You don’t have to look too far on the web to find football stats, some good, many pointless and usually topping the list is that of previous form (usually a team's last 5 games). In my experience you would probably be better of playing roulette than trying to make decisions using only 5 previous games but sadly this counts as “statistical analysis” for the many.
Go on any sports website or even to the bookies themselves and they will happily show you the last 5 results for a team before placing you bet – the point been it probably doesn’t mean much on its own. Even if a team is on a good run this will usually be noticed and priced into the odds accordingly.
In order to make money in the long term we need to find an ongoing stream of good value bets! With this in mind I have been looking at how using recent form might help us identify a statistical edge based on nothing except a teams recent previous results.
So, with all this in mind I have fired up my football stats database and have been looking at the last 10 years of UK and European results to see if previous form can help us gain an edge?
NB – Im my charts and tables below I use abbreviated league names, this table below is a key for those abbreviations:
Also note when I refer to a “6th game” this is a game played by a team which has won or lost the previous 5 consecutive games (excluding international tournaments).
Scenario 1 – How well do teams with 5 previous wins perform in their 6th game?
Here I have used the last 10 years of data from the main UK and European leagues to see how teams will, on average, perform in their 6th game following 5 consecutive wins. In other words how useful is it knowing the team is on a small winning streak and could this potentially make for a profitable strategy over the long term?
Only 123 teams out of a possible 408 have played a 6th game following 5 consecutive wins during this period! In the table above I have listed these teams along with the number of games played on the back of 5 wins (Total_5W_games), the number of these games won (Wins) and the percentage in the final column. What’s great about this table it shows how the top teams in Europe have not only played many more games coming off 5 straight wins but they also maintain a very strong win rate with the top 20 teams all maintaining a 69%+ win ratio!
These percentages are certainly interesting but what about profitability if I had been blindly betting on these teams during such games? Below I am going to investigate what would have happened had I placed a £1 bet on each of these “6th games” for the 123 teams above? I will be using odds from Bet365 so bear in mind their “profit margin” will have been priced in meaning the odds will be slightly lower than the true odds!
As can be seen from the chart above had we bet £1 on each of the 1668 games we would have made a loss of £96.52 over the 10 year period, this is approx -5.8% of the £1668 bet!
Scenario 2 – How Well do Teams with 5 Previous Loses Perform in Their 6th Game?
Here I am looking at the same ten years of data to see how teams with 5 consecutive losses perform in their 6th game.
In the table above we have 177 teams in this position compared to 123 teams with 5 wins in the first scenario, I have ordered these teams by the number of games played. Since a lot of these have only played 1 game the win percentage in such cases will be either 0% or 100% which is not of much, as such these teams are ordered down towards the bottom of the table.
What I find very interesting is by looking at the top 20 rows in this table the teams with the most loosing streaks generally go on to perform very poorly in their 6th game. Also note how the highest win rate of such games is 40% with quite a few teams winning much less than 1 in 10 games! A definite contrast with the teams coming of 5 consecutive wins!
Again I am going to see how I would have fared betting £1 with Bet365 on each of these “6th games” for each of the 177 teams above:
As can be seen above just like backing teams on a winning streak this also produces another very clear long term loss! Betting £1 on each of the 1045 games involved would have delivered a loss of -£144.56 equating to a percentage return of -13.8%, ouch!
Scenario 3 – Looking at performance in 6th games on a league-by-league basis
Here I am going to group all of these “6th games” by league to see how different leagues compare with one another:
As can be seen in the table above some leagues including the Spanish La Liga, the Italia Serie A and the Portuguese Primeira Liga vastly outnumber certain leagues such as the French second division and the Bundesligua second division! Below I have created a chart to highlight this difference between leagues:
In the chart above I have picked 5 of the main footballing European countries which have a strong second league (England, France, Italy, Germany and Spain). The most striking thing for me is how in every instance the first league strongly outnumbers the second division.
Below is the same data as used in the chart above but showing the win rate for each league as a percentage. This highlights how different leagues and specifically the first and second divisions perform against each other.
Finally what if I had bet £1 on all “6th games” in each of these leagues over the 10 year period, what would my return look like on a league-by-league basis:
As the chart above clearly shows the majority of leagues register a loss with the noticeable exception of the Italian Series A which delivers a respectable 7% profit!
Scenario 4 – Average no. of goals in a game following 5 wins vs those games without
Finally I am going to look at the number of goals in these “6th games” versus games played by teams without 5 previous wins. As we have seen above teams playing a "6th game" generally win more of the time which must therefore translate to more goals been scored in such matches!
As the chart above confirms in almost every league this holds true and in every first division league this is the case 100% of the time!
I have always maintained looking at short term form is meaningless when sports betting and hopefully the scenarios above help to prove this is in fact the case. Simply backing a team based soley on good (or bad) recent form will deliver only losses in the long run.
What the data above does show however is that we can look at recent form from a different angle and extrapolate that certain leagues out-perform others over the long term in certain metrics such as goals scored. Whilst such a discovery might not deliver a profitable strategy by itself it can be combined with other analysis to help in identifying mispriced markets, I find stats like this specifically useful in the over/under markets myself!
For more information on using football stats please see my article on supercharging your football stats