Orpheus Posted November 16, 2008 Share Posted November 16, 2008 The existence of a favourite-longshot bias is well-documented - mostly positive (i.e. there is better value in backing short-priced horses than outsiders) and occasionally negative (i.e. there is better value in backing underdogs in American team sports). Naturally, all these studies look at the odds/results after the completion of the event. I want to see if the assumed existence of a positive favourite-longshot bias in football can provide any usefulness in forecasting. Has anyone else used this within their betting systems? The odds/results have been taken from football-data.co.uk and cover the 22 different leagues available for download on that site. The odds used are the BetBrain maximum odds. I've split the odds into intervals of 0.05 up to a maximum of 2.25 (else it is not a positive favourite-longshot bias system) and so I've tracked the profit/loss from blindly backing teams in each odds interval. Here is an example of the results: In the English Premier League, blindly backing every home team up to odds of 2.05 produced a profit of 15.17 units from 1 unit stakes in 2007; in 2006 backing every home team up to odds of 2.10 produced a profit of 18.23 units from 1 unit stakes. However, the size of these profit-making regions of odds is rare - most leagues have only small regions of odds in which consistent profits have been made by blindly backing home/away teams. Indeed, that has been the case with the Premier League in 2008 unlike the last two years. So, my system is to take the last three years of odds/results and calculate which odds intervals have shown a cumulative profit by backing teams whose highest odds fall into these intervals. These 'profitable odds' intervals then have two further filters: (i) profits in each interval must have been positive in two of the last three years (to eliminate the effect of one very good year against two small negative ones); (ii) cumulative profits must have been positive in each one neighbouring odds interval (so if 1.86-1.90 has been profitable in the Premier League in total over the last three years and in at least two of the last three years, it is designated a 'profitable odds' region only if the same holds for either the 1.81-1.85 or 1.91-1.95 regions). Teams are then blindly backed whose highest odds fall into these 'profitable' odds intervals in each league. Results have been encouraging this season, but I'm posting it here to see if I can learn from anyone who has looked at a similar 'system'. Teams backed today from this system: Werder Bremen 1.50 Catania 2.19 Udinese 1.45 Alkmaar AZ 1.81 Celtic 1.29 Athletico Madrid 1.67 Espanyol 1.73 Eskisehirspor 1.91 Let see if the poster's curse kicks in :tongue2 Quote Link to comment Share on other sites More sharing options...
Orpheus Posted November 18, 2008 Author Share Posted November 18, 2008 Re: Football Favourite-Longshot Bias System Bets: 8 Wins: 4 Win Percentage: 50% Total Staked: 8pts Profit: -1.35pts Yield: -16.9% Not the best start :eyes Just one qualifying bet from the midweek's slim pickings: Cambridge Utd 1.96 [Pinnacle] Quote Link to comment Share on other sites More sharing options...
Orpheus Posted November 20, 2008 Author Share Posted November 20, 2008 Re: Football Favourite-Longshot Bias System Bets: 9 Wins: 5 Win Percentage: 56% Total Staked: 9pts Profit: -0.39pts Yield: -4.3% One qualifying bet for tonight: Crawley Town 1.94 [Pinnacle] Quote Link to comment Share on other sites More sharing options...
Orpheus Posted November 22, 2008 Author Share Posted November 22, 2008 Re: Football Favourite-Longshot Bias System Bets: 10 Wins: 5 Win Percentage: 50% Total Staked: 10pts Profit: -1.39pts Yield: -13.9% Today's qualifying bets: Portsmouth 1.92 [Pinnacle] Man Utd 1.88 [Pinnacle] Preston 1.91 [Pinnacle] Ipswich 2.1 [Paddy Power] Carlisle 1.80 [Paddy Power] Burton 2.00 [Paddy Power] Schalke 04 1.44 [stan James] Inter Milan 2.15 [Pinnacle] Albinoleffe 2.10 [betFred] Grosseto 2.12 [Pinnacle] PSV 1.25 [Paddy Power] Dundee Utd 1.57 [Pinnacle] Motherwell 2.24 [Pinnacle] Celtic 1.39 [Pinnacle] Real Madrid 1.34 [Pinnacle] Hercules Alicante 2.20 [stan James] Antalyaspor 2.05 [Centrebet] Gaziantepspor 2.00 [Paddy Power] Quote Link to comment Share on other sites More sharing options...
Orpheus Posted November 23, 2008 Author Share Posted November 23, 2008 Re: Football Favourite-Longshot Bias System Bets: 28 Wins: 17 Win Percentage: 61% Total Staked: 28pts Profit: +2.21pts Yield: +7.9% Today's qualifying bets: Tottenham 1.75 [unibet] Lazio 2.03 [Pinnacle] Barcelona 1.22 [Pinnacle] Buyuksehir 2.00 [bet365] Trabzonspor 2.05 [Pinnacle] Quote Link to comment Share on other sites More sharing options...
Orpheus Posted December 5, 2008 Author Share Posted December 5, 2008 Re: Football Favourite-Longshot Bias System Today's qualifying bets: Sedan 2.00 [bet365] Chateauroux 1.93 [Pinnacle] Montpellier 1.92 [Pinnacle] Brest 1.95 [Fonbet] Boulogne 1.90 [VCbet] Metz 1.83 [betFred] Lens 2.05 [Pinnacle] Ingolstadt 1.80 [Pinnacle] Mainz 1.73 [Paddy Power] Quote Link to comment Share on other sites More sharing options...
Orpheus Posted December 5, 2008 Author Share Posted December 5, 2008 Re: Football Favourite-Longshot Bias System Bets: 33 Wins: 19 Win Percentage: 58% Total Staked: 33pts Profit: +0.96pts Yield: +2.9% Quote Link to comment Share on other sites More sharing options...
Orpheus Posted December 6, 2008 Author Share Posted December 6, 2008 Re: Football Favourite-Longshot Bias System Bets: 42 Wins: 20 Win Percentage: 48% Total Staked: 42pts Profit: -5.99pts Yield: -14.3% If I did believe in the posting curse ... :wall Today's qualifying bets: Liverpool 1.81 [betDaq] Newcastle 1.83 [blueSq] Cardiff 2.10 [blueSq] Sheffield Utd 2.07 [betDaq] Leicester 1.68 [Fonbet] Colchester 1.73 [betFred] Hartlepool 1.80 [VCBet] Millwall 1.90 [betDaq] Oldham 1.82 [betfair] Ebbsfleet 1.91 [stan James] Burton 2.20 [Expekt] Torquay 1.91 [stan James] Wrexham 1.95 [blueSq] Lyon 1.85 [Expekt] Modena 2.12 [betDaq] AZ Alkmaar 1.23 [betfair] Stenhousemuir 1.85 [unibet] Villareal 1.78 [betfair] Elche 2.20 [Eurobet] Antalyaspor 2.10 [stan James] Besiktas 1.53 [stan James] Quote Link to comment Share on other sites More sharing options...
Orpheus Posted December 7, 2008 Author Share Posted December 7, 2008 Re: Football Favourite-Longshot Bias System Bets: 63 Wins: 30 Win Percentage: 48% Total Staked: 63pts Profit: -7.89ptsYield: -12.5% System is throwing up too many plays at the weekend ... time for first revision: the full sample period from football-data.co.uk is used (00/01 season onwards instead of just last three years; other filters still apply). Today's qualifying bets under the revised system: St Pauli 1.70 [bet365] Celtic 1.65 [Expekt] Mallorca 1.91 [betFred] Quote Link to comment Share on other sites More sharing options...
PAULM03 Posted December 7, 2008 Share Posted December 7, 2008 Re: Football Favourite-Longshot Bias System Interesting approach. Have you read the stuff by the guy who owns football-data? According to that (and I think most agree with him) the favourite-longshot bias only exists in football at ultra-short odds (less than about 1.25 I think). Look for datapunters hot favourites system. Quote Link to comment Share on other sites More sharing options...
Orpheus Posted December 7, 2008 Author Share Posted December 7, 2008 Re: Football Favourite-Longshot Bias System Thanks for your input PAULM03 :D Odds of 1.25 or lower certainly is the most common range across the different leagues for positive returns and, as should be expected, this is most common in the leagues at the top of each national pyramid - particularly Spain, Italy, Scotland, Netherlands and Belgium. Interestingly, though, this has never been the case in the Premier League. Backing teams with odds of 1.25 or lower in this league would have produced a profit in only three of the last nine seasons. As such, and having now had a quick look at Datapunter's 'hot favourites' thread, any favourite-longshot bias that does exist only exists under conditions specific to each league and therefore at odds ranges specific to each league. I wouldn't apply this to Cup matches in any situation. Whether or not testing that a favourite-longshot bias exists in historical odds/results is largely an academic exercise. I've found that it can be profitable (despite the record since posting) for forecasting, which is a different statistical approach. While I found that predicted profits up to this season (when I started to put money on this system) were largest when using a rolling sample of last three seasons' of odds/results, it was producing a lot of plays at the weekends ... great when the matches go as planned, but can lead to some uncomfortable losses when results don't go well. Changes the sample period reduce expected profits from the forecasting analysis, but it should limit the daily downsides when they occur. Quote Link to comment Share on other sites More sharing options...
Orpheus Posted December 8, 2008 Author Share Posted December 8, 2008 Re: Football Favourite-Longshot Bias System Today's qualifying bet: Nurnberg 1.67 [stan James] Quote Link to comment Share on other sites More sharing options...
Orpheus Posted December 9, 2008 Author Share Posted December 9, 2008 Re: Football Favourite-Longshot Bias System Reverting back to original sample period of last 3 years of odds/results ... systems shouldn't be tweaked after a couple of bad days if they had been properly tested beforehand. Bets: 67 Wins: 32 Win Percentage: 48% Total Staked: 67pts Profit: -8.52pts Yield: -12.7% Today's qualifying bets: Preston 1.86 [Pinnacle] Stevenage 1.70 [Canbet] Quote Link to comment Share on other sites More sharing options...
Orpheus Posted December 13, 2008 Author Share Posted December 13, 2008 Re: Football Favourite-Longshot Bias System Bets: 69 Wins: 33 Win Percentage: 48% Total Staked: 69pts Profit: -8.66pts Yield: -12.6% Today's qualifying bets: Arsenal 1.80 [unibet] Aston Villa 1.74 [Pinnacle] Liverpool 1.40 [boylesports] Leyton Orient 1.95 [Pinnacle] Scunthorpe 1.90 [Totesport] Hamburger 1.45 [Pinnacle] Werder Bremen 2.15 [Centrebet] Bayer Leverkusen 1.36 [stan James] Napoli 1.53 [Five Dimes] Bari 1.73 [Pinnacle] Grosseto 2.06 [Pinnacle] Brechin 1.45 [Centrebet] East Stirlingshire 1.80 [Totesport] 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.