Gooner80 Posted March 21, 2011 Share Posted March 21, 2011 Hi there forgive me if anyone else has mentioned this before, but the guy who runs football data and who wrote that magnificent book mentioned that truer odds can be obtained from markets where there is only two outcomes, so for 1x2 betting I decided to get the asian handicap 0 ball odds in this case for mK dons versus peterboro so let me break it down 1x2 odds ( best on odds checker are) 1 2.62 x 3.6 2 2.7 asian handicap (0 ball) home 1.91 away 1.98 then I took arb crunchers draw no bet calculator and cranked in the over round which is 2.98% with the best odds I put this in as the commission. so using the best 1x2 odds and the draw in this calculator I played with the odds until they matched the asian handicap odds these are the odds I came out with home 2.69 (37.17%) draw 3.72 (26.88%) away 2.78 (35.97) may be I am totally wrong but does anyone have any ideas/ suggestions of how I can extract the true odds from asian handicap, I understand it is guess work but there must be some link between the two markets? any help is appreciated Quote Link to comment Share on other sites More sharing options...
Datapunter Posted March 21, 2011 Share Posted March 21, 2011 Re: Extracting true odds need a bit of help with theory You want to deduct the odds for the 1X2 market on the basis of the odds for the Asian handicap market ? Quote Link to comment Share on other sites More sharing options...
Donkson Posted March 22, 2011 Share Posted March 22, 2011 Re: Extracting true odds need a bit of help with theory Yeah that's what he's asking. Interesting idea, but I'm not certain how you're quantifying the draw odds from the asian handicap. Quote Link to comment Share on other sites More sharing options...
Gooner80 Posted March 24, 2011 Author Share Posted March 24, 2011 Re: Extracting true odds need a bit of help with theory Yes I have been looking at the asian handicap and trying to extract the1x2 odds, based on that figure. I try and take a educated guess on the draw odds, based on my own research and the expected goals. i feel the over round is harder to disguise in two outcome markets and my hunch is something like a 45% to 55% split of the over round on fairly even teams, from what I can see because the price is higher on the away team and also the liability and more people likely to accept it because it is a higher price they add more onto the away team. from what I have been doing the hardest part is working out the draw odds. I work on the basis that the more likely the game is going to be under 2.5 goals the more likely it is going to be a draw and then I add on to probability of 0-0 and 1-1 scores the 2-2 and 3-3 scores and come up with my draw figure. once I know my draw figure I can work out the overround on the asian handicap deduct it according to the split which ATM is just a hunch and work out my home and away scores I was hoping anyone might have some suggestions that might be able to refine what I am doing. Quote Link to comment Share on other sites More sharing options...
froment Posted March 28, 2011 Share Posted March 28, 2011 Re: Extracting true odds need a bit of help with theory This was interesting question, and I hoped someone to answer, while trying to figure it out myself; however, it seems no "mathematical" answer has been available, so I decided to look for an "empirical" solution: connection between 1X2 and AH(0) odds offered by bookmakers. In particular, local bookie here, whose odds I happened to have when I decided to give this a go (and which appeared to have terrible overround, but all local bookies here do have, I believe). Here is simple thing that I did: converted 1X2 and AH(0) into percents, and looked for a ratio home win / home(0) and away win / away (0). This ratio appeared to be pretty consistent, especially in narrow range of odds. Odds quoted by Gooner80 are all without overround; so, I assumed that after overround, home AH(0) odds of 1.91 will drop to around 1.70 to 1.80; I took several matches with those odds, and did the above calculation; for that range of odds, ratio for both home win/home(0) and away win/away(0) is 1.43, with standard deviation of only about 0.011 and 0.015. I extended my research to another range of home(0), 1.25 to 1.35, and just out of interest, added several more matches from other ranges. Results are somewhat similar; in order not to interprete them here, look at the attached Excel if you think this might be useful. That Excel contains another set of data, odds for under/over 2.5, as I thought it would impact AH(0) odds (because in matches where lower number of goals is expected, ratio of correct scores probabilities ((0-0)+(1-1)+(2-2)+...) / ((1-0)+(2-0)+(3-0)+(3-1)+...+(0-1)+(0-2)+(0-3)+(1-3)...) is higher, thus making draw odds lower for the same ratio home win/away win, as you noticed yourself), but it seems I was wrong - I don't see major influence of expected number of goals to this ratio. To conclude: once I established that ratio, next steps are these: Home(0) odds 1.91 = 52.4% Home win probability: 52.4/1.431 = 36.6% Home win odds = 2.73 Away(0) odds 1.98 = 50.5% Away win probability: 50.5/1.429 = 35.3% Away win odds = 2.83 Draw probability = 100 - 36.6 - 35.3 = 28.1% Draw odds = 3.56 You see, home win and away win odds are similar to ones you got yourself, but draw odds differ significantly... I believe main reason is high overround in odds I used, almost fourfold in comparison to 3% you used for calculation, but once I started with these odds, I couldn't bother to re-start again with odds from another bookie; idea can be also basically flawed, as AH(0) cannot be really converted into percents in the same way as 1X2, since there is "void" option, but if you have no other idea for this "extraction", you might try to explore this ratio with odds from your bookie - there must be a reason why this ratio is consistent. Here's Excel (right click and "Save As'; Excel 2003, 70 kB): excel.gubbed.me/files/Extracting_1X2_from_AH.xls Quote Link to comment Share on other sites More sharing options...
Gooner80 Posted April 5, 2011 Author Share Posted April 5, 2011 Re: Extracting true odds need a bit of help with theory Thanks for you help froment :clap Quote Link to comment Share on other sites More sharing options...
chikilicuatre Posted January 1, 2012 Share Posted January 1, 2012 Re: Extracting true odds need a bit of help with theory excel.gubbed.me/files/Extracting_1X2_from_AH.xls Happy New Year. The file is not available, can re-upload? Thanks¡¡ Quote Link to comment Share on other sites More sharing options...
froment Posted January 2, 2012 Share Posted January 2, 2012 Re: Extracting true odds need a bit of help with theory Happy New Year. Thanks! Happy New Year you, too! :ok The file is not available, can re-upload? Thanks¡¡ I re-arranged in the meantime structure of the site, so I forgot to move some files in proper destination... :spank Anyway, thanks for notifying me, and it's here now: www.gubbed.me/fair/?p=895 Quote Link to comment Share on other sites More sharing options...
chikilicuatre Posted January 2, 2012 Share Posted January 2, 2012 Re: Extracting true odds need a bit of help with theory Good morning, Thank you too for the work. And apologize for my bad English, I need more practice. On the other hand, can you explain what you have to fill and how? I have some doubts about the operation, and what to fill specific Thanks Quote Link to comment Share on other sites More sharing options...
froment Posted January 2, 2012 Share Posted January 2, 2012 Re: Extracting true odds need a bit of help with theory You welcome, and don't worry too much about your English - it is understandable, and we all need to practice it... :ok Purpose of this spreadsheet was to try to find correlation between 1X2 and AH oods, and I did it for specific case above, using odds from my local bookmaker. Original poster asked if there was "mathematical" way to find that relationship. No-one replied, so I decided to take odds from my local bookmaker and compare odds. It appeared that ratio between probabilities of Home Win and Home (0) was pretty consistent, 1.41 to 1.43; ratio of probabilities of Away win and Away (0) was not that consistent, but anyway was in similar range, 1.43 to 1.52. That "coefficient" enabled me to calculate 1X2 odds from AH(0) odds, as displayed in smaller table "Conversion". So, if you want to check if your bookmaker maintains the same ratio between 1X2 and AH(0) oods, you'd need to replace all odds in blue-coloured cells with odds from your bookmaker (you can insert more rows if you need); after that, you can use table "Conversion" to calculate 1X2 odds if you know only AH(0) odds. Though, I'm not sure if this table has some practical value - bookmaker will be always glad to provide you with all odds you want; this was rather "theoretical" question and challenge how to link these two types of odds. Quote Link to comment Share on other sites More sharing options...
chikilicuatre Posted January 2, 2012 Share Posted January 2, 2012 Re: Extracting true odds need a bit of help with theory Good afternoon, Thank you very much for the prompt response. I am conducting a similar study to see where a bookmaker can make mistakes, or when he is right. After all variables would be an event that is to be treated separately. Regards and Happy New Year PD: And this http://www.gubbed.me/fair//betting/Poisson_Number_of_goals.xls ?? Quote Link to comment Share on other sites More sharing options...
chifoobear Posted July 23, 2013 Share Posted July 23, 2013 Re: Extracting true odds need a bit of help with theory Thank you for the in depth explanation Froment! However, when I click on: www.gubbed.me/fair/?p=895 I was unable to find the Extracting_1X2_from_AH.xls file. Can you re-post it again please? I am currently doing research on this topic and I find it fascinating! Quote Link to comment Share on other sites More sharing options...
chifoobear Posted July 23, 2013 Share Posted July 23, 2013 Re: Extracting true odds need a bit of help with theory More specifically, how would one convert 1x2 fixed odds to other, more complicated, Asian Handicap Odds? e.g. +/-0.25, +/- 1 Quote Link to comment Share on other sites More sharing options...
dalkent Posted November 27, 2013 Share Posted November 27, 2013 Re: Extracting true odds need a bit of help with theory Been playing around with this myself today. Here is what I determined Let H,D,A = the decimal odds of 1,X,2 respectively. Then implied Asian +0 odds are as follows H(+0) = H(D-1)/D H(-0)=A(D-1)/D Similarly H(+0.5)= DH/(D+H) A(-0.5) =A H(-0.5)= H A(+0.5) =DA/(D+A) So in the example using 2.62, 3.6 and 2.7 we get implied Asian odds as H(+0)=2.62*(3.6-1)/3.6 = 1.892 H(-0) =2.7*(3.6-1)/3.6= 1.95 H(+0.5)= 2.62*3.6/(2.6+3.6)=1.516 A(-0.5) =2.7 H(-0.5)= 2.62 A(+0.5) =3.6*2.7/(3.6+2.7)=1.543 To go back is a bit harder as you can’t just do it from H(+0) and A(-0). Bu obviously if you have A(-0.5) and H(-0.5) then you can derive D easily. I tested it looking at Pinnacle Odds for Spurs – Man United this week. I implied the Asian odds from the 1x2 oddsodds of 2.62, 3.44 and 2.87 H(0) 1.87 implied 1.86 A(0) 2.07 implied 2.04 H+.5 1.47 implied 1.49 A-.5 2.87 implied 2.87 H-.5 2.61 implied 2.62 A+.5 1.56 implied 1.56 Quote Link to comment Share on other sites More sharing options...
froment Posted November 28, 2013 Share Posted November 28, 2013 Re: Extracting true odds need a bit of help with theory Thank you for the in depth explanation Froment! However' date=' when I click on: www.gubbed.me/fair/?p=895 I was unable to find the Extracting_1X2_from_AH.xls file. Can you re-post it again please? I am currently doing research on this topic and I find it fascinating! Somehow I missed this post... :\ Yes, I changed the site structure once again, and file can be found now at www.betgps.com/excel_files/betting/Extracting_1X2_from_AH.xls , but I guess you have given it up by now... Been playing around with this myself today. Here is what I determined Let H,D,A = the decimal odds of 1,X,2 respectively. Then implied Asian +0 odds are as follows There is an Excel AH(0) calculator available at www.betgps.com/excel_files/betting/Double_Chance_and_DNB_Odds_Calc.xls , it yields the same calculation as you did. I'd recommend googling out "Teppo Salonen", he has a nice paper how to calculate other AH odds. Quote Link to comment Share on other sites More sharing options...
dalkent Posted November 28, 2013 Share Posted November 28, 2013 Re: Extracting true odds need a bit of help with theory Great at least i can trust my calculations. Will have a look for the paper Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.