Calculating/Comparing Edge

[Arimathea]

This has provoked some good discussion but I still don't know the answer!! :( I thought I'd post on this forum where there are lots of experts! Please see below and thanks in advance!... Here's a confusing one (for me at least!) Do you compare edge in terms of probabibility difference or difference in the odds? E.g. Scenario 1: We back player A at 1.5 knowing that his true odds are 1.01 (Giving us a 32% advantage in probabililit but only a 0.49 advantage in odds) Scenario 2: We back player B at 100 knowing his true odds are 50 )Giving us a 1% adavantage in probability but an advantage of 50 in odds. I would say Scenario 1 provides a greater edge but if you backed the same amount 100 times on both scenarios, Scenario 2 would be more profitable. Which has the greater edge?

[gloryhunters]

Re: Calculating/Comparing Edge Scenario 1 has the greater edge as if you bet 100 times on each you would win more through scenario 1. This is because it happens more frequently and so you have less losing bets to overcome.

[relf]

Re: Calculating/Comparing Edge Assuming if everything falls into place.. S1: You will win 99 out of 100 times. That's a 66% yield. (99 * 1.5 / 1) S2: You will win 2 out of 100 times. That's a 100% yield. (2 * 50 / 1) Yes, S2 produces a higher yield. But the occurrence of S2 is very low, and this means it is subjected to a high variance, i.e. there could be a chance where you can play 200 times without a win. Is your bank big enough to cope with that? Whereas for S1, a true odds of 1.01 (or near 99% chance) is highly unlikely to fail more than 3 consecutive times, which means you can work with a smaller bank or a higher staking strategy.

[gloryhunters]

Re: Calculating/Comparing Edge Whoops made an error in my previous post. I agree with the previous post about being able to bet bigger with more certainty. Another attraction with scenario 1 is that we very rarely know the true odds, so go through an estimation process. Even if our estimation is out by a few percentage points then we stil win in scenario 1 whereas in scenario 2 we could be losing (and we would never be able to find this out until a very long time).

[muppet77]

Re: Calculating/Comparing Edge Those yields are wrong bud. yield is profit over stake. Scenario 2, assuming 100 bets....... Staked = 100 Win = 2 x 50 Minus 2 stake = 98 Losses = 98 Profit = 0 Yield = 0 %

[muppet77]

Re: Calculating/Comparing Edge In fact true odds are 50 to one so you will win 1 in 51, so a little under 2 in 100....

[Arimathea]

Re: Calculating/Comparing Edge Thanks guys, glad that you agree with me and that scenario 1 is better... but these are the extremes... If Scenario 1 was real price 1.25 available at 1.67 (Difference between 80% and 60% probability i.e. 20%) and Sceario 2 was real price 2.00 available at 3.00 (Difference between 50% and 33% probability i.e. 17%) Would you still say scenario 1 had the greater edge and should be the bet?

[Lardonio]

Re: Calculating/Comparing Edge Math was my worst subject in school, but aren't you guys wrong and isn't this correct (assuming that everything goes exactly according to the odds) First post: Scenario 1 (back at 1.5, true odds 1.01) produces a profit of 48.5 units on 100 bets. Scenario 2(back at 100, true odds 50) produces a profit of 100 units on 100 bets Post above: Scenario 1 (back at 1.67, true odds 1.25) produces a profit of 33.6 units on 100 bets Scenario 2 (back at 3, true odds 2) produces a profit of 50 units on 100 bets

[relf]

Re: Calculating/Comparing Edge

Assuming if everything falls into place.. S1: You will win 99 out of 100 times. That's a 48.5% yield. (99 * 1.5 / 1) S2: You will win 2 out of 100 times. That's a 100% yield. (2 * 100 / 1) Yes, S2 produces a higher yield. But the occurrence of S2 is very low, and this means it is subjected to a high variance, i.e. there could be a chance where you can play 200 times without a win. Is your bank big enough to cope with that? Whereas for S1, a true odds of 1.01 (or near 99% chance) is highly unlikely to fail more than 3 consecutive times, which means you can work with a smaller bank or a higher staking strategy.

I have edited the errors in the numbers. Must have been sleeping early in the morning. Thanks to those who have pointed it out.

[relf]

Re: Calculating/Comparing Edge

Thanks guys, glad that you agree with me and that scenario 1 is better... but these are the extremes... If Scenario 1 was real price 1.25 available at 1.67 (Difference between 80% and 60% probability i.e. 20%) and Sceario 2 was real price 2.00 available at 3.00 (Difference between 50% and 33% probability i.e. 17%) Would you still say scenario 1 had the greater edge and should be the bet?

Lardonio has worked out the correct numbers. Scenario 2 has the higher 'expected' yield. But the variance or range will definitely be greater due to higher uncertainty. It all depends on individual preferences for these 2 scenarios. I would say they are both good. Perhaps a scenario I could pose you would be.. S1: The expected yield is 5%, but it has a +/- of 1%, meaning it could go from 4 to 6%. S2: The expected yield is 10%, but it has a +/- of 20%, meaning it could go from a loss of 10% to a profit of 30%. Which one would you take? The one with the higher 'expected' yield, or something that is sure to make a profit.

[Arimathea]

Re: Calculating/Comparing Edge Thanks again guys, I guess in football odds the highest odds are never going to exceed much more than 10-1 so the variance will not be so bad and knowing how random football matches can be it might prove more profitable to go with the higher yield in these circumstances. I guess it all depends on your picks and backtesting them to see what would have worked best to decide which to go with in future. In your example I would choose the first 'safer' option myself but certainly if I knew that the second scenario proved more profitable in the past I would go for that. The thing is sports betting will never go completely to the real odds or your odds so there isn't necessarily a better option but if your estimations and the real odds played out as expected I think that by using kelly mathematically, you will make more profit from betting on the selection with the best difference in percentage probability than you would betting on the selection with the best difference of odds. Am I right?

[Gingertipster]

Re: Calculating/Comparing Edge Not exactly what was asked but..... My own staking plan tries to take in to account both the percentage chance the selection has of winning. And the amount of value there is in the bet. e.g. If I believe something has a 25% (fair 3/1) chance of winning, and can get 4/1 (20%), I'll have 25 points (because my opinion of 25%) + 5 points (because the difference between my price and best price (25% and 20%) is 5%. So 25 + 5 = 30 points @ 4/1. This way I tend to bet more on the shor priced selections (because they've got a better chance of winning) yet win more on the outsiders (because they've got less money risked on the). Am thinking of upping it to 2 points per percentage point difference, to add more to the value aspect of the bet.

[Shy10ck]

Re: Calculating/Comparing Edge

This has provoked some good discussion but I still don't know the answer!! :( I thought I'd post on this forum where there are lots of experts! Please see below and thanks in advance!... Here's a confusing one (for me at least!) Do you compare edge in terms of probabibility difference or difference in the odds? E.g. Scenario 1: We back player A at 1.5 knowing that his true odds are 1.01 (Giving us a 32% advantage in probabililit but only a 0.49 advantage in odds) Scenario 2: We back player B at 100 knowing his true odds are 50 )Giving us a 1% adavantage in probability but an advantage of 50 in odds. I would say Scenario 1 provides a greater edge but if you backed the same amount 100 times on both scenarios, Scenario 2 would be more profitable. Which has the greater edge?

Scenerio 1 is by far the best choice for producing profit. Lets say you have a betting account of 1000 units. Scenerio 1 has a 99% success rate pr bet Scenerio 2 has a 2% success rate pr bet The stacking can be discussed, but a very roughly estimate could be: In scenerio 1 you can easily bet 100 units pr bet without fearing bankruptcy In scenerio 2 you can bet 2 units per bet and still fear bankruptcy After 100 bets in scenerio 1 you will have a profit of 4850 (1,5*100 *99)-(100*100) After 100 bets in scenerio 2 you will have a profit of 200 (100*2*2)-(100*2) So in my opinion there no doubt that Scenerio 1 is the better choice.