Dumyat
-
Posts
193 -
Joined
-
Last visited
Posts posted by Dumyat
-
-
Re: Lack of draws methodology Adding in the aways so that both the home and away team has a recent sequence of 10 games with a low strike rate doesn't seem to improve the results. There are 296 matches in that 20 year period of the English Premier where both the home and away side had a sequence involving a maximum of one draw in their immediately preceding 10 home or away results. ie exceptionally low draw strike rates which you would think would continue. Of those 296 matches 82 were drawn. That is 27.7% which is exactly the draw strike rate you would expect if you picked your matches with a pin. So many systems rely on sequences of results - and so many times when you investigate these sequences they dont actually predict the future - all they do is observe the past. There's the most recent "unexpected" draw from two teams with low draw sequences.
120071020
2
A
Everton
Liverpool
20071103
3
1
H
Everton
Birmingham
20071124
7
1
H
Everton
Sunderland
20071208
3
0
H
Everton
Fulham
20071226
2
0
H
Everton
Bolton
20071229
1
4
A
Everton
Arsenal
20080112
1
0
H
Everton
Man City
20080130
0
0
X
Everton
Tottenham
20080209
1
0
H
Everton
Reading
20080302
3
1
H
Everton
Portsmouth
20071110
0
5
A
Derby
West Ham
20071201
1
0
H
Chelsea
West Ham
20071209
0
1
A
Blackburn
West Ham
20071222
1
2
A
Middlesbrough
West Ham
20080101
2
0
H
Arsenal
West Ham
20080120
1
1
X
Man City
West Ham
20080202
1
0
H
Wigan
West Ham
20080223
0
1
A
Fulham
West Ham
20080305
4
0
H
Liverpool
West Ham
20080309
4
0
H
Tottenham
West Ham
20080322
1
1
X
Everton
West Ham -
Re: Lack of draws methodology The idea sounds good - if a team has a recent record of not drawing it is less likely to be involved in a draw in its next game. I like to see how systems would have worked on previous data. OK so lets take a look at the English Premier all the way from August 1989 to August 2008. There were 2047 draws from 7548 matches = 27.12%. I choose EP since I happen to have all the data for an extended period. Lets see what happens in the next game if the 10 preceding games had zero draws - (just choosing 10 since easy to calculate). There were 255 examples where a home team had not drawn any of its preceding 10 home fixtures - how many of the next 255 games were drawn when a home side had that low draw record - it was 86 for a strike rate of 34% ie it was much worse than random. How about only 1 of last 10 drawn - there were 1175 such matches but 316 were drawn on the next game which is back at the random strike rate of 27%. So it would seem very logical that a home team with a low draw strike rate would be less likely to draw but on this sample size and this set of English Premier data it's not happening. When I get a moment I'll add the away teams so that both have a low strike - I wonder if that improves things.
-
Re: Holland v Scotland 28th March 2009 Scotland dont play that many away games against elite teams in World and Euro qualifiers. Here's the recent record: 19931013 3 1 H H 2 1 Italy Scotland 19990609 3 2 A H 0 1 Czech_R Scotland 20030910 2 1 H H 1 0 Germany Scotland 20050326 2 0 H H 1 0 Italy Scotland 20070328 2 0 H H 1 0 Italy Scotland 20070912 0 1 X A 0 0 France Scotland Not a massive sample but 4 times out of 6 we have scored away v the quality sides. However the Burley period has all but extinguished the revival begun by Walter Smith so we could indeed get a doing.
-
Re: Should Bookies be reported to the Gambling Commission? >>>You were 'field checked' by a manager/manageress, and over a set period of time, you made a profit, therefore, you are bad for business. Yes - I made a substantial sum in a 6 month period in 2004 on a new 2nd half coupon Coral tried. Banning abusive customers is entirely reasonable but I was always perfectly polite - even after having a bet refused. Though I did try some sarcastic remarks to the effect that I doubt I would be refused if I wanted to bet the cartoon racing or the roulette. However my point is that if you offer a football coupon or whatever you have to take all bets from any valid customer (not the drunks, cretins etc) to that minimum stake. The minimum is sufficiently low that its not going to kill the business to honour the bet and the punter will get something on that desired trade. It doesn't matter if the bookie doesn't like the rule since it would be enshrined in the legislation of how bookies have to operate - ie if u want a licence this is the way you have to do business whether you like it or not.
-
Re: Should Bookies be reported to the Gambling Commission? Gambling firms requires a licence. Legislators can impose their conditions on how gambling operates. One condition might be that to offer any price in a shop you have to accommodate all punters to some minimum liability. Say £100. You can of course withdraw that price later if you dont fancy further trade but while it exists you have to honour all trade. I am banned to zero in my local branch of Corals. I went back after three years and the settler woman remembered me - even criminals are allowed to start afresh.
-
Re: Portuguese CS + Italian Draws >>>For the Italian league I propose a similar method to these using the CS of 0-0, 1-1 and 2-2 Wouldn't it be better value, not to mention easier, just to back the straight draw. The dutched CS odds on these scores are probably going to be a lower price than the draw odds - and you are not going to be happy if it's a 3-3 or a Chelsea 4-4!
-
Re: Euro 2008 Qualifiers - Weds 21st Nov Let's see - that was Israel on Saturday and Croatia today were not "motivated" to win. Maybe punters will have to re-think the idea that motivated teams are more likely to win and can justify artificially short odds.
-
Re: Euro 2008 Qualifiers - Weds 21st Nov I would make the England match 55% home, 15% away, 30% draw based on ELO Ratings. However ELO doesn't weight motivation, wanting to finish 2nd in group etc. The best value bet if the above figures are correct is CROATS +1 Asian at 2.14 now, 2.16 earlier. This is over a 20% edge. Not quite in the mega edges on the Israel-Russia game but another wrong price based on one of the teams needs a result and other doesn't care. If I was Croat am I not more likely to win Euro 2008 if Russia rather than England are participants. Is that not a motivation to try and eliminate them now in a risk free fixture?
-
Re: Euro 2008 Qualifiers Sat 17th Nov As mentioned by several contributors Russia have been chalked up at one of the most "wrong" prices ever. On ELO ratings it would be something like 30% home, 37% away, 33% draw. The true price on Russia is 2.7 but punters are betting at 1.5 - madness. So my punter's paranoia suggests something dodgy is going on but I just have to bet ISRAEL somehow. So I have £500 on Israel +1 goal ASIAN at a price of EVENS with CANBET. At least this way you get money back if its a margin one Russian victory.
-
Re: UNDER 2.5 goals using previous results Back to that database of 130 matches the SERALDINI correct scores are: 2-1, 10/130 = 7.69% - needs a break even price of 13 1-2, 6/130 = 4.6% - needs a break even price of 21.7 2-2, 5/130 = 3.84% - needs a break even price of 26.04 So it will not be possible to ever get a price bigger than the true chance of the score happening - so it will be adding negative expectation bets to create insurance with too high a premium. ******************************* Another interesting sort of the dbase is by price received. Do you think that the profits are made from the bigger or shorter prices? I split the dbase at 1.5 and shorter and bigger than 1.5. 1.5 and shorter - 58 selections, 580 points on 662 returned - an incredible 14% profit on turnover. Whereas the longer prices performed worse. 72 selections, 720 points staked and 713 returned. Just a statistical blip?
-
Re: UNDER 2.5 goals using previous results I have input all the selections in a spreadsheet since the thread started here in January 2006. I have ignored the minor dabble with a double and treble (!) and the internationals and just collated the data relating to the actual system bets. There have been 130 bets. If you use JTW's actual named stakes it would be 3645 units on and 4001 units back for a 9.5% profit on turnover. This ignores all the side bets on insurance etc. If you just put a flat £10 on each selection it would be 1300 units on and 1375 back for a 5.7% profit on flat stake turnover. The performance of all leagues used (at least 10 selections in that league) on flat stakes is: French 1: 37 selections = 370 units on 345 returned. French 2: 26 selections = 260 units on 234 returned. Greece: 18 selections = 180 units on 237 returned. Italy 2: 19 selections = 190 units on 258 returned. Spain 2: 13 selections = 130 units on 108 returned. So it's Greece and Italy 2 where the money has been made - however these are very small samples and it's hard to judge if these results are significant. Collating this material was tricky so errors are possible - I can pass on a copy of the sheet to JTW if interested.
-
Re: Fleet's "Not To Lose" Idea
Indeed if you are backing WIN/DRAW Newcastle then that is a lay of Spurs. However the lay price was astronomic - the win/draw prices means you were offering 3.6 to B365 on a Spurs win when best price anywhere was just 3.2. So even although you have a good system for finding teams not to lose the prices will often be too rubbish to make any long term profit because of the way the bet is made. ie backing win/draw with same bookie. So IMHO Asian 0 is a better way - asian always offered at lower margin. DRAW NO BET is a two option bet which is always lower margin than 3 etc. If you really can keep coming up with a series of teams not to lose then an acca of XNB, even if each is odds on and with some ending up void, will still be a good win.or...Lay multiples on Betfair? :ok -
Re: Fleet's "Not To Lose" Idea 1) If you back win and draw with the same book you are absorbing two chunks of their margin rather than one. 2) The value part of a bet is almost never the draw part so why waste stakes on that part - especially when you are paying such a massive premium price to get with the draw. 3) Why not use DRAW NO BET or the ASIAN 0 option to eliminate these pesky draws.
-
Re: Euro 2008 Matches (Sep 8) 19910327 1 1 X X 1 1 England Ireland 19890603 3 0 H H 1 0 England Poland 20051012 2 1 X H 1 1 England Poland 19990327 3 1 H H 2 1 England Poland 19921014 1 1 X X 0 0 England Norway 19901017 2 0 H H 1 0 England Poland 20010324 2 1 X H 1 1 England Finland 19981010 0 0 X X 0 0 England Bulgaria 20051008 1 0 H H 1 0 England Austria 19961009 2 1 H H 2 1 England Poland 20030611 2 1 A H 0 1 England Slovakia 20011006 2 2 A X 0 1 England Greece There's a dozen England EC, WC qualifying matches with a similar gap in strength to the task involved in beating Israel. 8/12 wins is bang on the 66% estimate I'd make for win chance. The chance of England winning by a large margin looks unlikely - it only happens against Poland! BET: Israel +1.25 means you win half on a margin 1 England victory - other half stake returned. This is available at 1/1 or better PINN which should be about a 10% edge.
-
Re: Euro 2008 Matches (Sep 8) Croatia -2 on Asians have shortened up from 1.85 to 1.8 to beat Estonia. The reverse fixture was just 1-0 away to the Croats in June - so it's by no means a cert win but I still make a price of 1.8 over a 20%+ edge. In the exact home v away matchup the teams have met a couple of times with Estonia holding to a draw in the more recent. However the stats reckon a Croatia win is over 90% chance and it's unlikely to be by just 1.
19950903 7 1 H H 4 1 Croatia Estonia 20020907 0 0 X X 0 0 Croatia Estonia Some important team news UEFA.COM: Estonia's preparations for their Group E double-header against Croatia and F.Y.R. Macedonia have suffered a significant blow after No1 goalkeeper Mart Poom was ruled out with a back injury. With Poom's understudy, Artur Kotenko, out with a broken finger, FK Bodo/Glimt's Pavel Londak is likely to add to his three international appearances. Estonia coach Viggo Jensen has called up FC Flora Tallinn's uncapped Mihkel Aksalu, 22, as backup. The price on -2 keeps contracting now about 1.72 - still a good edge. Stan James straight handicap on Croats -2 ie win by 3 or more to win is 2.375 and a big edge if you have already missed 1.8 on -2 Asian.
-
Re: Converting Elo ratings into odds Asc: how did the system work in practice ie 2006-2007 season - did u actually achieve 8% yield.
-
Re: Euro 2008 Qualifiers Aug 22, 2007 Serbia +(0) away to Belgium 1.8 :wall San Marino +2 home to Cyprus 1.79 :clap Andorra +1.75 away to Estonia 2.12 :clap Armenia +1.75 home to Portugal 1.95 :clap That went pretty well although I had thought Serbia was the most likely winner!
-
Anyone got any views on the 6 Euro 2008 qualifiers I have tried these ASIANS - one away fav and three weak teams trying to keep the margin of defeat low. Serbia +(0) away to Belgium 1.8 San Marino +2 home to Cyprus 1.79 Andorra +1.75 away to Estonia 2.12 Armenia +1.75 home to Portugal 1.95
European Championship Qualifying Group A Armenia v Portugal 17.00 Belgium v Serbia 19.45 Finland v Kazakhstan 17.00 European Championship Qualifying Group D San Marino v Cyprus 20.00 European Championship Qualifying Group E Estonia v Andorra 17.00 European Championship Qualifying Group F Northern Ireland v Liechtenstein 19.45 -
Re: Laying Correct Scores Hearts were around 10/11 to win - so the true 1-0 chance would be somewhere in the range 10-12%. If say 11% then laying edge after BF tax of 5% is less than 1% - but still a profit. As you mention bookies take vast margin on CS but much less on the more likely scores. I suspect if you offer 10% more than bookie best price then edge will go negative. eg at 8.8 and true 11% then -0.16%. But if you only go with 2.5% then may not get any action!
Lack of draws methodology
in Betting Systems & Strategy
Posted
Re: Lack of draws methodology First off I salute anyone who puts their system online. It is so much easier to knock something down that build something up. I am not trying to knock it down I am just pointing out that the plain meaning of the way you introduced this thread suggested that if you could find matches where there was a consistent pattern of few draws then that was likely to predict that the next match would be less likely than random to be a draw. The title is "Lack of draws A system based on consistency of results - Lack of draws". OK So I have looked at teams who have had that lack of draws and as you can see I cannot find any such pattern - at least in the English Premier. "In LOD you have teams that have much less than 10 games without a draw." Well how many games should you look at - I am going to guess it doesn't matter the length of the sequence, it doesn't predict the next match even if a team has a low or high draw strike rate in relation to that league. "You have to look at when those draws occurred and vs who?" Well by all means that might work - but what has that got to do with low draw sequences. It all down to personal assessment.