Historically football punditry has been the domain of talking heads on TV. Most of these pundits consist of former players and managers. They will often try to use their back channels for obtaining “inside information”. This information, together with their football experiences, will use to formulate their opinions on various topics of the game. Most of their analysis revolves around players that have been signed or sold and the soundness of the various clubs’ approaches or recently played or upcoming matches.
As most news media generate revenue by creating attention, it makes sense for them to attain mostly uncompromising characters, expressing brash viewpoints. This is because controversial opinions, expressed in uncompromising terms, tend to create more buzz, than anodyne statements articulated in a reserved tone.
However, as you likely are well aware already, it is almost always better to employ a probabilistic approach when trying to analyze any sport for betting purposes.
Luckily nowadays the spectrum of analysis has widened. In addition to the usual cast of ex-players and ex-managers, a small industry of quantitative analysis has emerged.
The starting point for making a good analysis is to start with valuable information. The most valuable information about injuries and suspensions is available right here.
A good example of a website working with football data not revolving around injuries and suspensions is Statsbomb.
Other sites will use data from Statsbomb or data from other suppliers to feed their models. These models will predict the results of future matches and competitions. There are several such models available on free websites, the problem is that such publically available models will in most cases not be able to outperform the market and produce a profit.
For smaller sports, there are some indications that some models available in the public domain are capable of beating the betting markets. This is likely because the odds for less turnover-intensive sports are not as accurate as the odds for mainstream sports like football.
Even if it is difficult to follow a publically available model and make a profit. There is a lot of good content available, where a data-driven approach has been employed to gain insights about a topic. A good example of such a site could be Godsofodds.
The current trend is that more and more sports are being modelled. Interestingly even somewhat obscure sports have been modelled with publically available projections free of charge. A good example of this would be Sportindepth’s projected results of the biathlon world cup.
I cannot recommend that you use data from any freely available model, to quantify the likelihood of the outcomes of the different sports, for betting purposes. However, if you want to give this a try, I would definitely recommend that you go with a model working with smaller betting sports.
The best way forward would in my opinion be to team up with someone to build your own model. If you, like most people, do not have the resources to do that, my best advice would be to use your knowhow of a sport together with a probabilistic mindset and data, provided by sources like injuries and suspensions, to try to weigh up the chances of the different outcomes and find value bets.
Author: Mathis Brorstad
Subscribe to see the most useful team news from each league.
Like this? Like us on FACEBOOK and follow us on TWITTER and INSTAGRAM!
Explanation: player statistics (a/b c)
a - games played this season
b - goals scored
c - team position (d - defender, m - midfielder, f - forward, g - goalkeeper).
