• Quick note - the problem with Youtube videos not embedding on the forum appears to have been fixed, thanks to ZiprHead. If you do still see problems let me know.

Help needed with Excel.

Art Vandelay said:
II seems to be saying that you can.
Click to expand...

Yes that's right. With the more poular matches you can bet right up to just before the whistle is blown for the end of the match. At least you can for the who will win market (team A, team B or draw), total goals market, correct scores market and who will score next goal market (maybe other markets too, can't remember now).
 
Interesting Ian said:
Yes that's right. With the more poular matches you can bet right up to just before the whistle is blown for the end of the match. At least you can for the who will win market (team A, team B or draw), total goals market, correct scores market and who will score next goal market (maybe other markets too, can't remember now).
Click to expand...
However we were discussing the half-time/full-time bets for which I imagine this is not so.
 
Oh well, I reckon with a thread like this you had better put your money on the line. (Of course this isn't really money as I expect to lose most of it)

Actually most of the odds I calculate are just about the same as the betting agency, but these are the exceptions:
Code: 
Soccer - English Premier League
Date       Fixture                        Category              Odds Amount
2006-05-07 Aston Villa - Sunderland       Full time draw        3.65   $10
2006-05-07 Aston Villa - Sunderland       Half time draw         2.5   $10
2006-05-07 Aston Villa - Sunderland       Draw/Draw              6.5   $10
2006-05-05 Manchester City - Arsenal      Draw/Win                10   $10
2006-05-05 Sunderland - Fulham            Lose/Lose                4   $10
2006-05-07 Blackburn - Manchester City    Full time lose        4.85   $10
2006-05-07 Blackburn - Manchester City    Half time lose         4.9   $10
2006-05-07 Blackburn - Manchester City    Lose/Lose                9   $10
2006-05-07 Bolton - Birmingham            Full time lose        5.55   $10
2006-05-07 Bolton - Birmingham            Half time lose         5.2   $10
2006-05-07 Bolton - Birmingham            Lose/Lose               10   $10
2006-05-07 Everton - West Brom            Full time lose        5.55   $10
2006-05-07 Everton - West Brom            Half time lose         5.2   $10
2006-05-07 Everton - West Brom            Lose/Lose               10   $10
2006-05-07 Fulham - Middlesbrough         Full time lose        4.65   $10
2006-05-07 Fulham - Middlesbrough         Half time lose         4.8   $10
2006-05-07 Fulham - Middlesbrough         Lose/Lose              8.5   $10
2006-05-07 Manchester United - Charlton   Full time lose           9   $10
2006-05-07 Manchester United - Charlton   Half time lose         7.5   $10
2006-05-07 Manchester United - Charlton   Lose/Lose               21   $10
2006-05-07 Portsmouth - Liverpool         Win                   4.75   $10
2006-05-07 West Ham - Tottenham           Win/Win                  7   $10
 
Interesting Post Ian, but I fear your going to disappear into a wood of maths and not see the betting tree. There are several stumbling blocks on your road to riches, so lets get rid of the mathmatical theory ones first. Your base of 2.59 goals per game will leave your model wanting in some cases if not most, maybe only a couple of ticks on the price but enough to change profit into loss. By taking your base score prediction of 2.2 goals from another source you are effectively putting your money in their hands.

So on to other pastures. Betting markets arent efficient because they are built on opinions, and nobody's can be right all the time. Your Mathmatical Theory is an opinion in reality as soon as you put your money down, would I favour that over someone who knows his Soccer inside out? or a Person with up to the minute information on Team selection and Injury? Then you have the Money that dictates the market regardless of the Maths and Probability, The Big Punter who has made this Match one of his biggest bets of the Year, the weight of Money on a popular opinion contrary to the theory. Then there is the subtle things that has to be taken into account when trading, holding a winning position against a Favourite and trying to bale out at the right theoretical price but not being able to because the losing people are holding on for dear life.

What can be done then. Firstly think of the Maths as a safety net but you need to broaden its parameters to take in weighted money, trade at the top end of those markets and pass if you are not accomodated, unless you have been watching and get a feel for where its going. Always be aware that weak markets can be your friend for being laid but hell to hedge out of.

It may be worth starting a World Cup Thread in the Sports Forum specifically for Betting as it will have its oddities. Hope I've helped. :cool:
 
Waster said:
Interesting Post Ian, but I fear your going to disappear into a wood of maths and not see the betting tree. There are several stumbling blocks on your road to riches, so lets get rid of the mathmatical theory ones first. Your base of 2.59 goals per game will leave your model wanting in some cases if not most, maybe only a couple of ticks on the price but enough to change profit into loss. By taking your base score prediction of 2.2 goals from another source you are effectively putting your money in their hands.
Click to expand...

Eh?? I have no idea what you are talking about. The average number of goals per game is not important, it's the expected average for any particular game.

So on to other pastures. Betting markets arent efficient because they are built on opinions, and nobody's can be right all the time. Your Mathmatical Theory is an opinion in reality as soon as you put your money down, would I favour that over someone who knows his Soccer inside out? or a Person with up to the minute information on Team selection and Injury?
Click to expand...

But I am taking goal total, and goal supremacy, and the expected result in percentage terms (1x2) (from which I work out probs for correct scores and total goals) from the exchanges themselves. Obviously these figures will factor in injuries and the like.

Then you have the Money that dictates the market regardless of the Maths and Probability, The Big Punter who has made this Match one of his biggest bets of the Year, the weight of Money on a popular opinion contrary to the theory. Then there is the subtle things that has to be taken into account when trading, holding a winning position against a Favourite and trying to bale out at the right theoretical price but not being able to because the losing people are holding on for dear life.
Click to expand...

{shrugs} If I can't trade out then so be it. Providing I have a value bet then that's fine (in the long run).


It may be worth starting a World Cup Thread in the Sports Forum specifically for Betting as it will have its oddities. Hope I've helped. :cool:
Click to expand...

Yes that would be a good idea if other people are interested.
 
Interesting Ian said:
How did you calculate them?
Click to expand...
Probably not an interesting question any more in view of the result (not a sausage), but I will post the spreadsheet sometime today so that you can see.

Actually I think the model (which is adapted from an earlier suggestion of yours) is reasonably sound - but I just don't have good goal expectations to go on.
 
Robin said:
Probably not an interesting question any more in view of the result (not a sausage), but I will post the spreadsheet sometime today so that you can see.

Actually I think the model (which is adapted from an earlier suggestion of yours) is reasonably sound - but I just don't have good goal expectations to go on.
Click to expand...

You can get goal expectations both from spreadfair and betfair. People on spreadfair bet directly on total goals and superiority which means the average expected score can be derived very easily. On Betfair people bet on the outcome of the match (which Team will win or whether it will be a draw), and they bet on whether or not there will be more or less than 3 goals. From the betting on the over/under 2.5 goals market you can derive average expected total goals. From the match outcome market in combination with total goals it is possible to derive goal superiority. We also know that 44% of goals on average are scored in first half and 56% in 2nd half.

Ideally we need to take a look at many thousands of match results including the scores at half time, the quoted odds for the half-time/full-time before each of these matches as well as the expected total goals and superiority before each of these matches. Then we could see if the calculated odds would have generated a profit. But it might be rather difficult to get hold of these stats.

Do you have this spreadsheet then? I don't think you can post it on here can you? If you'd like to email it to me so I can take a look at it that would be great. I have a gmail address. Just stick interesting.ian in front of the @ part.
 
Interesting Ian said:
You can get goal expectations both from spreadfair and betfair. People on spreadfair bet directly on total goals and superiority which means the average expected score can be derived very easily. On Betfair people bet on the outcome of the match (which Team will win or whether it will be a draw), and they bet on whether or not there will be more or less than 3 goals. From the betting on the over/under 2.5 goals market you can derive average expected total goals. From the match outcome market in combination with total goals it is possible to derive goal superiority. We also know that 44% of goals on average are scored in first half and 56% in 2nd half.

Ideally we need to take a look at many thousands of match results including the scores at half time, the quoted odds for the half-time/full-time before each of these matches as well as the expected total goals and superiority before each of these matches. Then we could see if the calculated odds would have generated a profit. But it might be rather difficult to get hold of these stats.

Do you have this spreadsheet then? I don't think you can post it on here can you? If you'd like to email it to me so I can take a look at it that would be great. I have a gmail address. Just stick interesting.ian in front of the @ part.
Click to expand...

Are all of the world cup markets for the 90 minute only result or do they include the possibility of extra time?

Some interesting statistical analysis of previous world cups here:

http://www.planetworldcup.com/GUESTS/peter20020202.html

which suggest that goal expectation will be higher for knockout rather than group games.

A very quick look at the knockout games in the last world cup showed no bias between first half and second half goals (ignoring injury time entirely) - 12 first half, 11 second. Not sure if the type of match (knockout as opposed to league) would have an impact on the proportion first and second half as that is obviously a very small sample.
 
Jaggy Bunnet said:
Are all of the world cup markets for the 90 minute only result or do they include the possibility of extra time?
Click to expand...

There's always the injury time at the end of each half which typically tends to be around 1 to 4 mins. When it says 90 mins it includes this injury time. The odds quoted are for these "90" minutes (90 in quotes as that includes injury time).

Some interesting statistical analysis of previous world cups here:

http://www.planetworldcup.com/GUESTS/peter20020202.html

which suggest that goal expectation will be higher for knockout rather than group games.

A very quick look at the knockout games in the last world cup showed no bias between first half and second half goals (ignoring injury time entirely) - 12 first half, 11 second. Not sure if the type of match (knockout as opposed to league) would have an impact on the proportion first and second half as that is obviously a very small sample.
Click to expand...

Yes, the 44:56 ratio is from many thousands of matches. I really have no idea if this ratio might be different for the world cup. There doesn't seem to me to be any partciular reason why it should be. But I invite peoples' comments!

I'll take a look at the link.
 
Interesting Ian said:
There's always the injury time at the end of each half which typically tends to be around 1 to 4 mins. When it says 90 mins it includes this injury time. The odds quoted are for these "90" minutes (90 in quotes as that includes injury time).



Yes, the 44:56 ratio is from many thousands of matches. I really have no idea if this ratio might be different for the world cup. There doesn't seem to me to be any partciular reason why it should be. But I invite peoples' comments!

I'll take a look at the link.
Click to expand...

I can think of several possible reasons why it could be different, just not sure if it actually is or not!

For example it may be that knockout games result in more defensive football early in the game (losing the first goal is arguably more negative in a knockout match than a league match as there is less opportunity to recover) which would suggest a lower percantage in the first half.

Alternatively, the fact that the teams will be less familiar with each other than in a domestic league competition might mean that early in the game they take fewer chances as they "test out" the opposition.

However I would also have expected less goals per game in the knockout stage rather than the group stages which is not borne out by the stats in the article so what do I know?
 
Jaggy Bunnet said:
Are all of the world cup markets for the 90 minute only result or do they include the possibility of extra time?

Some interesting statistical analysis of previous world cups here:

http://www.planetworldcup.com/GUESTS/peter20020202.html

which suggest that goal expectation will be higher for knockout rather than group games.
Click to expand...
I disagree about the significance, though. The group games have an sd of .33. The knockout games have an sd of .79. That's an sd of .85 for their difference. So a difference of .5 really isn't that big.
 
Art Vandelay said:
I disagree about the significance, though. The group games have an sd of .33. The knockout games have an sd of .79. That's an sd of .85 for their difference. So a difference of .5 really isn't that big.
Click to expand...

I'm afraid I am no statistician so cannot make any sort of informed comment. The thing that struck me was that the expectation was higher when I would have expected it to be lower due to there being fewer mismatches (like the Germany 8-0 Saudi Arabia game in the group stages of WC2002) and also no reward for winning by a bigger margin (whereas groups could be settled by goal difference).

It may be that if the analysis was done for the last world cup the results would be different as there have arguably been more weak teams admitted as the places available to teams from outside Europe & South America have increased.
 
You must log in or register to reply here.

Back
Top Bottom