Is this a fair estimate of edge?

**hutennis** · 02-09-12, 11:40 AM

Originally posted by SportsMushroom

Lets say historically I win 53.3% of my nba wagers (for example), 51.2% of my soccer wagers, 54.9% of my tennis wagers....

The fact that you posted you winning %% without even mentioning corresponding implied probabilities, which is the only way win%% can have any kind of relevancy, makes your ability to draw correct conclusion from
your observation kind of a suspect.

Assuming the variables which I use to select a wager remain constant, could I possibly use those historical rates to calculate my edge in specific sports in order to apply the kelly criterion? or do you believe that historical data is non indicative of future performance?

Past performance is only indicative of future results in situations where CLT rules indisputably.
If you can not quantify your edge as presisly as casino does, then "past performance does not guaranty future results" is the rule to follow by default. Anyone claiming diffidently has to prove it to be taking seriously.

I know many of you calculate your own price and then compare it to the market price to determine your edge, but that seems too subjective to me, in determining which factors affect the price and by how much requires judgment and I dont believe that with my little laptop and excel spreadsheet I can outperform a harvard guy with a million dollars worth of software and hardware

I cant blame you for for being a not believer.

**SportsMushroom** · 02-09-12, 12:01 PM

Thank you for your reply

forgive me, I am relatively new to quantitative methods of handicapping and I am still trying to familiarize myself with it, I was just looking for an opinion as to whether past performance can be used as an estimate of the edge on future bets

Let me give you a figure from my spreadsheet, historically in soccer, I am winning 50.88% of my wagers at average odds of +115

please correct me if I have this wrong, but does this mean that implied probability of odds is 46.51%, and since the wagers are winning at 50.88% that impies a 1.09% edge?

this would be my historical edge, my question I guess is should I use this historical edge for future bets, or is it not relevant any more

appreciate any help you can give me

Originally posted by hutennis

The fact that you posted you winning %% without even mentioning corresponding implied probabilities, which is the only way win%% can have any kind of relevancy, makes your ability to draw correct conclusion from
your observation kind of a suspect.

**hutennis** · 02-09-12, 12:16 PM

I see.

Well, it is wrong.

First of all 50.88 has an edge of 9.61% over 46.51. It's calculated like this 50.88/46.51-1

But that's not all. Implied probability is not your dds converted to %%. Its no vig line converted to %%.
So, another side of your market is important. And since you use average it is important that average is calculated correctly.

**tukkk** · 02-09-12, 12:25 PM

9% edge is pretty bad
but dont worry, if you learn kelly and get your betsize correct, you will double that !

**SportsMushroom** · 02-09-12, 12:27 PM

Originally posted by hutennis

I see.

Well, it is wrong.

First of all 50.88 has an edge of 9.61% over 46.51. It's calculated like this 50.88/46.51-1

But that's not all. Implied probability is not your dds converted to %%. Its no vig line converted to %%.
So, another side of your market is important. And since you use average it is important that average is calculated correctly.

thank you for your help, if I could send you more points I would

a query I have regarding what you pointed out:

I know that the no-vig line is the true implied probability, BUT, no-vig lines are only in theory, no book will offer you a no-vig line. By using the no vig line to calculate edge, you essentially dont factor in the juice and you overstate your edge, which will result to non-efficient use of the kelly criterion, I could be mistaken, I fell asleep during math class in school, but that is why I am asking for help