What percentages for different factors?

[iroquois]

What percentage is ideal for the crucial factors when analising a game? for example: 1) Home/away = ? 2) form = ? 3) h2h = ? 4) offensive capacity = ? 5) defensive capacity = ? 6)last home/away game = ? 7) last form game = ? 8) h2h goals between the two teams = ? what are for you the crucial ones? some missing? what you think of that?

[Grex]

Re: What percentages for different factors? I don't think there is a set of "concrete" rules regarding the weighting of the various factors you have listed. My advice is to try and keep it simple.... if you find a "factor" that produces good results in back-testing, then stick to it.... don't complicate matters by introducing too many variables. This is only my personal advice.... probably some statistics experts on this forum have other ideas. By the way, the Moderators will probably remove the link to your web-site... I think you have to give recognition to Punters Lounge on your site...... I am afraid my Portugese is not very good (i.e. = 0) so I couldn't really appreciate the commentary on your site ;)

[lamas1984]

Re: What percentages for different factors? és tuga iroquois? XD eu também 2 portugueses in punterslounge? maybe there s more XD

[Gingertipster]

Re: What percentages for different factors? Am not a football punter, but it is similar to racing in that every game/race is different. e.g. If Rooney hasn't played for some time and Berbatov is out of form, then Man U's offensive capacity will recieve a big share of their percentage chance. If they were in reasonable form then it would not have such a big share and another aspect would have a greater share. If three of the normal back four is missing, then defensive capacity will be a big share. Similarly, if either offence or defence is in outstanding form then it would also get a big share of the percentage chance. Every game / race has different percentages given to different aspects.

[iroquois]

Re: What percentages for different factors?

és tuga iroquois? XD eu também 2 portugueses in punterslounge? maybe there s more XD

claro pá!!! que alegria hoje...4-0.

[iroquois]

Re: What percentages for different factors? i understand you. But, always have a but...imagine you want to bet in 10, 20 games in one saturday...you spend a lot of time looking for notices in the newspapers and internet. if you look some excepcional soccer statistic sites you can economize a lot of time, attending the statistics. i believe that a junction of the three (statistics, hot notices and intuiton) are the best way to analise a game...in the end is always a question of time. sorry my english.

[Mulkis]

Re: What percentages for different factors? IMHO all factors are already incorporated in the odds for a specific game. I see no sense to analyse all those statistics and team news any futher. Its already in the price.

[leftism]

Re: What percentages for different factors? If all data is already in the price then how are you supposed to find value? Surely the whole point of any successful system is to find and exploit discrepencies between statistical probabilities and actual market prices... Or have i got this whole betting thing wrong? :\

[Jase82]

Re: What percentages for different factors?

IMHO all factors are already incorporated in the odds for a specific game. I see no sense to analyse all those statistics and team news any futher. Its already in the price.

Eh? are you saying there's no point working out your own odds, and what is a fair price? Am I reading that right, because that just seems crazy.

[Gingertipster]

Re: What percentages for different factors?

If all data is already in the price then how are you supposed to find value? Surely the whole point of any successful system is to find and exploit discrepencies between statistical probabilities and actual market prices... Or have i got this whole betting thing wrong? :\

To misquote Eric Morcombe. All data is in the price, but not neccessarily in the right order. A bookmaker probably does does allow for all data, but he sometimes over or underestimates a particular statistic in a particular market. Therefore making a value bet for the punter. It is however, unlikely the same stat is over or under estimated every time, which is why most systems fail.

[leftism]

Re: What percentages for different factors?

To misquote Eric Morcombe. All data is in the price, but not neccessarily in the right order. A bookmaker probably does does allow for all data, but he sometimes over or underestimates a particular statistic in a particular market. Therefore making a value bet for the punter. It is however, unlikely the same stat is over or under estimated every time, which is why most systems fail.

fair point! Statistical systems tend to run out of steam in the long run either because enough people cop onto it and the market corrects itself OR the bookie figures it out.... Just trying to get as much £££ out of it before this happens i guess!

[leftism]

Re: What percentages for different factors? I suppose a very good example of the danger with statistics is the Chelsea fiasco over the last month. They kept cropping up in a statistical system i've been running over the last few months, but no data in the world could have predicted the effect Ray Wilkins departure would have on dressing room morale and subsequent performance on the field... Guess thats why they say its a funny old game! :rollin

[Mulkis]

Re: What percentages for different factors?

If all data is already in the price then how are you supposed to find value? Surely the whole point of any successful system is to find and exploit discrepencies between statistical probabilities and actual market prices... Or have i got this whole betting thing wrong? :\

Thats the point. Bookmakers odds are quite accurate. I've checked this. The corelation line is almost straight and R2 number is close to 1. Yes, there is bookmakers overround deducted from the price, but anyway overround makes only a few % difference. "Value betting" is solid and scientific theory based on bullet proof maths and statistic. But its unlikely ordinary punter could capitalise on this. Bookmakers odds are very much "true" odds. One should better concentrate his effort to "find winners", than "true odds". Just take a look at Grex or jtw treads - all they are doing is looking for winners "and odds will care for themselves":)

[Pepen]

Re: What percentages for different factors?

Bookmakers odds are very much "true" odds. One should better concentrate his effort to "find winners"' date=' than "true odds". Just take a look at Grex or jtw treads - all they are doing is looking for winners "and odds will care for themselves":)[/quote'] Actually Mulkis, isn't what Grex is doing, in his o/u 2.5 thread a statistical system? Afaik he's looking some games back and estimates his fair odds (= the system true odds), according to various variables, actually estimating some parameters for the different lagged ones (of which variables I'm not aware)

[Gingertipster]

Re: What percentages for different factors?

Thats the point. Bookmakers odds are quite accurate. I've checked this. The corelation line is almost straight and R2 number is close to 1. Yes' date=' there is bookmakers overround deducted from the price, but anyway overround makes only a few % difference. "Value betting" is solid and scientific theory based on bullet proof maths and statistic.[b'] But its unlikely ordinary punter could capitalise on this. Bookmakers odds are very much "true" odds. One should better concentrate his effort to "find winners", than "true odds". Just take a look at Grex or jtw treads - all they are doing is looking for winners "and odds will care for themselves":)

Who is this ""ordinary punter"? Is it a "mug punter"? Take a look at some "early prices", if you take the over-round off the bookies prices, there are times when one bookies odds compiler believes the odds given by another bookmaker is VALUE. So they can't be that "accurate" can they?:eyes What you are probably judging your "true odds" on are SP's, which are bound to be more accurate. It is up to the punter to find the value, before the odds become "true". If all you are doing is "looking for winners", you'd just as well back the favourite every time; as in most cases have the best chance of winning. But you won't make an over all profit. I think every punter has some inkling for "value", whether they are looking for value or "winners". :ok

[Mulkis]

Re: What percentages for different factors? Yes yes, "value betting" is the only way to move around in betting, to achieve profit in long run, bla bla bla... If I were a bookmaker I would do my best to persuade all the folks that my odds are not true odds, that there are some other true odds, and if you are "smart" enough and not a "mug punter" you would find those odds. So keep trying, keep on betting... Bullshit. One could check accuracy of bookmakers odds, it simple statistic, not rocket science, and one would get almost straight corelation line.

[Gingertipster]

Re: What percentages for different factors?

Yes yes' date=' "value betting" is the only way to move around in betting, to achieve profit in long run, bla bla bla... If I were a bookmaker I would do my best to persuade all the folks that my odds are not true odds, that there are some other true odds, and if you are "smart" enough and not a "mug punter" you would find those odds. So keep trying, keep on betting... Bullshit. One could check accuracy of bookmakers odds, it simple statistic, not rocket science, and one would get almost straight corelation line.[/quote'] Mulkis, Just because the vast majority of bookmakers odds are fairly accurate, does not mean there aren't one or two in each market who's chance has been under-estimated. If you don't believe me, then just take a look at my thread in the At The Races section. All I do is look for horses who I believe have been under-estimated by bookmakes. As you can see from the difference with most prices I take to SP's; it is possible to find the value bet/s often enough to make a profit. Monterosso also makes a good profit in a similar way. I am sure it is possible for those who know a lot about football to do the same. Just because you are incapable of doing it, does not mean it is impossible for everyone. Of course most people are incapable. To know the form is not good enough, a punter needs to be good at evaluating form. :ok You need to know the subject very well indeed. I know enough about horse racing to be in a team that takes on the Racing Post (NO BS) and other teams, made up of well known and less well known racing journalists and professionals, in a racing quiz once a year. Not the one with the best general knowledge in my team; but am the one who knows most about form. Suggest you study your sport. Good Luck Mulkis, you're going to need it. ;)

[Mulkis]

Re: What percentages for different factors? Have no idea about horses. But feel quite comfortable at statistic science. There cant be two correct probabilities at the same time. There is only one corect answer. And bookies reflect that answer in their odds quite precisely. And I dont think its feasible to spend time looking for some other probabilities from another dimension. So how to play it then? My answer is "active betting". It involves taking starting position and then at some predetermined time exiting, redistributing or keeping it going. I am happy with my results and I dont care about "value". Actually it is all I wanted to say - "value betting" is not the only one way to move around.:ok

[Gingertipster]

Re: What percentages for different factors? But you were not saying "but it is not the only way" Mulkis. You said it was....

... Bullshit.

[Gingertipster]

Re: What percentages for different factors? A punter can occasionally take best bookmakers prices for a game of football, home win, draw and away win; it is possible to back all three outcomes and still come out on top. One bookie might offer: 13/8 (38.1%) Home, 2/1 (33.3%) Draw and 15/8 (34.8%) Away. 38.1 + 33.3 + 34.8 = 106.2. Betting to 106.2% Another might offer: 11/8 (42.1%) Home, 9/4 (30.8%) Draw and 2/1 (33.3%) Away. Betting to 106.2% Another might offer: 11/8 (42.1%) Home, 2/1 (33.3%) Draw and 9/4 (30.8%) Away. Betting to 106.2% Another might offer: 11/8 (42.1%) Home, 15/8 (34.8%) Draw and 5/2 (28.6%) Away. Betting to 105.5% Another might offer: 6/4 (40%) Home, 85/40 (32%) Draw and 2/1 (33.3%) Away. Betting to 105.3% It might look as though all bookmakers prices are accurate. Yet with top prices of: Home 13/8 Draw 9/4 Away 5/2 A punter could put 38 points on the Home win, 31 on a Draw and 29 on an away win. 38 + 31 + 29 = total stake is 98 points. 38 X 1.625 (13/8) = 61.75 + 38 points stake = 99.75 returned for a profit of 1.75 points. 31 X 2.25 (9/4) = 69.75 + 31 points stake = 100.75 returned for a profit of 2.75 points. 29 X 2.5 (5/2) = 72.5 + 29 points stake = 101.5 returned for a profit of 3.5 points. Whichever outcome rusults in a profit. It may look as though bookmakers have accurate early prices. But in reality, there does not have to be much variation for a punter to find value. Even when the best prices percentages come to over 100%, does NOT mean there isn't any value there. One or two individual prices may well be value. After all, with Early Prices and value, it's only a bookmaker's odds compiler's opinion against an individual punter's opinion.

[Mulkis]

Re: What percentages for different factors? Gingertipster, Your example above, you could call it "value betting", but it is just simple arbitrage. And ROI is only around 2%... And you didn't mentioned all the risks involved. There are a lot of unexperienced folks in this forum looking for ideas and inspirations. I think more experienced punters should take more responsible approach what they are posting here and what advices they give. It could inflict much more damage than just stepping in bullshit:cow

[Gingertipster]

Re: What percentages for different factors? Mulkis, My example was not an example of what a punter should do. It was to establish the bookmakers prices are often not accurate enough. A bookmaker will usually put a mark up on any individual price of between 0.25% and 3.5%. ie. In a competitive market a bookmaker who believes a horse to be a 25% (fair 3/1 shot); will only add around 2% to that percentage and offer 11/4 (just 1.7% difference). Unless it is a difficult horse / team to assess, he might add around 3.5% and offer 5/2. He can not afford to offer less, because he won't get any trade, with other bookies offering bigger prices. So with any individual selection believed to be a 25% chance that is offered at 11/4 (26.7%), a bookie only has to be 1.7% wrong and it is "value". No odds compiler will be able to work out EVERY price that accurately. He is bound to get it wrong sometimes. And even if he did; he has to react to the betting market: If there was no money for this horse it would lengthen in price. So whether the odds-compiler thinks it should be an 11/4 shot does not matter. If there is little money for the horse it will drift out. It may drift to a backable price.

[Gingertipster]

Re: What percentages for different factors?

Gingertipster, Your example above, you could call it "value betting", but it is just simple arbitrage. And ROI is only around 2%... And you didn't mentioned all the risks involved. There are a lot of unexperienced folks in this forum looking for ideas and inspirations. I think more experienced punters should take more responsible approach what they are posting here and what advices they give. It could inflict much more damage than just stepping in bullshit:cow

:puke How exactly have I been irresponsible? :unsure If it stops novice punters trying to just back "winners", without taking value in to account, then I have saved them money.

[DM99]

Re: What percentages for different factors?

So how to play it then? My answer is "active betting". It involves taking starting position and then at some predetermined time exiting, redistributing or keeping it going. I am happy with my results and I dont care about "value". Actually it is all I wanted to say - "value betting" is not the only one way to move around.:ok

Odds = probability of winning, just expressed differently. If the bookmaker always has the correct odds (and thus knows the exact chance of teams winning), then how the hell can you personally ever make a profit? There is no way to make any money at all with any betting theory if the odds were always precise. If what you say is correct, there is no point going for a higher frequency of winners either as the bookie will always pay out according to the exact implied chance. You claim to be long term profitable - are you personally just exceptionally lucky, the equivalent of a lottery winner? The betting markets are no different to any other liquid financial markets. Easy to participate in with a simple exterior, but inside all the wealth trickles to a very few participants. The reasons why most people lose in betting are the same reasons as the other markets - too many stakes, wrong sized stakes, having no idea of losing runs, undercapitalised, too high expectations, unwilling to do the real hard yards in research, reliance on "experts" instead of own experience, etc.

[protop]

Re: What percentages for different factors? The way I see this argument is like this. First of all it is impossible, to put an exact probability on the outcome of an event such as football as it is not quantifiable. It is just an opinion, odds makers offer opinions. If we new the exact probabilities, odds would remain static and they would all offer the same? As for the value concept, to come out ahead in this game, you don’t need value in your bets if you can find 100% winners, for the rest of us there has to be value in the bet. Value can be found in one of two ways; Trying to find value in each individual, outcome or by backing several selections that overall show a profit; value in there some where, but not sure where the value lies. The bookmakers basically operate the second principle, they don’t sweat on assuming they have priced every result to have a negative outcome, but know that overall they make a profit. Take this scenario; I toss a fair coin. On one outcome I pay 2.5 on the other outcome I pay 1.66. A 100% book. However you don’t know in advance what outcome pays what, but let’s say I write it down before each spin of the coin. Would you make a profit long term? No, but there is value each and every time on one of the outcomes

[Mulkis]

Re: What percentages for different factors?

If the bookmaker always has the correct odds (and thus knows the exact chance of teams winning), then how the hell can you personally ever make a profit? There is no way to make any money at all with any betting theory if the odds were always precise. If what you say is correct, there is no point going for a higher frequency of winners either as the bookie will always pay out according to the exact implied chance. You claim to be long term profitable - are you personally just exceptionally lucky, the equivalent of a lottery winner?

Right statements, wrong conclusions, DM99. When one "finds winners" instead of chasing "value", different set of rules comes into play - implied probability is no longer important, it is the strike rate and average odds that matters. And "finding winners" does not mean backing favorites! In fact it is often the opposite - laying favs. For better understanding here is an example of "winners" in football: Surprisingly, but after away wins some teams fail to perfom best in their next match at home. Not all teams, just some selected teams. (One could identify those teams by simple automated backtesting - it is easy when you know what to look for.) Probably there is something associated with team's psichological ability to prepare themselves for next game after away win. By laying such teams during last 2 and a half seasons one would get more or less stable ROI around 12%. In conjunction with "active betting" profit from this set-up is even better - more than 20%.

[Gingertipster]

Re: What percentages for different factors?

it is the strike rate and average odds that matters.

That is exactly what "value betting" is! :wall With level stakes a punter must achieve a better than 20% strike rate of his bets @ 4/1 to make a profit. Or to put it another way: To get VALUE, if a punter has a 20% strike rate he must achieve better average odds than 4/1. What you are saying Mulkis is: A punter does not need to chase value as long as he gets value. :eyes

[Mulkis]

Re: What percentages for different factors?

That is exactly what "value betting" is! :wall

No, it is not. And the main difference is how "value" punter and ... hm...lets call him "stupid" punter would approach the odds. "Value" punter would not bet at all if the odds are below certain level of his own calculated "true odds". Theoreticaly he is right, but in reality he will most likely miss "winners" on shorter odds and engage with "losers" on so called "true odds". "Stupid" punter in opposite would pay no or much less attention to odds and would bet on almost any odds whenever hi finds a potential "winner". Of course, the most important thing for "stupid" punter is to identify a "winner". It is not an easy task, but it looks more feasible, than looking for some mistic "true odds". One of the examples how to find "winner" I presented in previous post. But anyway, lets us stop this theoretical discussion and lets us better engage in some "wintersports" in the pub :cheers

[Gingertipster]

Re: What percentages for different factors?

No, it is not. "Value" punter would not bet at all if the odds are below certain level of his own calculated "true odds". Theoreticaly he is right, but in reality he will most likely miss "winners" on shorter odds and engage with "losers" on so called "true odds".

What you don't seem to realise Mulkis, is a bookmaker (or rather his odds compiler) works out "his own calculated true odds". Is he wasting his time too? May be he should just offer the same odds about every horse?

[Mulkis]

Re: What percentages for different factors? Bookmakers odds are quite accurate, quite "true" odds. I dont think it is worth an effort to compile your own even "more true" odds for 2-3% revard. It is a little monkey bussiness.