Now finished reading Nate Silver's (http://en.wikipedia.org/wiki/Nate_Silver) book, "The Signal and the Noise, Why so many predictions fail but some don't".
The book is well worth the read IMHO and there are a few chapters and observations that should be particularly educating and interesting those making money and those attempting to make money from betfair. Even though I studied applied maths at Uni (but never used it in my day job) I don't think you need any advanced maths education to read or enjoy this book. If you are making money on betfair, you've no doubt got all the maths knowledge you need.
To whet the appetite, midway through the book there are a few chapters devoted to sports betting and poker. In fact, Nate writes that he was making up to US$400,000 per year playing poker online. But, when the market changed and there were fewer fish (suckers/newbies) losing money he turned from an on average winner per hundred hands, to an on average loser per hundred hands. Nate explains the lessons he learned from watching the online poker market change. (He realised he had lost his edge and moved on.)
The book focuses on Bayes' theorem. http://en.wikipedia.org/wiki/Bayes'_theorem
Through various examples, be it baseball, sports betting, political forecasting, poker, the stock market, terrorism, climate change - the author uses Bayes' theorem to calculate probability and analyse prediction problems.
I did find myself skim reading a few of the later chapters and I concentrated most of my attention on those chapters I considered having direct benefit to betfair strategies.
The book will appeal to those people who like to make predictions by building models. Those who observe patterns in the market and then exploit or attempt to exploit those patterns for profit. Those that tinker, those that like to think about problems in their study. Those trying to separate the signal from the noise.
PS Nate would have made some serious money in the US election given his blogs predictive rate of success in just about every political contest in the election. http://fivethirtyeight.blogs.nytimes.com/
PPS Nate is also a big fan of http://www.intrade.com/v4/home/
The Signal and the Noise (completed book review)
Hi G
Thanks for taking the time to post the review
I'll look it up on Amazon
Regards
Peter
PS Good luck with the trading. I know there are many ways of profiting on betfair ie Value bet horses & using opinion etc..But with a background in maths you will find angles down that route too for sure.
Thanks for taking the time to post the review
I'll look it up on Amazon
Regards
Peter
PS Good luck with the trading. I know there are many ways of profiting on betfair ie Value bet horses & using opinion etc..But with a background in maths you will find angles down that route too for sure.
Thanks Peter
I got the book on the kindle for iPad (and had to use a US address).
I'm able to make enough on betfair each Saturday making books on the horses to satisfy the ego. I use it as stress relief from my day job.
But from reading the book I will be focusing more attention on sports stats and politics. There are a few patterns I can see that may be worth exploring further. Interestingly, the book relates that pattern recognition is a key skill humans have over computers. This is explained through the chess matches between Kasparov http://en.wikipedia.org/wiki/Garry_Kasparov and Deep Blue.
Gazuty
I got the book on the kindle for iPad (and had to use a US address).
I'm able to make enough on betfair each Saturday making books on the horses to satisfy the ego. I use it as stress relief from my day job.
But from reading the book I will be focusing more attention on sports stats and politics. There are a few patterns I can see that may be worth exploring further. Interestingly, the book relates that pattern recognition is a key skill humans have over computers. This is explained through the chess matches between Kasparov http://en.wikipedia.org/wiki/Garry_Kasparov and Deep Blue.
Gazuty
Asked for this book as a Christmas present and really enjoying reading it. Ive only partially read it, but it has given me some ideas to explore for 2014, partially the Bayes (which Gazuty mentions), is explained in simple way and leaves you wanting to find out more.
He wrote about Bob Voulgaris (who is the US equivalent of Peter Webb). Peter he is based in LA so maybe you could meet up with him next time you're out there?
http://espn.go.com/blog/playbook/dollar ... ba-gambler
He wrote about Bob Voulgaris (who is the US equivalent of Peter Webb). Peter he is based in LA so maybe you could meet up with him next time you're out there?
http://espn.go.com/blog/playbook/dollar ... ba-gambler
Peter you did well to digest it so quickly. I had to re-read it several times for it to sink in. The chapter on Bob Voulgaris is very interesting. I went back and had a look at the chapter and this sticks out:
"Voulgaris is among the best sports bettors in the world today - he still only gets about 57% of his bets right. Its exceptionally difficult to do better than that."
It's starting to dawn on me that in trading and betting the line between success and failure is very thin (most worryingly you may often have a winning strategy and never actually know) and your return against your investment is always likely to be very low.
"Voulgaris is among the best sports bettors in the world today - he still only gets about 57% of his bets right. Its exceptionally difficult to do better than that."
It's starting to dawn on me that in trading and betting the line between success and failure is very thin (most worryingly you may often have a winning strategy and never actually know) and your return against your investment is always likely to be very low.
Hi Steve
still working throught it at the mo.
57% on its own doesn't really tell you much. For instance if he is winning twice as much money as he loses, then even 50% would be remarkable
The bayes has spurned me on to start collecting my own data as that seems to be the route for long term profitability. I was up till 2:30 this morning building my first ever database! I recon it will take me a good ten days to pull in the data and start playing around with it.
I'm going to start looking for patterns in the pre off and in play horse markets (uk only for now) looking at all the criteria, horses, trainers, jockeys, etc based on a hypothesis that there is corruption and manipulation...not sure where it will take me but if I,don't look I'll never know!
I can do this in parallel with my normal trading so it's just time and effort
regards
Peter
still working throught it at the mo.
57% on its own doesn't really tell you much. For instance if he is winning twice as much money as he loses, then even 50% would be remarkable
The bayes has spurned me on to start collecting my own data as that seems to be the route for long term profitability. I was up till 2:30 this morning building my first ever database! I recon it will take me a good ten days to pull in the data and start playing around with it.
I'm going to start looking for patterns in the pre off and in play horse markets (uk only for now) looking at all the criteria, horses, trainers, jockeys, etc based on a hypothesis that there is corruption and manipulation...not sure where it will take me but if I,don't look I'll never know!
I can do this in parallel with my normal trading so it's just time and effort
regards
Peter
-
convoysur-2
- Posts: 1110
- Joined: Thu Jan 12, 2012 10:00 am
im reading it at the moment ,
its a very interesting read,
Marc
its a very interesting read,
Marc
Yes, it would have nice to have known the returns, but I get the impression that he works to a small margin of error. When talking about the 53 - 56% varience, Silver mentions later that "this is the narrow margin that pro gamblers work to".57% on its own doesn't really tell you much.
I would love to know what margins professional gamblers work to and I have developed the feeling in recent years that it's much less than we imagine. It's probably far from the Hollywood depiction.
By trade I am a web developer, so I am fortunate to have a decent knowledge of database development. I am 3/4 through building one for pre-racing which has taken me ages to compile.
I am doing some football stuff in excel, which has been useful. However I can't help think that if you want to use a lot of data then databases is the way to go.