Arbusers

hello
My goal is to code a bot that will automate value betting. The coding itself shoudn't be the problem.
As far as I understood when valuebetting on asian markets different basebooks are neccessary so that I have a basebook for football for example.
My problem is now that I don't think that I have understood how to determin the most efficient bookmaker in a sport. I would try and use historical odds from bookmakers to look at their efficiency.
Is anyone nice enough to explain how I can see how efficient odds of a bookie are?
thank you 

Use Betfair (exchange) as the line, its about as accurate as possible.

dealer wins wrote: Use Betfair (exchange) as the line, its about as accurate as possible.

if I understood that right that would mean that all odds above the betfair line are value bets?
I thought that there are specific bookies for every sport

dealer wins wrote: Use Betfair (exchange) as the line, its about as accurate as possible.

Betunfair is only sharp where they use bots to follow pinnacle and the asian books, such as football match odds. For anything outside the main market they are weak due to low liquidity and premium charge/high commission stopping any chance of finding efficient odds.

Look at the state of 21bet exchange, a betfair white label. They took some of the risk and ended up going bust and still haven't paid out some winning customers after over a year.

Smarkets copy betfair exchange prices and their profits went down 50%. https://egr.global/intel/news/smarkets- ... g-decline/

Smarkets are a strange one. Instead of addressing the issue of weak odds they have introduced commission tiers and unfair t's and c's to stop winning traders on their exchange (they put you on 1% commission on wins and losses).

I agree that betfair should not be used as a base book for the above mentioned reasons.
This question has no easy answer. I m observing a number of value betting projects both manual and automated. I would say that your filtering should be multidimensional.
Because it is not only a matter of bookmakers, but a matter of odds as well.
For example pinnacle might be a good reference for odds 2-3 while bet365 might be a good option for 3+ odds.
Just to give you an idea, I would focus on pinnacle, SBOBET and bet365.

arbusers wrote: I agree that betfair should not be used as a base book for the above mentioned reasons.
This question has no easy answer. I m observing a number of value betting projects both manual and automated. I would say that your filtering should be multidimensional.
Because it is not only a matter of bookmakers, but a matter of odds as well.
For example pinnacle might be a good reference for odds 2-3 while bet365 might be a good option for 3+ odds.
Just to give you an idea, I would focus on pinnacle, SBOBET and bet365.

thank you for your answer. I will look into that when analysing the data.
Is there maybe a good known article where calcuting if odds are efficient is described? I googled a bit but didn't find such. I have still problems to imagine how to calculate that mathematically.

I would avoid an academic approach to this matter. I m quoting Charlie Munger:
“I can’t give you a formulaic approach, because I don’t use one. And I just mix all the factors, and if the gap between value and price is not attractive, I go on to something else. And sometimes it’s just quantitative. If you want a formula, you should go back to graduate school. They’ll give you lots of formulas that won’t work.”

At the same time, I would like to give you a very general guideline. So you basically need to find the bookmakers and the odds range where these bookmakers offer value. Notice that every bookmaker might offer value at a different odds range and markets. Then you need to cover their lines, plus their margin, no matter what this is.
For example, if you find the markets and odds range in Pinnacle, you then need to cover their fee, 1-1.5%+ to be on the right side of the value.

You might want to take a look at the following video in our youtube channel: