[allin1]

I was wondering if there is a way to determine the probability for total goals over/under on soccer matches. (1.5, 2.5, 3.5)

I am looking at soccerstats.com where they have the average probability of the league and also the average probability of over/under goals for the total goals scored in a game in which a team is playing considering results from current season.

For example I can have:

- in any match when team 1 has played there have been over 2.5 goals x% of the time
- in any match when team 2 has played there have been over 2.5 goals y% of the time
- the average probability of the league to have over 2.5goals in a match is z%

Can I calculate the probability of having over 2.5 goals in the game when 1 and 2 play together?

I was told to look at how tight is the distribution around that average. Has anybody every done this? Is there some site where I could see this?

Any other ideas on this issue?

[allin1]

Also I was wondering how do bookmakers determine the odds for over/under goals (0.5, 1.5, 2.5) if there is no goal in the first half...

Depending on the situation I guess it can be from -400 to +300 (1.25 - 4.0) but I can't understand how they are coming up with these numbers.

Usually if goals are scored in the ifrst half, it's more likely to have other goals scored because the game changes strategically, also some players could get tired, while substitutions can change the game. I was thinking at 0:0 at half time because it is a tricky situation.It really depends on who is playing and what the teams want from the match. If one will be satisfied with a draw it will keep a defensive approach, but if both try to win it it't more possible to see more goals.

But how do bookmakers determine the odds for in-play betting?

Is it possible to determine the probabilities for some scenarios in order to be prepared at half time to calculate the potential bet with kelly criterion?

[Pokerjoe]

Scoring is so rare in soccer, compared to other sports, that you're unlikely to develop a profitable calculation that depends upon them. Also, I believe (have not proven) that totals in soccer are more dependent upon the disparity in team strength than in other sports.

I'd even guess (have not proven) that if I had to choose between two pieces of info, and these two only--the game's AH, or the team's scoring averages for the year--I would choose, for making a totals bet, the game's AH. IOW, I think the game's AH, the difference between the team's strength, is a better indicator of the number of goals to be scored in a game than the team's past scoring history (I have done no work on this; if anyone has, chime in, I have no ego in this and won't mind being proven wrong).

As to "how do bookmakers determine the odds for in-play betting?", I think they make the line with DB crunching, adjusting the game line by score, cards, injuries, like that. I doubt they'd put much instinct into it, aside from injury appraisal (and may have DBs big enough to confidently put a number on those, too). IOW, it's a math question. And I'd say that 0:0 at the half is probably not tricky at all. 4:0 at the half, I'd think, would be trickier, because the DB on it would be so much smaller.

DB = database, AH = asian handicap.

You said, "I can't understand how they are coming up with these numbers." Until you do, and can come up with them yourself, don't bet. But you aren't really going to get good answers by asking, you're only really going to get them by deducing what must be (which is all I've done here, btw; I've never sat in a book's war room as they generate lines. Then again, only a few books generate lines anymore anyway.)

Books are neither brilliant nor stupid, they have no magic, most just clone the generators, and the generators mostly just number crunch and small-opin/lean an opening number and let the early bettors smooth it out from there.

[podonne]

I was wondering if there is a way to determine the probability for total goals over/under on soccer matches. (1.5, 2.5, 3.5)

I am looking at soccerstats.com where they have the average probability of the league and also the average probability of over/under goals for the total goals scored in a game in which a team is playing considering results from current season.

For example I can have:

- in any match when team 1 has played there have been over 2.5 goals x% of the time
- in any match when team 2 has played there have been over 2.5 goals y% of the time
- the average probability of the league to have over 2.5goals in a match is z%

Can I calculate the probability of having over 2.5 goals in the game when 1 and 2 play together?

I was told to look at how tight is the distribution around that average. Has anybody every done this? Is there some site where I could see this?

Any other ideas on this issue?

Short answer is no, these numbers won't simply tell you the probability of having over 2.5 goals in the game. You'll need to have a record of the times that these two teams play each other. For illustration, look at the LSU 'Bama game. Every game those two teams played -not against other- was high scoring, they play each other, 15 points.

Longer answer does involve the distribution of totals. When someone told you "look at how tight the distribution was", they are telling you to look at the total from every game, and calculate the mean and standard deviation of the totals. The smaller the standard deviation is, the "tighter" the distribution is. Look up "normal distribution" on wikipedia for more info.

[allin1]

Thank you for your answers.

So in conclusion I could manipulate the stats I mentioned in any possible mathematical way but I wouldn't get close to a probability that is near accurate.

I know what the distribution is and maybe it could be calculated entering the stats in excel and working them out, but my question was if this could be found already done somewhere on the internet.

[Ruifgalmeida]

Must live-in-play U/O odds are computator generated, the Trader only adjusts when there are a send off or weather condicion etc... as for the formula it is a secret