Interesting article. I like stats and when there are so many stats working out which one to focus on is the key. The WOWY number is interesting, if you have enough games. In this case we don’t. 
 

https://www.espn.com.au/sports/soccer/insider/story/_/id/32745800/chelsea-need-romelu-lukaku-far-better-him

Anymore of those articles and I'll be claiming copyright.

Edited November 30, 20214 yr by Argo

Sorry don't trust stats unless they are written and publish exclusively by @Gol15 😉

It is certainly food for thought but it is written in a way that would get raked over the coals in undergrad. The problem is that he draws conclusions before explaining the assumptions of his 'model', and the result is that it downplays their role or suggests they are minor considerations.  Sample size and the general unsuitability of football to WOWY stats are not minor concerns but major flaws that prevent the numbers being used in any serious analysis. The potential for random action and the sheer interconnectedness of an 11 v 11 game leads to so much variance that an accurate statistical model is impractical if not outright impossible.

The numbers do not really provide any definitive answer, either. Supporters of Lukaku would point to xG numbers and say tht the low xG is because the team doesn't use him to his full potential; detractors will say that he takes away from the team's style of play as whole. Given the statistical context both lines of argument are justified and to be honest, the truth lies probably somewhere in the middle. 

What is interesting to observe is that scoring more goals is not merely about signing a more prolific striker. The team need to adapt to the striker as well as be able to create chances independently of that player. At Chelsea, using Lukaku as a focal point seems to result in a situation where he is expected to be both primary creator and primary goalscorer similar to Harry Kane at Tottenham, whereas the likes of Salah and Mane are not really expected to provide for each other as much as get on the end of moves created by their midfielders/wingbacks. The Man City game showed this ugly truth where Lukaku was relegated to an ineffective dummy whom they simply allowed on the ball and counted on him being pressured to create. It creates a catch-22 where you have a sub-optimal creator trying to provide for a sub-optimal finisher. A similar argument could be made for Werner played as a lone striker against Man U (or really any time he has played as such...)

Edited November 30, 20214 yr by SydneyChelsea

Well written SydneyC.   The team does need to adapt to the striker. I think Salah does a bit more than get on the end of wingback runs. Each of Liverpool, ManC and us see a number of goals created down one touchline but finished at the far post by someone coming through.

Against Man C Lukaku had an impossible job, completely isolated and surrounded by at least 2 defenders on the halfway line. That game we played too defensive.

I'll feel more optimistic when he's fit and we can see how things go.

All that said the WOWY stat is an interesting one and despite the statistical issues I'm sure if I was a coach I'd still be looking at it.

Probably also worth noting that the analysis doesn't go very far beyond surface level, only comparing goals/xG/goal difference per game with and without Lukaku. Obviously there's quite a lot more going on during a football match than that, and it doesn't even begin to consider other environmental factors (e.g. relative quality/qualities of opposition, formations, which teammates he is playing alongside, etc.)

Of course a lot of this is hard to do, as has been pointed out in this thread (and is in fact is acknowledged in the article itself), when we're not even halfway through a single season. We haven't even played every other team in the Premier League yet.

Largely pointless analysis at this moment. If he had concluded with "Well anyway... food for thought" rather than "So far this season, Chelsea have been better without their club-record signing" it might have a little more credibility...

There is really nothing to say. The xg number is better without Lukaku, so? 

Is it because of Lukaku? Is it because of the opp? Is it because of CHO? Is it because we suddenly score using set pieces and the opp has to open up? 

Like @SydneyChelseasaid, there are so many variables in football game, impossible to judge. 

I personally think we do look better without Lukaku right now but I prefer Lukaku to play. It sounds crazy but I believe if we can integrate Lukaku into this team better, the ceiling for this team is so much higher compare to the one without him. 

 

6 hours ago, Bob stark said:

There is really nothing to say. The xg number is better without Lukaku, so? 

Is it because of Lukaku? Is it because of the opp? Is it because of CHO? Is it because we suddenly score using set pieces and the opp has to open up? 

Like @SydneyChelseasaid, there are so many variables in football game, impossible to judge. 

I personally think we do look better without Lukaku right now but I prefer Lukaku to play. It sounds crazy but I believe if we can integrate Lukaku into this team better, the ceiling for this team is so much higher compare to the one without him. 

 

Well we played City and Juve with him and look at our performances in them compared to when we faced the same sides without him. The signs were there against Villa and Zenit but he took what little he/we created (and it the case of Villa Edou played like superman) so we got away with it but it was clear as early as then he was compromising the system.

The main issue with Lukaku is his pressing. His so so general play could probably be worked around but the pressing is a huge huge issue, without him we're arguably the best team in Europe at it, with him earlier this season we were marginally better than Spurs, yes Nuno's lifeless Spurs.