EV Question.

I am a smalltime bookie. Lets assume someone places a bet with me, Manchester United at +110 three hours before the game. One hour before game, I take a look at Pinnacle and see they have Manchester at +125. I want to max EV and don't mind variance. Should I ever make the bet myself at Pinny at +125 to guarantee profit or would I be sacrificing EV (I think its the lattter)?

Thanks,

Oscar

Depends on the size of the bet and your bankroll. Bookies been laying off 10 cent action on the nickle line since day one.

Quote: Buzzard

Depends on the size of the bet and your bankroll. Bookies been laying off 10 cent action on the nickle line since day one.

Assume bankroll and bet size are not problems. Assume I just want to maximize longterm EV.

Oscar

Quote: oscar33

I want to max EV and don't mind variance.

How do you calculate EV of a sporting event without making a few assumptions?

You can, however, bet to maximize your guaranteed profit.
Assume the original bet is 100, and you bet 100 times X.
If ManU win, you get 125X from Pinnacle, but pay off 110, so your profit is 125X - 110.
If ManU don't win, you lose 100X, but keep the 100 that was bet with you, so your profit is 100 x (1 - X).
These are the same when 125X - 110 = 100 - 100X, or X = 210/225 = 14/15.
If you bet 93 with Pinnacle, your profit is 6.25 if ManU win, and 7 if they don't.

Anything less than 14/15 of the original bet, and your profit goes down if ManU win.
Anything more than 14/15 of the original bet, and your profit goes down if ManU don't win.

Quote: oscar33

Assume bankroll and bet size are not problems. Assume I just want to maximize longterm EV.

Oscar

If you want to maximize your longterm EV, you should never place bets with other books, only take your own (unless you have inside info). Otherwise you can only assume the line is correct at Pinnacle, and they are getting their 5% or 10%. Don't hedge.

Quote: dwheatley

If you want to maximize your longterm EV, you should never place bets with other books, only take your own (unless you have inside info). Otherwise you can only assume the line is correct at Pinnacle, and they are getting their 5% or 10%. Don't hedge.

Ok, thanks. That's what I thought...but couldn't put it into words.

Oscar

In the long term you are better not hedging. However sometimes you have a bet that's too big for the book and (using on-course or old fashioned bookmarking logic) have to lay it off.

On one hand having taken 11/10 and the general price is now 5/4, think yourself lucky.
On the other hand I remember (at Perth races) one bookie had placed a bet of £2500/400 (i.e. at 6/1 - for US people, sometimes on-course big bets are rounded i.e. 10k/1600 = 6/1, 5k/300 = 16/1 etc.) and was quite happy to offer 9/2 while other books had 4/1.

Some rails bookmakers just take a few large bets and, essentially, pray. Win some, lose some, but in the long term they earn 1%.

Whether you hedge depends on your view on risk and how much you are prepared to lose on each event. It's quite a mathematical dilemma, whether to lay off, alter the odds etc, Stay too rigid and you risk taking too many bets and occasionally losing a bundle when all the bets come in; stay too flexible and you land up giving all the advantage away by offering better odds on the other outcomes or laying off too often.

For instance considering offering 10/11 (-110) on both heads and tails. In theory you shouldn't ever offer over Evens, but are you better off taking (i) small bets or (ii) a big bet of £22k to win £20k and taking a big bet the other way at 21/20 (+105)?

Quote: charliepatrick

Whether you hedge depends on your view on risk and how much you are prepared to lose on each event. It's quite a mathematical dilemma, whether to lay off, alter the odds etc,

Where is the "mathematical dilemma" ? Kelly criterion answers this, i.e. it tells you how much loss you can tolerate in relation to your bankroll. The only problem is you cannot find true probabilities of sport events. But you can use price-implied probabilities which I hear are equally good (does somebody care to explain?)