Tennis
Welcome to Live View – Take the tour to learn more
Start Tour
There is currently 1 person viewing this thread.
As stated in the title. This has been bothering me for a long time. I'm sure they use a complicated algorithm, but purely for interest purposes, what are the basics?
For example, the game between Voracova v Date Krumm at 10:30 is listed as Voracova (Back: 2.92, Lay: 2.98) and Date Krumm (B:1.51, L:1.52).
Are these odds calculated based on ALL previous matches played by each player? Or their matches for the last 1 month/1 year/5 years..? Or their head to head (what if there's no head to head)?
And how do they turn historical stats into specific odds? and how do they calculate the changing odds throughout the match...e.g. if Voracova leads Date Krumm 15 - 0 in the first game, Voracova's odds will drop to a specific value..how is that calculated? 0.01 drop in odds per tennis point scored? Is the relationship with time linear, exponential, logarithmic?
I realise I've asked a million and one questions...but please, someone who knows, answer any one of them.
For example, the game between Voracova v Date Krumm at 10:30 is listed as Voracova (Back: 2.92, Lay: 2.98) and Date Krumm (B:1.51, L:1.52).
Are these odds calculated based on ALL previous matches played by each player? Or their matches for the last 1 month/1 year/5 years..? Or their head to head (what if there's no head to head)?
And how do they turn historical stats into specific odds? and how do they calculate the changing odds throughout the match...e.g. if Voracova leads Date Krumm 15 - 0 in the first game, Voracova's odds will drop to a specific value..how is that calculated? 0.01 drop in odds per tennis point scored? Is the relationship with time linear, exponential, logarithmic?
I realise I've asked a million and one questions...but please, someone who knows, answer any one of them.
Show More
BF does nothing but loading the markets, it's people like you and me offering the money at specific odds. Your questions are valid for any other bookies, but not bet exchanges.
Betfair offer markets unlike a traditional bookmaker. The people that set the odds are the people that place the bets - they populate the market with odds themselves. These are usually loosely based on the odds bookmakers offer.
here's a little snippet i came across. clearly it pertains to US markets, but the idea can be extended to any market .....
"Odds makers will tell you their job is NOT to predict an event's outcome but, rather divide the public as to who it thinks will win. "When the oddsmakers have set a point spread properly," said Howard Martin, a noted expert on how odds are calculated, "there will be an equal number of people betting each side of the line." "
"Odds makers will tell you their job is NOT to predict an event's outcome but, rather divide the public as to who it thinks will win. "When the oddsmakers have set a point spread properly," said Howard Martin, a noted expert on how odds are calculated, "there will be an equal number of people betting each side of the line." "
not an equal number of people, but roughly an equal profit for the bookie on each outcome from money placed at given odds
whatever, the main point being it's fckall to do with trying to predict the outcome. it's just mass psychology.
im not sure they actually have a system for calculating odds. They have professionals staff, probably using theyre experience and circumstances to define odds, which are then obviously subjective looking from a strictly mathematical point of view cause they are defined by a limited number of people.
(not betfair but most bookies that is not an exchange markets i meant)
the interesting question for me is, how much are early markets on BF influenced by bookies' odds? Clearly bookies react to shifts in markets on exchanges later, but how is it in the early phase?
i would say the other way round, odds compilers will often use bf as a stating point, before adding about ten per cent to their margins
Well for those sports I check regularly (NBA, tennis) Pinnacle publishes full odds lists even before Betfair puts up markets.
Bookmakers have been doing business LONG before Betfair got started, so thinking they have to wait for random people to put up stupidly juiced odds on an exchange is idiotic.
Bookmakers have been doing business LONG before Betfair got started, so thinking they have to wait for random people to put up stupidly juiced odds on an exchange is idiotic.
no not suggesting that, i just think that odds compilers dont always rely purely on their methods and instincts, that using any form of algorithm to create a market is little short of idiotic, and that, generally speaking the market is what the punters believe it is - ie if a compiler gets something wrong it wont last long. so i believe odds compilers prob use bf or pinny as a guide....another tool for reassurance
Also, here is some more help: http://lmgtfy.com/?q=how+bookies+calculate+odds 
Few comments:
1. "real" oddsmakers use Elo rating system for basis, modified for injuries and some other factors
2. odds are historically accurate so they ARE predictions of who will win matches
3. most of the money is always put on favorites, and bookies never have "all green" books - they don't have to
Few comments:
1. "real" oddsmakers use Elo rating system for basis, modified for injuries and some other factors
2. odds are historically accurate so they ARE predictions of who will win matches
3. most of the money is always put on favorites, and bookies never have "all green" books - they don't have to
i'm fascinated by how compilers create markets for tennis. are they all tennis experts or are they simply being scientific. you talk about a rating system being modified marin. presumably those modifications are key, as a ratings system wouldnt take into account things as obvious as opponent, current form, weather, surface.....or would it...please enlighten me
Pinnacle work differently: they set opening lines with low limits, gradually increase them, and let the money dictate where it ends up. It's essentially an exchange model.
If you analyse those opening lines over time, they're pretty solid as a predictor. For example, atp 1.66-2.00, wins 2845/5214, returns -2% (which equates to -104 or 1.96 odds). They have less juice on lower odds, more on higher, but overall they're pretty damn good.
Of course, it's just one book, but it's also the biggest in tennis betting by miles.
If you analyse those opening lines over time, they're pretty solid as a predictor. For example, atp 1.66-2.00, wins 2845/5214, returns -2% (which equates to -104 or 1.96 odds). They have less juice on lower odds, more on higher, but overall they're pretty damn good.
Of course, it's just one book, but it's also the biggest in tennis betting by miles.
The line about bookies setting prices to balance the action is wide of the mark. It's just a rent-a-quote trotted out largely by Americans.
99% of all odds are calculated and "published" by betradar, but you can't see them because you don't have a contract with them. Then bookies do their own math: put odds up and down slightly first for their profit margin, then for some eventual info if they find it and in the end they have to see how exposed are they regarding their risk limits.
Maybe some bookies do their own for some small part of not even sports, but competitions inside these sports, but in particular there is to much work any you should have army of employees for something like that. Second thing is that top markets (tennis GS, american sports, top football leagues etc) are quite self efficient and odds include more or less all known facts. Something like Markowitz theory in world of finance.
Somebody mentioned complicated algorithms and that is also false. Try with google a little bit and you will find that creating odds is quite simple, and even if you get something wrong you always have ton of others bookies to compare. They also use betfair, betbrain, oddschecker and similar tools and that's pure logic, as odds movements just represents where is the money and "why" there is a money..
Maybe some bookies do their own for some small part of not even sports, but competitions inside these sports, but in particular there is to much work any you should have army of employees for something like that. Second thing is that top markets (tennis GS, american sports, top football leagues etc) are quite self efficient and odds include more or less all known facts. Something like Markowitz theory in world of finance.
Somebody mentioned complicated algorithms and that is also false. Try with google a little bit and you will find that creating odds is quite simple, and even if you get something wrong you always have ton of others bookies to compare. They also use betfair, betbrain, oddschecker and similar tools and that's pure logic, as odds movements just represents where is the money and "why" there is a money..
Betfair's tennis odds are done by a very well-known ex-player.
For the big games, he adjusts odds live after each point.
For other games, it is a bot working it out based on past matches.
For the big games, he adjusts odds live after each point.
For other games, it is a bot working it out based on past matches.
Also, if you are trying to calculate odds and see if there is any value you should put your own expectations based on possibilities you give to every possible outcome to happen but.... Here it comes the big BUT..
For any given possibility you will get some odd, and then you say "Yeah, I find value". Nice, but what if you put wrong possibilities (and automatically) and get the wrong odds based on that. For thousand of times I find people talking there is no value in 1,10, 1,15 etc and in long term you will get bankrupt. And if you put value bets over the long run you will get rich. The same thing as not every odd of 1,80 has real value of 1,80 the same thing is with smaller odds. Nadal in first round of GS have all the value of the world in 1,01 because he will get it every time. And that's the fact.
Answer for this is one of strongest cognitive biases called overconfidence. All punters think they can find value odds, that they can count real possibilities for outcomes, etc etc.. But in the end, when you draw the line, NET profit is all that meters.
For any given possibility you will get some odd, and then you say "Yeah, I find value". Nice, but what if you put wrong possibilities (and automatically) and get the wrong odds based on that. For thousand of times I find people talking there is no value in 1,10, 1,15 etc and in long term you will get bankrupt. And if you put value bets over the long run you will get rich. The same thing as not every odd of 1,80 has real value of 1,80 the same thing is with smaller odds. Nadal in first round of GS have all the value of the world in 1,01 because he will get it every time. And that's the fact.
Answer for this is one of strongest cognitive biases called overconfidence. All punters think they can find value odds, that they can count real possibilities for outcomes, etc etc.. But in the end, when you draw the line, NET profit is all that meters.
Nadal in first round of GS have all the value of the world in 1,01 because he will get it every time. And that's the fact.
Bumping old post here, but last night at Wimbledon he was no 1.01 shot.
Got the job done in the end, but he was $1.40 to $1.50 odds in reality, prior to match turning in play.
Bumping old post here, but last night at Wimbledon he was no 1.01 shot.
Got the job done in the end, but he was $1.40 to $1.50 odds in reality, prior to match turning in play.
Post Your Reply
Please login to post a reply.
Wonder
Instance ID: 13539