In this post, I will explain how to analyse a football team’s last 6 matches.
Many people just scroll down the list of the last 6 matches and say to themselves something like:
“Arsenal” should have won that match because they were odds-on favourites” or
“Everton did well in that match because they were the outsider and they drew”.
However, we can do much better than that.
In this article, I will explain how I assess recent form. There is nothing difficult in this article. You are probably already familiar with the information that I use. However, what may be new, is how I put the pieces of information together to reach a conclusion.
So, let’s get started.
A. Football is Not Just About Skill – It is Also About Luck
It’s important to remember that odds reflect probabilities and not certainties. Each team has a probability of winning, drawing and losing.
The first topic addresses the question, “Has the team, that we are researching, been over-achieving, under-achieving, or performing as expected in terms of results?” To do this, we will work out the number of points a team is expected to win from the last 6 matches. This is different from just looking at whether the team should have won, drawn or lost individual matches.
We always need to remember that there is a large element of luck in football. This is why the bookmakers present odds. Each team has a probability of winning, drawing and losing.
The second topic addresses the question, “Has the team been lucky or unlucky or has their luck been just about right?”
In this analysis, I will look at home form for the home team and away form for the away team.
B. Step 1: Determine Whether a Team Been Over-Achieving, Under-Achieving, or Performing as Expected
So, I’ll use a fictitious match and put Everton at home and Arsenal away into the Poisson calculator. I won’t go through the whole match analysis. I showed how I would go about the first part of the analysis that in the last video in the Brighton – Liverpool example.
When you put your teams in the Poisson Calculator, 4 tabs will appear at the top of the calculator. These tabs are:
Stats (1yr) – 1 year statistics
Last 6 games (H vs A) – Last 6 Home games for the home team and last 6 away games for the away team
Last 6 games (All) – The last 6 matches, regardless of whether the matches were home or away matches
Stats (Last 6) – Statistics for the last 6 matches.
We are going to use 3 of these for our analysis.
So, I will start by looking at the last 6 matches home vs away.
So, this shows the last 6 matches. I will explain what everything means.
The last 6 match ratings at the top are the most important numbers. If you were in a rush, you could just look at these numbers.
1. Last 6 Match Home or Away Rating and Last 6 Match Away Rating
The ratings at the top reflect how many points the team has gained compared to how many points the team was expected to get from the last 6 matches. Therefore, if a team won exactly the number of points, that it was expected to get, the percentage would be 100%.
If a team has a rating greater than 100%, like Everton, in this example, the team has over-achieved. By contrast, if a team has a rating of less than 100%, like Arsenal in this example, the team has under-achieved.
You will see that Everton has a Last 6 Match Home Rating of 159.04%.
This means that Everton have won 59.04% more points from their last 6 matches than expected. In other words, Everton has been over-achieving by 59.04% in their last 6 matches.
Arsenal have a Last 6 Match Away Rating is 65.55%.
This means that Arsenal have won only 65.55% of the points that would be expected in their last 6 matches.
I won’t go deeply into the maths because some people don’t like maths. The article on my website explains the maths. The link is under the video. If you are a maths person, you’’ll know how to do the calculations anyway.
You don’t really need to understand where the numbers come from. However, I will explain where the numbers come from so, that you know that these numbers have a logical basis.
2. Probability
On the right of the grey box with “Prob” inside, there is a list of probabilities for the Home Win, Draw and Away win.
The historical football data, that I get, contains average bookmakers’ odds.
I convert these odds into percentages and then, in order to remove the effects of the bookies overround, I scale the percentages to a total of 100.
These odds can be converted into percentages in the following way.
You divide 100 by each of the odds. So, let’s say that you have odds of:
Home: 1.94
Draw: 3.48
Away: 4.12
You would end up with the following percentages:
Home: 100 ÷ 1.94 = 51.55%
Draw: 100 ÷ 3.48 = 28.74%
Away: 100 ÷ 4.12 = 24.27%
If we add these percentages up, we get 104.56%.
They add up to more than 100% due to the bookmakers’ overround.
So, to turn them into probabilities, that are scaled to a total of 100%, you just multiply the above probabilities by 100 and divide by 104.56.
Home: (100 x 51.55) ÷ 104.56 = 49.30%
Draw: (100 x 28.74) ÷ 104.56 = 27.49%
Away: (100 x 24.27) ÷ 104.56 = 23.21%
This adds up to 100%. These are close to the same probabilities that Everton had against Crystal Palace. The numbers aren’t exactly the same because, when I calculate this manually, I round every step of the calculation to 2 decimal points. Obviously, the calculator doesn’t do this.
3. XPts/Game (Expected Points Per Game)
XPts/Game reflects the theoretical average number of points that Everton would get, if the match against Crystal Palace was played an infinite number of times.
This is a simple calculation. We multiply the probability of the home win by 3 pts and the probability of a draw by 1 pt and add these 2 numbers together.
XPts = (Probability Home Win x 3 pts) + (Probability Draw x 1 pt)
So, instead of saying that Everton should win this match because they are favourites, we are saying that, Everton has a 49.24% chance of getting 3 pts and a 27.67% chance of getting 1 pt. We don’t have to do anything about the away win because this would be zero points and we would be multiplying by zero.
So, (49.24% x 3) + (27.67% x1) = 1.75 pts
We would do the same procedure for all 6 matches.
For Arsenal, we would do a similar calculation. XPts would be the (Probability of an Away Win x 3 pts) + (Probability of a Draw x 1 pt)
4. Pts (Actual Number of Points Won)
This refers to the actual number of points that Everton got from each game. So, because Everton won against Crystal Palace, they got 3 pts.
5. The Last 6 Match Home Rating
To calculate the Last 6 match Home Rating, the calculator adds up the points that Everton obtained in the 6 matches and divide this by the total number of expected points. The result of this is multiplied by 100.
So, the formula is:
(Total Pts/Total XPts) x 100
To calculate the away team’s rating, we follow exactly the same procedure, except we would be calculating XPts from the away team’s perspective.
So, that is the most objective way, that I can think of, to work out how well a team has performed over the last 6 games.
As I have said, you don’t need to do any of this maths. You can just look at the last 6 match ratings at the top of the calculator.
C. Step 2: Figuring Out the Luck Factor
So, we know from the above calculations that, based on points gained in the last 6 games, Everton has over-achieved, while Arsenal has under-achieved. However, the analysis doesn’t stop there.
There is a luck factor in football. This is one of the reasons that we can bet on it.
The questions are:
Have Everton just been lucky or have they been playing well?
Have Arsenal been unlucky or have they been playing badly?
To answer these questions, we need to look at the statistics.
1. Shots on Target and Luck
So, we click on the 6 match statistics link to bring up the stats for the last 6 matches.
With all my statistics, I like to set them out so you can see how each statistical variable influences the win rate of teams. So, for shots on target, you’ll see the average shots on target for all matches. Then, you’ll see the average shots on target when the team wins; the average when they draw and the average when they lose.
I’ve listed the standard statistics. However, in my opinion, there is only one statistic variable, that has power as a standalone variable. That is shots on target both for and against the team.
That’s because, when you look at the long-term stats of the better, or what you might call teams that emphasise possession, you can see a clear pattern of more shots on target when they win, compared to when they draw or lose. In addition, you will often see less shots on target faced when they win, compared to when they draw or lose.
You don’t see such a clear pattern with any other variable.
2. Percentages
So, before I explain the trick to analysing short-term statistics, I want to emphasise that every play in football has a percentage chance of success. It’s not just shots on target. For example, when you see a winger cross a ball perfectly to the feet of a striker, it might look amazing. However, the chances are that the winner will get the cross perfect a small percentage of the time. He will put in an adequate cross a larger percentage of the time and he will completely mess it up a percentage of the time. Obviously, the better wingers have better percentages. This idea of percentages in football works for normal passes, tackles and saves by goalkeepers.
However, a team can only score goals if it is getting shots on target.
Conversion Percentages of Shots on Target to Goals from the 2021/22 Premier League
In the video, I show the conversion rate of shots on target to goals from last season. The lists are ranked by their conversion percentage. So, Man City (at home) have a conversion percentage of 43.61%. They got around 7 shots on target/home match on average last season. Although that’s the highest, relative to any other team, it is a small number in absolute terms. This means that there can be a variability in shots on target being converted to goals in an individual match.
In addition, 43.61% is worse than the odds of a coin-flip. Even with the highest percentage in the league, they can expect to have the occasional bad run. You won’t see many bad runs with Man City in the match result market, because they are also good at stopping other teams from getting shots on target.
When you look at teams, that are below Man City and Liverpool and possibly, Chelsea, you will find more variability.
When a team’s is close to 30%, the odds are comparable to backgammon, when you are on the bar against a 5 pt board. Anyone, who has played backgammon will know that you can be rolling for a long time before hitting your number.
So, that how I view shots on target. If a team is having a bad run, while still getting their shots on target, they may not be playing badly because they are still getting to roll the dice. If a team isn’t getting many shots on target, they are not getting to roll the dice often and if a team isn’t rolling the dice, they can’t win.
3. The Secret to Analysing Short-Term Statistics
So, before I go through the luck analysis, I want to explain that I would never bet on the basis of 6 matches. Before looking at the last 6 matches, I would have done the long-term analysis. That would include looking at Poisson results and data from previous times that these 2 teams have met.
Let say that the long-term analysis was saying that Arsenal was a value bet. If the 6 match analysis said that Arsenal is a bad bet, I wouldn’t get involved in the match. However, if the 6 match analysis supported the bet or was even neutral, I would be happy to bet on the basis of the long-term analysis.
If you look at a list of statistics, it’s difficult to know whether the numbers are big or small. The secret to looking at short-term statistics, is to compare each team with itself over different time periods. So, you compare the team’s short-term statistics against its long-term statistics.
The long-term statistics don’t have to be a year’s worth of statistics. However, I am going to use a year here.
So, I will click on the tab labelled Stats (1 Yr).
So, I’ve got Everton and Arsenal’s 6 match and 1 year stats side by side. I’ll look at Everton first. So, there is no difference between the average shots on target for between the 2 datasets. If you look at shots on target faced, you will see that Everton have faced slightly more shots on target in the 6 match statistics, compared to the 1 year statistics.
Therefore, we can’t say that Everton has shown improvement.
If you look at Arsenal, they’ve had more or less the same average number of shots on target in the 6 match and the 1 year stats. However, they have faced less shots on target in the 6 match analysis, compared to the 1 year analysis. This suggests that Arsenal have been playing better and not worse. Even if you don’t want to go so far as to say Arsenal have improved, it can’t be argued that they have played worse than normal.
There are 2 things that you would have to check to reach a conclusion. Firstly, the fact that Arsenal have been facing less shots on target, but have been losing matches, could be because their goalkeeper is off form. Goalkeepers don’t usually lose their form. However, you would need to check that they are haven’t been playing their second best goalkeeper. For the premier league, you can find team sheets for previous matches on the BBC website.
Secondly, you have to check whether teams have been playing a representative sample of teams in the last 6 match analysis. For example, if a team has only played strong teams or all weak teams, you couldn’t compare 6 match stats with 1 years stats. This is because strong teams are better at getting shots on target and stopping other teams from getting shots on target, while weak teams are the opposite. Therefore, a 6 match sample, in which the opposition is at the extremes of strong or weak, should not be compared to a 1 year sample.
So far, from the analysis, I would say that Everton have been lucky, while Arsenal may have been unlucky. This is where you might find value. When other punters are making decisions solely on short-term results and you are doing a deeper analysis, you might find value opportunities when your opinion is the opposite to the majority.
These are tentative conclusions. However, there is one more thing we can do to help us to decide whether Arsenal have been unlucky. Let’s click on the link that says Last 6 (All)
When I started making this video, I didn’t know that Arsenal had played Everton at home in their last match. I just chose a random teams, until I found a pair with one that was above average on points and one that was below average.
The fact that the most recent match doesn’t make any difference to the idea that I’m putting across.
So, these are the 6 most recent matches for each team, regardless of whether the matches are home or away. I’m more interested in Arsenal’s overall performance here. They are over 100% on their last 6 match rating. This suggests that they have not lost their form. It’s unlikely that a team would only deteriorate on away from but not home form as well.
Although these 6 matches are more recent than the last 6 away matches, this, together with the fact, that Arsenal hadn’t shown deterioration on the 6 match stats compared to one year stats, would convince me that Arsenal’s poor results in their last 6 away matches was just variation or bad luck.
D. Conclusions
There’s a saying in Gestalt Psychology, “The Whole is Greater than the Sum of It’s Parts”. Often, individual stats don’t tell you anything. However, when you put different sets of stats together, they can sometimes tell a story.
So, I’ve explained a 2 step process for analysing the last 6 matches.
This isn’t a lot of work. It’s much quicker to do this analysis than it is to explain. At least it is, if you are using my calculators.
Although this is not a fool-proof formula for analysing the last 6 matches of a football team or recent form, it is a better method than just scrolling through the last 6 matches and just saying to yourself, I reckon that they should have won that match because they were favourite.
It is a way of weighing up how 2 teams have performed in their last 6 matches. In addition, this is part of a bigger analysis, in which we are trying to put the long-term analyses, together with short-term analyses.
In addition, the idea of comparing 2 sets of stats of different lengths of time can be applied to more than just this situation.