Cogge Posted January 20, 2009 Author Share Posted January 20, 2009 Re: Soccer Prediction Model Yeah, I think I'm gonna go down to level stakes and 4-5% of the total bank. Quote Link to comment Share on other sites More sharing options...
ProphetofProfit Posted January 20, 2009 Share Posted January 20, 2009 Re: Soccer Prediction Model 4-5% is still to much imo. You can see for yourself by doing the following in excel. Start with £1000 bank, change the bet size accordingly. Cell. A1-A500 =RAND() B1-A500 =IF(A1<0.5, stake*yield, -stake) C1-C500 = running profit/loss Column A creates random numbers, column B works out if your bet won of not, using a win % of 50% (A1<0.5 would mean that 50% are winners). stake * yield is what you win on each winning bet (eg 2.2 odds for a 50% shot where odds should be 2.0 is a yield of 2.2/2*100-100. Then create a graph from column C to track profit over the 500 bets. Change the bet size to see what happens. Then feel a little sick when you realise that the long term is a very long way away. I would personally not use more than 3% of bank unless I was gambling it up. That's only an estimate from looking at the graphs I made. Quote Link to comment Share on other sites More sharing options...
ProphetofProfit Posted January 20, 2009 Share Posted January 20, 2009 Re: Soccer Prediction Model I also suspect that kelly staking is superior, there are some articles on this at the punters paradise. I'm going to investigate another staking method where a fixed percentage of bank is multiplied by the win% of the bet. This reduces liability on longshots and is a lot like kelly staking since I'm betting more when my edge is large. Example: £1000 bank Bet 4% x win % Which is £40 x win % Odds are 2.2 50% win percentage = £20 bet 60% win percentage = £24 bet. Therefore when my edge is larger with the 60% win chance bet, my bet is also larger to capitalise, but not so large as to increase volatiliy, since the increased bet is compensated by an increased chance of winning and therefore less variance. Quote Link to comment Share on other sites More sharing options...
Cogge Posted January 20, 2009 Author Share Posted January 20, 2009 Re: Soccer Prediction Model But you are assuming that your edge is exact. Also, if win percentage = 100%, the bet should probably be 100% of bank instead of 4%, hehehe. Quote Link to comment Share on other sites More sharing options...
ProphetofProfit Posted January 20, 2009 Share Posted January 20, 2009 Re: Soccer Prediction Model But you are assuming that your edge is exact. Also, if win percentage = 100%, the bet should probably be 100% of bank instead of 4%, hehehe. No I'm not. But I can say for sure that when I bet on Man Utd at home to Stoke that the win % is greater than betting on Stoke away to Man Utd. And this doesn't change the fact that 10% of bank is a suicidal staking strategy unless your goal is to go broke. 5% is safer but I still wouldn't like to be in a situation where a bad few weeks nearly annihilates my bankroll. Quote Link to comment Share on other sites More sharing options...
Cogge Posted January 21, 2009 Author Share Posted January 21, 2009 Re: Soccer Prediction Model - Bank: €141.332 Profit: +€41.332 Yield: 8.5% No I'm not. But I can say for sure that when I bet on Man Utd at home to Stoke that the win % is greater than betting on Stoke away to Man Utd. And this doesn't change the fact that 10% of bank is a suicidal staking strategy unless your goal is to go broke. 5% is safer but I still wouldn't like to be in a situation where a bad few weeks nearly annihilates my bankroll. A bit harsh :P. Anyway, maybe I didn't get the staking plan but: £1000 bank Bet 4% x win % Which is £40 x win % Odds are 2.2 50% win percentage = £20 bet 60% win percentage = £24 bet When you say win %, I assumed you meant your calculated win %? Or did you mean 1/(odds*overround) or something? Of course the win % will be higher for Man U than for Stoke, but that's generally reflected in the odds to some extent. Basically, is your betting plan to stake more when you believe you have a larger edge, or is your betting plan to win the same amount on each bet? For example: Bank = £1000 1. Bet more on believed edge) Bet 1: Odds 2.2, Estimated probability of win: 50%. Stake= £40*0.5=£20 Bet 2: Odds 1.4, Estimated probability of win: 80%. Stake= £40*0.8=£32 etc... 2. Want to win the same amount) Want to win 4% of bank (£40) Bet 1: Odds = 2.2, stake = £40/1.2 = £33.33333 Bet 2: Odds = 1.5, stake = £40/0.5 = £80 etc... Quote Link to comment Share on other sites More sharing options...
Zoroaster Posted January 21, 2009 Share Posted January 21, 2009 Re: Soccer Prediction Model I also suspect that kelly staking is superior, there are some articles on this at the punters paradise. I'm going to investigate another staking method where a fixed percentage of bank is multiplied by the win% of the bet. This reduces liability on longshots and is a lot like kelly staking since I'm betting more when my edge is large. Example: £1000 bank Bet 4% x win % Which is £40 x win % Odds are 2.2 50% win percentage = £20 bet 60% win percentage = £24 bet. Therefore when my edge is larger with the 60% win chance bet, my bet is also larger to capitalise, but not so large as to increase volatiliy, since the increased bet is compensated by an increased chance of winning and therefore less variance. If you are going to vary the size of your bet stake on each bet, surely it should be related to the size of the edge (value) in the bet, rather than just lumping on tiny odds and being scared of long odds. Quote Link to comment Share on other sites More sharing options...
ProphetofProfit Posted January 21, 2009 Share Posted January 21, 2009 Re: Soccer Prediction Model If you are going to vary the size of your bet stake on each bet' date=' surely it should be related to the size of the edge (value) in the bet, rather than just lumping on tiny odds and being scared of long odds.[/quote'] It probably should, but I haven't fully looked into kelly for some reason. And I'm no expert on statistics but my idea makes sense to me. Flat stake of 2% works awesome with average win % of 50 but falls apart slightly when your average win % is 20, since the variance increases and it's similar to betting 4-5% on 50% shots in terms of bankroll variance. All I was trying to do was adjust for the variance caused by the win percentage being smaller than usual, and this was what I came up with. Say you're betting on teams with a 50% win chance, and you do this 100 times, and then all of a sudden you come up with a list of teams that are all 10% to win, and you have 50 bets to make. That's a lot of variance for 2% flat bets. Cogge, I might be a bit harsh but it's not with the intention of being a dick or annoying you but to save you from yourself and your crazy staking plan :) Your method of bet selection probably has a lot of potential but staking incorrectly could ruin it. I see so many mad strategies that I feel like I have to say something when I see people playing accumulators at bookmakers when betfair or betdaq is only a few keystrokes away. I thought people learned this years ago when Joe Buchdahl was doing the rounds on these forums. And to clarify I mean changing the bet size according to win % and not odds. And again I don't know if this works better than kelly or flat stakes. If anyone knows how to run simulations to test it then please do. Quote Link to comment Share on other sites More sharing options...
ProphetofProfit Posted January 21, 2009 Share Posted January 21, 2009 Re: Soccer Prediction Model - Bank: €141.332 Profit: +€41.332 Yield: 8.5% http://www.punterslounge.com/forum/f29/staking-systems-long-term-betting-7038/index3.html See posts 42 and 57 Quote Link to comment Share on other sites More sharing options...
Cogge Posted January 21, 2009 Author Share Posted January 21, 2009 Re: Soccer Prediction Model - Bank: €141.332 Profit: +€41.332 Yield: 8.5% And by win % you mean what you have estimated their win % is? For example: A two-way odds of 1.9 - 1.9 generally means the bookmaker has assessed the probability for each way to 50% - 50%. You then estimate the "true" probability to 60% for team one. Does this mean that you stake: (4% of bankroll) * 0.6 ? Quote Link to comment Share on other sites More sharing options...
ProphetofProfit Posted January 21, 2009 Share Posted January 21, 2009 Re: Soccer Prediction Model - Bank: €141.332 Profit: +€41.332 Yield: 8.5% Yes that's it, 4% of bankroll x 0.6. But I haven't done any research myself to prove that 4% is the optimal amount. I got that number using US sports bettors estimates of betting 1 or 2% of their bankroll per bet. They are betting handicaps where the win percentage is about 50%. So I took these numbers, and figured that if betting 2% on a 50% shot is a safe approach to bankroll management, then i'll use double this number and use that as a start to work out how much to bet. I hope that makes sense. Quote Link to comment Share on other sites More sharing options...
Zoroaster Posted January 22, 2009 Share Posted January 22, 2009 Re: Soccer Prediction Model - Bank: €141.332 Profit: +€41.332 Yield: 8.5% Yes that's it' date=' 4% of bankroll x 0.6. But I haven't done any research myself to prove that 4% is the optimal amount. I got that number using US sports bettors estimates of betting 1 or 2% of their bankroll per bet. They are betting handicaps where the win percentage is about 50%. So I took these numbers, and figured that if betting 2% on a 50% shot is a safe approach to bankroll management, then i'll use double this number and use that as a start to work out how much to bet. I hope that makes sense.[/quote'] Are your PM's not working mate ? Quote Link to comment Share on other sites More sharing options...
ProphetofProfit Posted January 22, 2009 Share Posted January 22, 2009 Re: Soccer Prediction Model - Bank: €141.332 Profit: +€41.332 Yield: 8.5% Apparently not, I can't see any. I'll send you mail. Quote Link to comment Share on other sites More sharing options...
ScottyXS Posted January 23, 2009 Share Posted January 23, 2009 Re: Soccer Prediction Model - Bank: €141.332 Profit: +€41.332 Yield: 8.5% Apparently not, I can't see any. I'll send you mail. Need 50 posts first in order for you to be able to send/recieve PMs :ok Quote Link to comment Share on other sites More sharing options...
Cogge Posted January 23, 2009 Author Share Posted January 23, 2009 Re: Soccer Prediction Model Going to change to level stakes according to: Lvl 1 - Stake: €5 Lvl 2 (When bank hits €200) - Stake: €10 Lvl 3 (When bank hits €300) - Stake: €15 etc... Oki guys, time for this weekends matches: All odds are from Goalwin Ligue 1 Not playing :cry Primera Division Sunday Valladolid vs Espanyol. Bet Valladolid @ 2.08 Almeria vs Athletic Bilbao. Bet: Almeria @ 2.46 Getafe vs Sporting Gijon. Bet: Getafe @ 2.02 Serie A Saturday Reggina vs Chievo. Bet: Reggina @ 2.12 Sunday Lecce vs Torino. Bet: Lecce @ 2.25 Palermo vs Udinese. Bet: Palermo @ 2.02 Eredivisie No bets found by model. :cry Premier League - Midweek matches 27th Sunderland vs Fulham. Bet: Sunderland @ 2.3 28th West Ham vs Hull. Bet: West Ham @ 1.81 Bundesliga Not playing this weekend. :cry Liga Sagres Today (Friday) Maritimo vs Lexioes. Bet: Maritimo @ 2.12 Starting bank: €141.332 Number of bets: 9 Stake per bet: €5 Total stake: €45 Remaining bank after placed bets: €96.332 Might throw up some more bets this weekend if odds start changing for some matches. Napoli v Roma looked like a good 1X bet, but the scales need to tip a bit more over on Roma (which I believe they will). :hope Wish me luck :P :hope Quote Link to comment Share on other sites More sharing options...
Cogge Posted January 29, 2009 Author Share Posted January 29, 2009 Re: Soccer Prediction Model Going to change to level stakes according to: Lvl 1 - Stake: €5 Lvl 2 (When bank hits €200) - Stake: €10 Lvl 3 (When bank hits €300) - Stake: €15 etc... Oki guys, time for this weekends matches: All odds are from Goalwin Ligue 1 Not playing :cry Primera Division Sunday Valladolid vs Espanyol. Bet Valladolid @ 2.08 DRAW :sad Almeria vs Athletic Bilbao. Bet: Almeria @ 2.46 :nana €12.3 returned Getafe vs Sporting Gijon. Bet: Getafe @ 2.02 :nana €10.1 returned Serie A Saturday Reggina vs Chievo. Bet: Reggina @ 2.12 :sad Sunday Lecce vs Torino. Bet: Lecce @ 2.25 DRAW :sad Palermo vs Udinese. Bet: Palermo @ 2.02 :nana €10.1 returned Eredivisie No bets found by model. :cry Premier League - Midweek matches 27th Sunderland vs Fulham. Bet: Sunderland @ 2.3 :nana €11.5 returned 28th West Ham vs Hull. Bet: West Ham @ 1.81 :nana €9.05 returned Bundesliga Not playing this weekend. :cry Liga Sagres Today (Friday) Maritimo vs Lexioes. Bet: Maritimo @ 2.12 DRAW:sad Starting bank: €141.332 Number of bets: 9 Stake per bet: €5 Total stake: €45 Bets won: 5 Bets lost: 4 Stake returned: €53.05 Weekend result: +€8.05 Bank after weekend: €149.832 Lots of draws this round, looked bad after weekend but the EPL-bets saved it :P Total Staked: €528.5 Total Profit: +€49.832 Bets won: 32/59 Strike rate: 54.2% Yield: 9.4% Quote Link to comment Share on other sites More sharing options...
Cogge Posted January 29, 2009 Author Share Posted January 29, 2009 Re: Soccer Prediction Model Last few weekends have been showing a nice growth, hopefully it will continue and with all leagues playing this weekend we're bound for some nice action! For this weekends matches, some matches haven't got odds yet. I'm waiting for the market fluctuations to settle, will place the bets later tonight and post the odds then. If the bets have odds I have placed those bets already.: All odds are from Goalwin Ligue 1 Saturday Auxerre v Lorient. Bet: Auxerre @ 2.56 Sunday Bordeaux v Lille. Bet: Bordeaux @ 1.88 Valenciennes v Nice. Bet: Valenciennes @ 2.54 Primera Division Saturday Athletic Bilbao v Malaga. Bet: Athletic Bilbao @ 2.12 Sunday Espanyol v Recreativo. Bet: Espanyol @ 2.04 Deportivo v Villarreal. Bet: Deportivo @ 2.56 Serie A Saturday Napoli v Udinese. Bet: Napoli @ 1.9 Sunday Siena v Lecce. Bet: Siena @ 1.95 Atalanta v Catania. Bet: Atalanta @ 2.02 Genoa v Palermo. Bet: Genoa @ 1.96 Eredivisie Sunday NEC v Feyenoord. Bet: NEC @ 2.13 Premier League Sunday Liverpool v Chelsea. Bet: Liverpool @ 2.5 Bundesliga Bochum v Karlsruher. Bet: Bochum @ 2.04 Liga Sagres Sunday Maritimo v Naval. Bet: Maritimo @ 1.84 Nacional v Leixoes. Bet: Nacional @ 2.02 Starting bank: €149.832 Number of bets: 15 Stake per bet: €5 Total Stake: €75 Remaining bank: €74.832 :hope Wish me luck :P :hope Quote Link to comment Share on other sites More sharing options...
marsel05 Posted January 29, 2009 Share Posted January 29, 2009 Re: Soccer Prediction Model - Bank: €149.832 Profit: +€49.832 Yield: 9.4% What is the criteria for these matches, i am thinking to follow you and bet on every match 5 euro. Quote Link to comment Share on other sites More sharing options...
Cogge Posted January 29, 2009 Author Share Posted January 29, 2009 Re: Soccer Prediction Model - Bank: €149.832 Profit: +€49.832 Yield: 9.4% Check the first post in this thread, it's quite extensive. :P Basically I have a model that works with a ton of statistics to determine the probability of each game. Then I bet on home team favorites where the model predict they are a larger favorite than the odds bookmakers are offering. Then I've set a limit at around 7% value, which means if a bookmaker puts out an odds at 2.0 and my model predicts that the odds should be 1.75 then the difference between the probabilities are 7.1% (1/1.75-1/2). In this case the difference is above 7% so I would place a bet, however if the model had predicted 1.85 instead, no bet would have been placed as the difference would have been 4%. If that makes any sense to you :P Short version: Predict game - bet when bookmaker has made a mistake. hehe Quote Link to comment Share on other sites More sharing options...
marsel05 Posted January 29, 2009 Share Posted January 29, 2009 Re: Soccer Prediction Model - Bank: €149.832 Profit: +€49.832 Yield: 9.4% Check the first post in this thread, it's quite extensive. :P Basically I have a model that works with a ton of statistics to determine the probability of each game. Then I bet on home team favorites where the model predict they are a larger favorite than the odds bookmakers are offering. Then I've set a limit at around 7% value, which means if a bookmaker puts out an odds at 2.0 and my model predicts that the odds should be 1.75 then the difference between the probabilities are 7.1% (1/1.75-1/2). In this case the difference is above 7% so I would place a bet, however if the model had predicted 1.85 instead, no bet would have been placed as the difference would have been 4%. If that makes any sense to you :P Short version: Predict game - bet when bookmaker has made a mistake. hehe I like it, i will try it this weekend, and also will bet on my System which you saw (Favorite win, first match was draw). Quote Link to comment Share on other sites More sharing options...
Cogge Posted January 29, 2009 Author Share Posted January 29, 2009 Re: Soccer Prediction Model - Bank: €149.832 Profit: +€49.832 Yield: 9.4% If you're following my bets, make sure you get around the same odds as me, otherwise it's pointless. For example if I get 2 odds somewhere, and the model estimates it to be 1.75 I will bet, but if you get 1.9 odds on the same match somewhere else, you shouldn't bet. If you bet at 1.9 in that case, then you aren't following my model :P Quote Link to comment Share on other sites More sharing options...
marsel05 Posted January 29, 2009 Share Posted January 29, 2009 Re: Soccer Prediction Model - Bank: €149.832 Profit: +€49.832 Yield: 9.4% I placed 15 bets on the next odds : Auxerre v Lorient. Bet: Auxerre @ 2.56............2.40 Bordeaux v Lille. Bet: Bordeaux @ 1.88............1.85 Valenciennes v Nice. Bet: Valenciennes @ 2.54........2.45 Athletic Bilbao v Malaga. Bet: Athletic Bilbao @ 2.12.....2.05 Espanyol v Recreativo. Bet: Espanyol @ 2.04......2.00 Deportivo v Villarreal. Bet: Deportivo @ 2.56.....2.50 Napoli v Udinese. Bet: Napoli @...............1.85 Siena v Lecce. Bet: Siena @ 1.95...........1.90 Atalanta v Catania. Bet: Atalanta @ 2.02.......1.90 Genoa v Palermo. Bet: Genoa @ 1.96..........1.90 NEC v Feyenoord. Bet: NEC @........2.10 Liverpool v Chelsea. Bet: Liverpool @ .......2.40 Bochum v Karlsruher. Bet: Bochum @ 2.04.......2.00 Maritimo v Naval. Bet: Maritimo @......1.80 Nacional v Leixoes. Bet: Nacional @.....1.95 Quote Link to comment Share on other sites More sharing options...
Cogge Posted January 29, 2009 Author Share Posted January 29, 2009 Re: Soccer Prediction Model - Bank: €149.832 Profit: +€49.832 Yield: 9.4% Best of luck to both of us then :P Quote Link to comment Share on other sites More sharing options...
marsel05 Posted January 29, 2009 Share Posted January 29, 2009 Re: Soccer Prediction Model - Bank: €149.832 Profit: +€49.832 Yield: 9.4% I found 5 matches for tomorrow where the first match finished draw, you can try if you want. Quote Link to comment Share on other sites More sharing options...
Cogge Posted February 1, 2009 Author Share Posted February 1, 2009 Re: Soccer Prediction Model All odds are from Goalwin Ligue 1 Saturday Auxerre v Lorient. Bet: Auxerre @ 2.56 :sadDRAW Sunday Bordeaux v Lille. Bet: Bordeaux @ 1.88 :sad DRAW Valenciennes v Nice. Bet: Valenciennes @ 2.54 :nana+€12.7 Primera Division Saturday Athletic Bilbao v Malaga. Bet: Athletic Bilbao @ 2.12 :nana+€10.6 Sunday Espanyol v Recreativo. Bet: Espanyol @ 2.04 :sad DRAW Deportivo v Villarreal. Bet: Deportivo @ 2.56 :nana +€12.8 Serie A Saturday Napoli v Udinese. Bet: Napoli @ 1.9 :sadDRAW Sunday Siena v Lecce. Bet: Siena @ 1.95 :sad Atalanta v Catania. Bet: Atalanta @ 2.02 :nana +€10.1 Genoa v Palermo. Bet: Genoa @ 1.96 :nana +€9.8 Eredivisie Sunday NEC v Feyenoord. Bet: NEC @ 2.13 :nana+€10.65 Premier League Sunday Liverpool v Chelsea. Bet: Liverpool @ 2.5 :nana Amazing push from Torres with 2 89th and 90th min goals! +€12.5 Bundesliga Bochum v Karlsruher. Bet: Bochum @ 2.04 :nana +€10.20 Liga Sagres Sunday Maritimo v Naval. Bet: Maritimo @ 1.84 :nana +€9.3 Nacional v Leixoes. Bet: Nacional @ 2.02 :sad DRAW Weekend Results: Starting bank: €149.832 Number of bets: 15 Stake per bet: €5 Total stake: €75 Bets won: 9 Bets lost: 6 Stake returned: €98.65 Weekend result: +€23.65 Bank after weekend: €173.482 Weekend Yield: 31.5% Total Staked: €603.5 Total Profit: +€73.482 Bets won: 41/74 Strike rate: 55.4% Yield: 12.1% Great result this weekend. Some later goals went the system's way. Next weekend's bets will be up sometime around tuesday / wednesday. Quote Link to comment Share on other sites More sharing options...
marsel05 Posted February 1, 2009 Share Posted February 1, 2009 Re: Soccer Prediction Model - Bank: €173.832 Profit: +€73.832 Yield: 12.1% Good matches, Well done, i will bet on your picks from now. :clap:clap:clap:clap:clap:clap:clap:clap Quote Link to comment Share on other sites More sharing options...
Cogge Posted February 3, 2009 Author Share Posted February 3, 2009 Re: Soccer Prediction Model - Bank: €173.832 Profit: +€73.832 Yield: 12.1% Figured I could do a recap of how the betting has gone so far: EPL Bets won: 7 / 9 Strike rate: 78% Average odds: 2.27 Eredivisie Bets won: 2 / 3 Strike rate: 67% Average odds: 2.19 Liga Sagres Bets won: 2 / 6 Strike rate: 33% Average odds: 1.935 Ligue 1 Bets won: 8 / 17 Strike rate: 47% Average odds: 2.17 Bundesliga Bets won: 3 / 3 Strike rate: 100% Average odds: 1.95 Serie A Bets won: 8 / 18 Strike rate: 44.5% Average odds: 2.06 Primera Divisione Bets won: 11 / 18 Strike rate: 61% Average odds: 2.11 Well, Bundesliga, EPL, Eredivisie and Primera Divisione are all going very well. Hopefully Liga Sagres will pick up soon and get above 50% as well :P Quote Link to comment Share on other sites More sharing options...
Cogge Posted February 3, 2009 Author Share Posted February 3, 2009 Re: Soccer Prediction Model - Bank: €173.832 Profit: +€73.832 Yield: 12.1% A late picks for tonights' Eredivisie game Bet placed at Goalwin Tonight Willem II vs FC Volendam. Bet Willem II @ 1.73 Starting Bank: €173.482 Bets: 1 Stake per bet: €5 Remaining bank: €168.482 Quote Link to comment Share on other sites More sharing options...
Gingertipster Posted February 3, 2009 Share Posted February 3, 2009 Re: Soccer Prediction Model - Bank: €173.832 Profit: +€73.832 Yield: 12.1% Early days Cogge, but impressive stuff so far. Will follow with interest (without betting). As far as the staking goes. What I do is probably something similar to Kelly (don't really know what that is). But I bet in points. With each selection having the same points as my own opinion of percentage. So if I believe something has an even money 2:00 chance, I back it 50 points @ the biggest price available. As long as it is bigger than Evens (plus a margin for error) of course. Something I think has 6/4 2.5 would have 40 points. You could try the same but just do stakes to the price your system tells you it should be. Even if you divide each figure by 5 or 10 or whatever. With 8 or 4 points on a 6/4 chance. Good luck Ginge Quote Link to comment Share on other sites More sharing options...
Cogge Posted February 4, 2009 Author Share Posted February 4, 2009 Re: Soccer Prediction Model - Bank: €173.832 Profit: +€73.832 Yield: 12.1% Bet placed at Goalwin Tonight Willem II vs FC Volendam. Bet Willem II @ 1.73 :sad Well that went in the crapper. One goal lead and they lost it :( Result: -€5 Remaining bank: €168.482 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.