Predicting Football Matches

[Graham Kendall]

Hi all

(New member)

I have just started a blog that really talks about my various research interests (see http://research-reflections.blogspot.com).

My latest post (http://research-reflections.blogspot.com/2009/07/predicting-results-of-football-matches.html) has started to discuss how to predict the outcome of football matches. I have some ideas already (e.g. use the shortest odds, select the home team, look at previous encounters etc.), but I wonder if Sports Exchange members have any other rules of thumb they use?

Cheers

Graham

[Graham Kendall]

Anton: Thank you. I read your Number9Systems blog. Very informative. I need to work through the Man U/Arsenal example but from an initial read it looks promising.

[Richard Stoddart]

Analysing each game on its merits and circumstances. A colleague of mine has always opted to side with a result he thinks won't happen and lay it, then he has 2 results running for him rather than the one if he had of backed it.

[Anton Whalley]

Hi Graham

Welcome to the sports exchange.

I agree with Ron on this topic though for me there is only one way to back for profit and that is on Price.
i.e. Make sure that you are getting over the odds for your bet so over time you will profit.

This can be done in a number of ways but the Price is always king!!

I posted an example of how you might do this the other month on my blog.
(Just reading over it now its very badly written but you should get the idea)
http://number9systems.blogspot.com/2009/05/view-on-champions-league...
and we had another example of this the other day on this site when Richard posted
http://testpunt.ning.com/profiles/blogs/quick-get-on-notts-co-for

In the latter example richard was event driven which I am not totally happy with but once I saw the price on BF was 9 and I looked it up on Betdaq and it was 11 so I took it on BetDaq and got off at between 8 and 6.
(Roll on the end of season now!)

For me the problem with pattern based approaches based on historical data is that it is very hard to tell if you are in an elaborate gamblers fallacy or not.
I guess its the same reason financial institutions have the clause "Past performance is not an indication of future performance" at the end of every advert. (But that doesn't stop people giving them money! much the same way as "Smoking causes brain tumors" doesn't stop me smoking).

I hope you enjoy your travels in this area as much as I do.

[Hristo]

I use a sort of a "pattern recognition" system i kind of developed my self and cross-check it with an estimate of the teams current form and sometimes that gives startlingly good predictions. The thing is from the angle i am looking, most of the football games in a certain championship are unpredictable. So i ended up having fewer and fewer bets.

I am though very interested to hear your opinions on the "current form" thingy. Do you calculate it somehow or just look at the stats and "feel" your way to a decision ? I do bit of both.
One pretty good sign for me is winning to zero with at least a two goals margin but not with too many goals: 2-0, 3-0 against middle to top side opponents to me is a sign of a pretty good form (mind you, goals should not come from penalty kicks) so depending on the opposition, i am pretty confident for the team's next game outcome. On the other hand, 1-0 is not good enough and 4+ - 0 kind of "burns the enthusiasm out" for the next match more often than not.
This, however, does not apply for some amateurish leagues as sometimes there are huge gaps between teams there.

On the other hand, a sign of pretty bad form is losing at home to zero with a 2+ goals margin. Most often if next game is away, it is a loss.

So do you guys have similar ideas ?

[Ron]

The only trouble with statistics is that bookmakers also have access to them. I prefer to let them do the legwork. Anyone who can predict the outcome of a match or any other event more accurately than the bookmakers is welcome on my friends list.

[Gordon McIvor Wilson]

Wow, the things you guys get up to! Fascinating, youll find it here on The Sports Exchange.

[john Tuohy]

On the other hand you could just back QPR to get beat every week. :-).

[Graham Kendall]

Mick

Just had a closer look and this is a VERY useful resource. Thank you. I owe you one.

Graham

[Graham Kendall]

Cheers Mick

I have come across this web site before though, to be honest, I had forgotten all about it (my use bookmarks more often!), so thank you.

I actually have all the results from last season (and just for fun I also recorded all the yellow and red cards that were handed out). This information was taken from the BBC sports pages.

But through your link, it seems that I can get 8 seasons of betting odds (which is nice) - though I need to track it down. Last season, I visited the bookies every week to get their fixed odds coupons. This coming season, I think I will subscribe to The Racing Post; at least on Saturdays.

[Mick Stone]

Are you familiar with the data available free of charge here :- http://www.football-data.co.uk/data.php ? I note from your blog that your data is incomplete and football-data will likely fill some of the gaps as well as providing a broader sample.