After almost a year in charge, how much improvement, if any, has the German been responsible for?
This is an introductory look at how the media love-in with Klopp actually contrasts with the raw data thrown out by his reign so far.
You can download this project in PDF format by clicking the button below
In the light of recent media adulation of Klopp and his hyperactive style of football, and having learned to think for myself a long time ago, I decided to set about physically gauging what improvements he has made to Liverpool, if any, since his appointment.
Now,it’s important you don’t get me wrong from the outset here. I admire the affable German and like his and Liverpool’s style of play so far this season. The intensity and desire to play entertaining football is to be lauded but that’s not to say it’s particularly effective over long periods of time.
Of course, his short time at the club so far does not afford us a sample that considers what we could call a ‘long’ period of time but it does give us a set to lay down some foundations now and then consider the progress or lack of it come the end of the season.
With this in mind, this article is only intended as a ‘Part 1’ of the story and we will revisit it at the end of the season when we have much more data to go off.
To monitor how Klopp is doing, we must first establish exactly what it was he inherited from his predecessor, Brendan Rodgers. That was, I think we’d all agree, a team in transition as soon as Suarez left but losing out on the title, which however unlikely looked theirs for the taking, certainly played a part too. All of this data is incorporated to give an idea of how Liverpool performed throughout their time with Rodgers in charge so as to give us an accurate picture of what state the team were in when Klopp took over.
If we know what he inherited, it’s then much easier to gauge whether he has actually improved the side or not and whilst on displays this season they look a more accomplished outfit, it’s much more sensible to base opinions over longer time frames.
The excitement of the moment leads to snap judgements and subjectivity whereas if we only look at the raw numbers as an indicator of performance we get a much more objective view and also a better idea of what the long term future may hold.
I have attempted to explain the figures generated in regular places throughout the article but if anything is unclear please don’t hesitate to contact me for clarification
Overall performance since Rodgers first game
Let’s first look at the ultimate arbiter of football performance – goals – or perhaps more precisely Liverpool ability to score more and concede less than an average team.
The chart below shows Liverpool’s ability to do the above better than an average team since Rodgers first game right through to Klopp’s most recent match.
Before we talk about what this chart shows, let’s take a look at the metric it actually uses so as to make better sense of it.
The ratings for goals are, in layman’s terms, how many goals are Liverpool better off than the average team in any given match. So if the average team was expected to score 1 goal and Liverpool were currently rating 0.5 on the chart above, we would expect them to score 1.5 goals as their rating is 0.5 goals better than the average side.
Makes sense? I hope so but we’ll come back to it.
It should be noted there’s a very bad start for Rodgers here but this doesn’t serve as a handicap as the rating is a cumulative average over all the games meaning the early matches look obscure on the chart but don’t have a negative impact on the actual ratings that we’ll be comparing in the most part – Klopp’s influence on the side. The whole point of incorporating Rodgers entire reign into the stats is to guard against this kind of issue and ensure we have a realistic rating on which to start Klopp’s reign further down the line.
So what does this show?
Well we see a tough start for Rodgers which is later catered for in the chart as we see gradual progression, peaking in the latter stages of Liverpool’s close shave with the title. Following that disappointment, the loss of Suarez (would affect any team) and the possible damage to chemistry caused by the resulting mass recruitment drive, we see a gradual sloping off in their effectiveness.
When you consider only headlines, you may think Rodgers was underachieving in the latter part of his reign. However, it’s more likely he was a victim of a freak season when they went close.
The numbers for that season, and we don’t look at it here but I can present the work if anybody is interested, were simply not sustainable with very high conversion ratios in terms of shots vs goals. The more retentive of you may recall the season prior to their title challenge when they were creating plenty of chances but missing far too many and simply not scoring enough goals.
Like the latter, the former could not last long term and I think it fair to say that Liverpool were probably performing exactly as we’d have expected them to, even as late as the latter stages of Rodgers reign. When we consider how much ridicule Rodgers attracted at his end and the seemingly never-ending praise of Klopp since then, it would be easy to imagine that the German had totally transformed the clubs fortunes.
Whilst it’s arguable that the overall atmosphere around the club has changed significantly since and the German will need some time to make it work for him, at this point the numbers don’t really prove this out. In fact, if we consider the chart that follows we can see there has been absolutely no improvement at all in their ability to win football matches since the appointment of Klopp.
As you can see here, Klopp inherited a team better at winning football matches (scoring more and conceding less than the average team) than the one he currently possesses almost a year on.
The actual team goals rating in Klopp’s first game was 0.661 (so for every 1 goals the average team scored against them, Liverpool would be expected to score 1.661 in reply) compared to the rating 12 months later (for a supposedly ‘improved’ team) of 0.656. So after 12 months of Klopp, on the face of it Liverpool are now 0.005 of a goal worse at winning football matches and at the top end of football with such small margins at play, a 1% deterioration in team performance simply does not back up the current media love-in with the German.
It will certainly be interesting to see where the numbers lie come the end of this season and as I mentioned previously, I will be going back to this with new data at that time. You will find updates athttp://www.betfairguru.coand you can subscribe to receive all updates and new content via email as soon as it is published.
In the meanwhile, there are of course other metrics which show the development of a team but are possibly more open to variance. These are things that we already touched upon, shots and shots on target.
Let’s see the chart for shot differential since Rodgers first game to begin with, then we’ll break it down for just games under Klopp to gauge progress. We’ll then repeat that process again for shots on target as it could be we’ve been in a period of low conversion and are about to see a period of high conversion, which with more games in the sample would give a better rating to Klopp currently.
This chart concurs somewhat with the goals chart up and to a point… the point being when Klopp takes over which is when we see a gradual improvement in shot differential – the creation of shots and prevention of shots also.
This would be explained by the change in style since his appointment and that winning the ball back higher up the pitch and controlling the game more will offer you more opportunities and your opponents fewer. This contrasts with Rodgers efforts where Liverpool would be much more reactionary, inviting teams out and scoring most of their goals getting in behind teams with pace.
Let’s break down what happened since Klopp arrived to see this idea in visual format
Of course, there is every chance that this is just a regression to what’s expected and that Rodgers style was inviting too many shots against the side rather than Klopp’s Liverpool creating more, all of which can be considered now but not really proved out without a lot more data to look at. I also think it only fair to give Klopp time to get the players that he wants playing the system that he wants before judging too much. In his early reign he was lumbered with a dysfunctional squad, little in the way of depth and low on confidence so over time I would have expected improvement anyway, the only question is how much.
On the face of it though, Liverpool are now around 0.5 of a shot better off per game, which in turn should equals a better goal rating over his time in charge. However, as we’ve already seen, that’s now what is happening so what is the explanation?
We should consider what kind of shots these are since a wild hack from 85 yards in the vague direction of the goal can be labelled a shot. Let’s look at Shots on Target ratings, first throughout the entire measured spell (Rodgers first game onwards) and then look at what happened since Klopp arrived.
As expected, this chart also mirrors the others we’ve already seen with a drop in performance since the peak of their title challenge. This would suggest that Liverpool’s style of play was pretty consistent throughout Rodgers reign and the goal shortage, followed by a goal glut were most likely just extremes of variance.
What you may note is that the rating is still falling even towards the end of the chart so let’s look at what happened explicitly since Klopp arrived.
It would seem that, whilst they are gradually improving the amount of shooting opportunities they have, they are not necessarily good opportunities. In fairness, the drop has been small from 2.79 to 2.65 but if you listen to the media, Liverpool are now all of a sudden a free-scoring side under Klopp creating lots of opportunities in the opponents final third.
The truth of it is that since he took over, according to the raw data we have at the moment, they are worse at the variables that directly determine whether or not you win football matches than they were under Rodgers, even in his last games in charge where the perception was that he’d totally lost the plot.
Now, at this point it’s easy to question what the German has brought to the party but, as I said previously, I am a huge fan of his refreshing brand of honesty and exciting style of football and have no doubts at all that it will ultimately prove successful. In fact, I am looking forward to coming back to this at the end of the season and using Part 1 as a case study example of why you should not place too much trust in short-term data sets.
Now I’m one that’s all for fairness and I know that an easy argument could be that we didn’t look at the league in it’s entirety. Maybe these figures drop off across the entire league and not just Liverpool under Klopp. However, the way the ratings are calculated means that this is already factored into the rating. So if the entire league dropped off, it would just show as stagnation on the Klopp chart, it wouldn’t show as a drop or a rise. The only time we get rises or drops in rating are when the team is performing better or worse than the average team over a the same time period.
I haven’t considered such metrics as Possession and distance covered since I don’t believe they bear much effect on winning football matches. If they did consider these, and conventional belief is that both of these things have improved under Klopp, we can confidently say they aren’t helping Liverpool to win more football matches over the current sample.
So finally here we are at a point where we can post a provisional conclusion with a view to revisiting it in 8 months time at the end of the 2016-2017 football season. At that point, we will have a better sample size of Klopp games but also a better sample size for the rest of the league during the same period of time.
I hope you enjoyed this brief guide and when I compile things like this, I like to look a little beyond what usually counts for “analysis” in modern-day lazy journalism.
Most of the time they just look at points won over the same sample or win% compared to the previous manager but this is far too simplistic as it gives no attention to who the opponents were during the same time frame.
Whilst extremes of form are temporary, team strength is quite organic and changes very slowly over long periods of time. Given what we have seen so far, it may look exciting and you’ll get the odd goal-fest but there’s been no improvement at all, perhaps the team even went backwards over this time.
If you wish to receive the follow-up to this at seasons end, just subscribe to my mail-list at – http://www.betfairguru.co – it’s completely free and I don’t spam you. In fact, I only ever send free content out as and when I publish it on my website.
With the season due to end on 21st May 2017, I’d expect the next part to be released towards the end of May 2017, possibly the first week of June. In the meanwhile, there’ll be a lot more projects like this for you to enjoy and my most popular that I’ve had to date was my Euro 2016 Project which you can download directly from the website.