Best Goalscorer in 2018/19?

I’m going to look at some stats to see who was the best striker in Europe last season.  As usual with this blog, I am using data from understat.com, a great resource.  I looked at players who scored 10+ goals in Europe’s top 5 leagues last season (England, Spain, France, Germany and Italy).  This gave me a total of 90 players, so a pretty big data source to look at.

It should be easy to find out who is the best striker around, surely the striker who scores the most goals is the best striker?

In theory that makes sense and last season that would mean that Lionel Messi’s 36 league goals for Barcelona would make him the top striker in Europe last season.  No surprise that the best player in the world is the best striker.  Lets have a look at the top ten goalscorers from last season.

But what happens if we say that penalties shouldn’t count when considering how good a striker is.  That’s fair enough I guess, as there is a lot of luck involved in penalties.  Let’s take them out of the equation and see what happens.  Well Messi’s goal tally goes down to a meagre(!) 32 goals, but this puts him level with PSG’s Kylian Mbappe.  Mbappe only scored 1 penalty which is why he jumps up.  As you can see from the below, this now also brings in players like Karim Benzema and Sergio Augero.

So Messi and Mbappe finish as the best strikers in Europe, end of argument, right?  Well no, Messi played 370 minutes more football than Mbappe, so he had more time to score his goals.  So let’s look at non penalty goals per 90 minutes played, this would give a fairer comparison between players.

This is where we see a new name, Paco Alcacer of Broussia Dortmund, who scored 1.30 non penalty goals per 90 minutes, ahead of Mbappe on 1.23 goals and Messi on 1.20 goals.  Also appearing on this list are Ishak Belfodil, Edinson Cavani and the late Emiliano Sala.  So now we have Alcacer as the best striker in Europe.  But is that the case?

By now, most people who read these types of blogs know what expected goals are.  This is the stat which gives each shot at goal a value between 0-1 depending on the location, etc of the shot on how likely it will be scored according to previous shots from the same location.

So we are already down to NPG90 with our stats, so let’s bring in a new stat which is xNPG90.  This is how many non penalty goals a player is expected to score every 90 minutes.  From this, Mbappe jumps back up to the top spot with an xNPG90 of 1.09, followed by Robert Lewandowski on 0.92, Messi on 0.74 and Alcacer on 0.72.  I have already blogged about Lewandowski which you can read here.  This chart also sees the debuts of Suarez and Ronaldo.

So is there anything else we can consider?  Well what about the difference between xNPG90 and NPG90, this will show who is performing the best over expectations.  There is a bit of a change here, Alcacer goes back to the top with an NPG90 Diff of 0.58 (basically he scores 0.58 goals more than expected every 90 minutes), followed by Messi on 0.33 and the late Emiliano Sala with 0.31.  Sala was having an incredible season before his tragic death in January 2019.  Mbappe falls down to number 10 in this list with 0.14.  Lewandowski falls all the way down to the worst performing striker with -0.34.  So, the top 10 strikers in Europe this season using this in depth stats are:

We also see on this list Iago Aspas who had a great season at Celta Vigo, along with Jadon Sancho, the young English player at Broussia Dortmund.

It will be interesting to follow Alcacer this season to see if he is able to keep up this strike rate, especially if he ends up getting a lot more minutes with Dortmund.  Without a doubt we can expect Messi to continue his excellent goalscoring record.

Advertisements

Comparing Haller v Perez – who scores more in 19/20?

In this blog I’m going to have a look at a couple of new strikers who have been brought in by West Ham and Leicester – Sebastien Haller and Ayoze Perez.

Once again the stats used for this blog will be from understat.com.

Leicester spent £30m to bring Perez from Newcastle.  This was to replace Okazaki whose contract ran out at the end of last season.  When it comes to goals, Okazaki should be easy to replace as he was never the biggest scorer for Leicester.

Perez last season scored 12 goals for Newcastle, working out at 0.39 NPG90 compared to an xNPG90 of 0.25.  Perez in his career at Newcastle always scores at a higher rate than his expected goals, which is always a good sign for a striker.

Perez should in theory see an improvement this season coming in his goal chances.  Last season with Newcastle he had an xNPG of 8.29 goals, despite this he ended up with 12 goals.  This came despite Newcastle only having a rate of 11.72 shots per 90 minutes.  This low rate is also shown in the Key Passes per 90 Minutes of 8.54.  This compares to Leicester last season of Sh90 of 13.63 and a KP90 of 10.13.

If Leicester keep up this rate of chance creation, then Perez can look to get 15+ goals this season which can take a lot of pressure off of Jamie Vardy, who has borne the burden of goals for Leicester in recent years.

Sebastien Haller scored the same amount (12) NPG goals as Perez last season before his £45m move to West Ham but from a higher xNPG of 11.42 goals.  Haller’s NPG90 of 0.50 and xNPG of 0.47 however, was higher than Perez.

One of the issues that Haller may find is West Ham’s lack of chance creation.  They were very close to Newcastle last season with a Sh90 of 11.62 and KP90 of 8.59.  Haller’s previous club Eintracht Frankfurt had a Sh90 of 12.37 and KP90 of 9.03.  For Haller to improve on his 12 goals last season, he needs the service. 

This blog is a primer for something I will be writing over the next few weeks, which is how teams always want to sign a goal scorer to take them to the next level, when sometimes it is best to focus on getting in a goal creator.

I’m going to stick my neck out and say that Perez will score more goals next season than Haller, mainly due to the additional chances he will get from his Leicester teammates.

Chelsea – Who can replace Hazard?

The biggest transfer so far this summer is Eden Hazard leaving Chelsea for Real Madrid for a reported £88m fee, that may rise up to £130m.  A quick cheap plug here, but the size of the upfront fee shows that the valuation of £80m that Crystal Palace have for Wilfried Zaha is insane.  I talked more about that here

Getting back to Hazard, because of Chelsea’s transfer ban they haven’t been able to replace him with a big transfer, so will have to relay on players who are still at the club.  I’m going to have a look at who is capable of stepping up to replace the Belgian or is it more likely that nobody can replace him.

Chelsea’s options as a direct replacement are Pedro, Willian, Callum Hudson-Odoi and Christian Pulisic.  Pulisic of course was signed last January but loaned straight back to Dortmund, now he has the chance to make his mark at Stamford Bridge.

The stats I will be using are supplied by understat.com.  I have also done the stats per 90 minutes play.  The first chart I have is comparing Key Passes Per 90 Minutes and Shots Per 90 Minutes.

As you can see from the chart, the close to the top right of the chart, the better the stat.  Willian actually averaged more KP90 than Hazard, whereas Willian, Hazard, Pulisic and Pedro had a similar Sh90 stats.  Callum Hudson-Odoi struggled to get shots off last season, but was averaging over 2 KP90.

The next chart compares Expected Goals Per 90 Minutes against the Shots Per 90 Minutes

What we can see from this despite Willian having a high number of shots, not many of his shots are of high quality.  Christian Pulisic however is averaging over 0.50 xG90 which is a great rate for an attacking midfielder.  Callum Hudson-Odoi again shows that goals will probably not be a big part of his game.

Up next is time to look more at creativity rather than goals.  While Hazard does bring goals to his game, his biggest asset is creating chances.

The above chart shows Expected Assists Per 90 Minutes and Key Passes Per 90 Minutes.  Callum Hudson-Odoi comes into his own here.  He actually has a higher xA90 rate than Hazard (and everyone else), so the chances he creates are of high quality.  Pedro struggles here, whereas Pulisic and Willian also stand out.

The final chart is going to compare Expected Assists Per 90 Minutes and Expected Goals Per 90 Minutes.  This will see who actually does mix in both creativity and goals.

As you can see, Hazard stands out here, not only as a goalscorer but as a creator.  Christian Pulisic could fill in the gap with the goals, averaging over 0.50 xG90, although he doesn’t create as much as Hazard did.  Callum Hudson-Odoi can be the creative force for Chelsea next season if he is given more of a chance.  There has been talk of a big money move to Germany for him but Chelsea have to keep a hold of the youngster and give him a chance next season.  Pedro and Willian are both more experienced, but as you can see from their stats that they aren’t going to cut the mustard.  If Chelsea continue playing with the front three then both Pulisic and Hudson-Odoi have to be first choice.

Neymar – Should he stay or should he go?

One of the biggest transfer stories of the summer has been Neymar requesting to leave PSG and hoping to go back to Barcelona.  In this blog I’m going to look at how well Neymar performed last season compared to both his PSG teammates and his peers at Barcelona.  I have also included Barcelona’s major summer signing, Antoine Griezmann, to see how he also compares.  Once again, the stats I’m using come from understat.com which is a great resource.

The first chart compares Non Penalty Goals per 90 Minutes, alongside their expected equivalent.  The closer to the top right a player is the better he has performed and of course the closer to the bottom left the worse they performed.

As you can see Kylian Mbappe stands out on his own here.  Mbappe scored 1.23 NPG90 last season off an xNPG90 of 1.09 goals.  This is an incredible output by the young Frenchman.  Lionel Messi is next with an NPG90 of 1.07 compared to his xNPG90 of 0.74 goals.  Neymar sits in the middle group along with Suarez and Cavani.  Interestingly enough, Griezmann didn’t have the best scoring season last year.

From this chart, it looks like PSG won’t really miss Neymar’s goals as long as Mbappe continues this sort of form.  Barcelona, however, may be strengthened with Neymar coming back to the Nou Camp, as Luis Suarez scored less NPG90 than his xNPG90 and with Griezmann, Coutinho and Dembele not looking like prolific scorers.

What does Neymar bring to PSG in terms of assists?  Lets have a look at that and see if this is something that PSG would miss if Neymar leaves this summer.

Neymar is definitely more of a creator for PSG than a goalscorer.  If Neymar leaves, this would have to do replaced as Mbappe could struggle for goals if the chances are not being created for him.  Barcelona having Messi and Neymar back together creating chances could be lethal for any opponent though!

Neymar can’t just be judged on one season, so lets have a look at what he has done over the past 5 seasons.

Neymar’s best performances seem to come when he was at his happiest, back in 2014/15 at Barcelona and his first season at PSG.  His most unhappy seasons have been 2016/17 at Barcelona and 2018/19 at PSG.  Neymar’s morale and happiness do seem to make a big difference to his goal scoring prowess.

Looking now at his assists over the past 5 seasons, you can see that his best season again came when he was happy at PSG in 2017/18.  Considering how good of a creative season he had in 2018/19, this chart shows how good he was in 17/18 and how consistent he is overall with setting up teammates.

So, should PSG sell Neymar?  Short answer is yes.  When Neymar is unhappy his performances do suffer.  Mbappe should be able to take over the goalscoring duties and should still improve with age.  Barcelona will benefit from Neymar’s goals considering they seem to be relying more and more on Messi.

Creatively PSG may suffer.  There has been talk of Dembele and Coutinho moving to PSG as part of the deal.  Despite not having the goals, they are creative.  Only Messi and Neymar had more xA90 than Dembele and Coutinho last season, so they would be relied on to create the chances for Mbappe.

This could be win-win for everyone involved (well, for maybe two seasons for Barcelona before Neymar gets unhappy again!)

Zaha to Arsenal – The £100m Question

This summer has seen Arsenal linked with a few players as they try to strengthen their squad for next season.  The main target seems to be Wilifried Zaha of Crystal Palace.  The fees quoted have ranged from £60-100m.

In this blog I’m going to look to see if Zaha can improve Arsenal enough for it to be worth spending nearly £100m on him, but I’m also going to look at a player that Arsenal were linked with earlier this summer, Bournemouth’s Ryan Fraser.  The young Scot has been valued at £40m if Bournemouth want to cash in this summer. Again, I have used http://understat.com to help with this article

Lets look at the Key Pass stats for Arsenal along with Zaha and Fraser from last season.

As you can see from the above chart, both Zaha and Fraser made more key passes than any Arsenal player, but it’s the 93 key passes from Fraser which stands out.  Fraser made more than double the amount of key passes than Mesut Ozil who lead the way for Arsenal.

When this gets broken down into Key Passes per 90 minutes, you can see that Ryan Fraser still leads the way with 2.63 key passes per 90 minutes.  Zaha falls behind Ozil, Iwobi and Ramsay with 1.70.  Both Zaha and Fraser played a lot of minutes last season, compared to the Arsenal players who were rotated more, so this explains why this has become closer.

The next chart below is where things get interesting.  I feel Expected Assists are a better judge of a players creativity than actual assists are.  For example, you can make a 3 yard pass to a player who goes and slams home a 35 yarder into the top corner and be given an assist, or you can play an excellent ball between defenders that a striker misses from 6 yards and you don’t get any reward for that.

As you can see, Ryan Fraser is way out in front here.  Fraser in fact had the highest xA stat of anyone in the top 5 leagues in Europe last season.  Zaha is middle of the pack here with 5.40.

When is comes to expected assists every 90 minutes, Fraser is still leading with 0.46 per 90.  Zaha is away down with only 0.16 per 90.  Arsenal are needing more creativity in their team and as you can see from these charts, not only is Fraser going to be cheaper to buy, he brings more creativity with him.  The charts above also show that Alex Iwobi should get more game time next season.

Now what I’m going to do is look closer between Fraser and Zaha.  Was this just a one off really good season for Fraser or has Zaha shown more consistency over the years?

I have stats on Fraser going back 3 seasons and I have 5 years worth of stats for Zaha.

As you can see both players have increased their Key Passes recently, only the 2016/17 season for Zaha saw a slight dip.  This is good to show that both players may still be improving and not hit their peaks yet.

When it comes to breaking the key passes down over 90 minutes, you can see that it is more up and down for Zaha, whereas Fraser has continued to improve.  Only Zaha’s 2017/18 season has beaten any of Fraser’s three seasons in the EPL.

Zaha looks better at the moment with Expected Assists.  Zaha has consistently improved season on season, whereas Fraser improved incredibly in 2018/19.  Can Fraser improve, or at least keep this level going next season, that would be the risk that Arsenal have to think about.

Expected Assists per 90 Minutes are much closer (apart from Fraser’s 18/19 season).  This shows that Fraser’s lack of xA in previous seasons were down to minutes played and if he plays a similar amount of time next season that his xA should be higher than Zaha.

Zaha does score slightly more goals than Fraser, although Fraser’s goals per season are increasing.  Zaha’s extra goals compared to his lack of xA may be because he is more likely to shoot himself when getting into good positions, rather than looking to set up a teammate.

The non penalty goals per 90 minutes, also shows that Zaha is more of a goal threat.

From the stats and charts above I’m convinced that Arsenal should save themselves some money and go back to trying to sign Ryan Fraser from Bournemouth.  Zaha is an excellent player and probably better than Arsenal’s current creative players, but he is not worth close to £100m.

The other related rumour to this was for Arsenal to afford the Zaha transfer they may need to see their top scorer Pierre-Emerick Aubameyang.  Lets finish off this blog, with a quick look to see if this is the time for Arsenal to cash in on Aubameyang.

Looking back at Aubameyang’s goals over the past 5 seasons and you can see that he peaked back in 2016/17 with 29 goals.  This was the season before he joined Arsenal in the winter transfer window of 2017/18.  You can see in his two full seasons after that peak that his goals have dropped to 19 and 18.

Aubameyang’s expected non penalty goals follow a similar pattern, in fact there isn’t much difference between expected goals and actual goals.

The non penalty goals per 90 minutes continue to follow the same pattern, decreasing since he joined Arsenal during the 2017/18 season.

Aubameyang’s xNPG90 has also decreased, so his downturn in goals in not due to any lack of minutes played.

The next two charts, however, may tell the story.

Aubameyang’s shots taken dropped dramatically last season by 24.  This was his first full season with Arsenal and the lack of creativity in the squad may be the season.

In conclusion, Arsenal should keep Aubameyang and sign Ryan Fraser.  Hopefully Fraser’s high output of key passes, along with his high expected assists should give not only more chances, but a higher quality of chances for Aubameyang.  If Arsenal want in the Champions League next season, they need Aubameyang and Lacazette scoring as many goals as possible.

Robert Lewandowski

In this blog I am going to have a deep dive into Bayern Munich’s Polish striker Robert Lewandowski. Lewandowski is of course a huge name in football, with a great goalscoring record, at club level and international level. Lewandowski recently signed a new 4 year deal with Bayern. Is this a good deal for Bayern, tying down one of the top strikers in Europe, or have they tied themselves down with a striker who is past his peak?

Now, simple stats like goals scored can tell some stories, but I need to go deeper to find out the true story. I am going to look at expected goals, which has come into the forefront of football stats over the past few years, along with a few other stats. I like to use a great site – Understat.com, which is one of the top resource websites on the net for expected goals, and if you are a fan of stats in football you should check them out. I have collated a lot of information from there, and added in some calculations myself to try to get a true picture of Lewandowski’s recent form.

In this blog, I will be talking expected goals, goals per 90 minutes, and shots per 90 minutes over the past 5 seasons.

I decided to look further into Lewandowski when I started looking at strikers who had scored 10 or more goals this season in the top 5 leagues in Europe (Germany, England, Italy, France and Spain). What stood out to me was the difference between Lewandowski’s goals scored to his expected goals.

In theory, these two stats should be very similar, but for a top level striker their actual goals scored should be higher than their expected goals. Basically, that shows that they score tougher chances. Someone who has a higher expected goals total than goals scored, shows that they are missing chances than they should be expected to score. What I’m also using is non penalty goals, which means that any penalties are not taken into account, so it is all based on open play goals.

To give an example, and this won’t be a shock to anyone, Lionel Messi had an xNPG (Exepected Non Penalty Goals) total of 22.28 goals this season, but actually scored 32 non penalty goals. That is the sign of a top level striker. On the other side of the spectrum, was Lewandowski. He has an xNPG of 33.14 goals, but only scored 22 goals for a difference of-11.14. This was more than double the next worst, which was Andrej Kramaric of Hoffenheim, who had a difference on -5.03 off an xNPG of 17.03 and goals of 12.

So what has caused this? Is it just a one off blip, or is it a sign of a regression of finishing skill for the big Pole?

First thing I look at was Lewandowski’s shot map for this and something stood out. You will see I have isolated the shots that Lewandowski missed the target with. There are 11 shots that missed inside the 6 yard box this past season. This isn’t shots that have been saved, or hit the post, this is actually missing the target completely. I have also posted shot maps for the previous seasons too and you can see the increase in shots missing the target (he has never missed more than 6 in one season).
2018/19

2017/18

2016/17

2015/16

2014/15

The next charts will look at the next two which are xNPG and NPG. Lewandowski’s NPG peaked in 15/16 with 28 goals, but this has slowly decreased to 25, 23 and then 19 last season. In the same time his xNPG has gone from 26.95, to 26.31, to 22.59 but then increased last season to 30.11 goals.

So, maybe the goals scored has decreased due to time spent on the pitch. So I looked at Lewandowski’s NPG per 90 minutes, to see if there is a similar pattern here.

The xNPG 90 has stayed steady over the past few seasons. In 2015/16, it was 0.91, then dropping slightly to 0.85 in 16/17 before climbing to 0.94 in 17/18 and then 0.92 last season.

For the actual NPG 90 this is where it gets worrying. His NPG90 goes from 0.95 in 15/16, to a dip in 16/17 of 0.81 and then climbing back up to 0.96 in 17/18. As you can see, these compare favourably to his xNPG90. There is now a huge drop last season, his xNPG90 is 0.92, but his actual NGP90 is only 0.58.

Now I should say that Robert Lewandowski is one of my favourite players, so I was wanting to find something that would explain this poor season. Perhaps then, he is taking less shots. So I did some analysis on his Shorts Per 90 Minutes.

He has taken less shots than normal last season than the previous few seasons (4.38 per 90 v 5.28, 4.63 and 5.15) but its not enough of a drop to explain the decreases in finishing. In fact, due to the increase in xNPG this season means he is actually getting much better chances this season than in any other season.

Lewandowski turns 31 in the next few months, so it is looking like he has hit his peak at club level and it might be that Bayern need to look for replacement. One player they have been linked with is Timo Werner, so I have looked into Werner to see if he looks like he has what it takes to replace Lewandowski.

Werner started off with low scoring seasons in 14/15 and 15/16 with 3 & 6 NPG respectively. But 16/17 saw big increase with 19 NPGs against an xNPG of only 13.02. 2017/18 saw a slight dip to 11 goals off an xNPG of 12.06 (not a worrying difference as he has scored close to what was expected of him).

This past season saw Werner’s NPGs increase again to 16 off an xNPG of 16.63. It makes you think if Werner had close to the chances that Lewandowski had last season, would this total NPG be in the high 20s?

Werner’s xNPG 90 over the past couple of seasons has stayed even at 0.49, 0.48, 0.43 before jumping up last season to 0.59, while his actual NPG90 has been inconsistent at 0.25, 0.70, 0.40 before jumping up to 0.57. It is interesting that his xNPG90 and actual NPG90 have been very close over the past two seasons.

Werner’s shots per 90 minutes has also increased season on season from 1.67 in 14/15, to 2.62, 2.72, 3.42 and then 3.80 last season.

From the stats available, this could be the perfect summer for Bayern to look at replacing Lewandowski. It does seem strange to say that a striker who only had 5 players score more non penalty goals than him last season in the top leagues in Europe.

Create your website at WordPress.com
Get started