03-20-2022 , 04:53 PM
A hypothetical question that would help me out a ton if I knew the answer to:

There’s a Yes/No bet being offered
Yes +1100
No -900
(true odds are somewhere around +\-700)

Does it make more sense to bet 50% of your bankroll on the Yes. Making a monster bet on a huge +ev spot(albeit one you’re still much more likely to lose than win)

Or

Does it make more sense to bet 10% of your bankroll on the Yes, then immediately betting the other 90% on the No. You’d be mitigating 100% of the risk, but also wiping out a ton of value.

If those were the only two options which is the right choice?
03-21-2022 , 01:41 AM
Sounds like you're looking for a Kelly Calculator.
03-21-2022 , 01:50 PM
id get a loan for as much money as humanly possible and bet 9% of it on yes and 91% on no
03-21-2022 , 05:09 PM
If the OP is talking about trying to arbitrage in-play lines between books, on the surface it sounds like a good idea, but it's like trying to pick up coins in front of a steam roller.

If you put 8.466% on Yes and 91.524% on No you'd lock in a profit of 1.602%. But if you also consider that there's a non-zero probability of something going wrong while trying to hedge, the line changes before you get the other number you want, etc., it starts to whittle away at your expected profits. Imagine you are only able to hedge both sides 90% of the time, and maybe 5% of the time you were only able to get the number you wanted on Yes and 5% of the time you were only able to get on the No. If we use your +-700 suggestion as the true odds of Yes/No occurring then your expected value drops to 0.798%. If you think you'll be able to successfully make the hedge 80% but get stuck on just one or the other side 10% each, then the arbitrage betting idea now has negative expected value. Assuming the betting limit was high enough for you to actually get down 91.524% of your bank roll, do you really want to risk losing that much on one shot if something goes wrong?

If you've accurately judged the fair line for Yes/No at +-700, then Full Kelly would tell you to bet 4.55% of your bankroll on the Yes. 87.5% of the time you'd lose it, and 12.5% of the time you'd win 4.55% * 11, for an expected value of 2.28%.
03-21-2022 , 11:04 PM
Excellent post^^and almost spot on w what I’ve experienced so far. I’m actually “arbing” at the same book, which when I take a step makes it way too obvious what I’m doing so I’m just going full Kelly and not buying back. I was making my initial plays at the kiosk then buying back on the app but still obviously looks sus when both bets come in within a second of one another.
05-09-2022 , 02:23 PM
Been a while since this was posted but yea, arbing in game is playing with fire for large sums - you’ll be too slow to get both sides way more than 5% of the time IME (though obviously this is dependent on the sport).

If this is a regular occurrence better to just take the side you lean to and don’t take huge positions. Unless you really have no clue it should have a higher expectation than the arb would.

m