prices
So that's a 9% POT laying them at those prices, but before commission.
It purely works because they have a very high place strike rate, and the percentage of unplaced horses, that did not trade lower at some point in the race, is tiny.
In fact it's profitable just above even money, where it is a clear cut sole favourite.
When going further out in the market, the further you go on price, the more you're laying winners.
I believe I have only lost 1% of the bookies premium when not getting matched on the one's that didn't trade lower, over a large set of races.
the 52 out of 100 are the last set this month on Aus and UK races combined, that I've covered.
Of the 820 in the sample, you'd have won 78.1 units before commission.
I allow for commission in my prices, but the bets are staggered.
I actuallly got the idea after reading the Betfair forum.
|