Henderson and Comolli
I finally found time to listen to the podcast of Sky Sports' debate on football analytics, from December last year. If you're interested in such things, I thoroughly recommend it (link). One of the contributors was Liverpool's former director of football, Damien Comolli. His comments on the scouting of Jordan Henderson in 2011 particularly piqued my interest, as they made me wonder if my chance quality work might have given the Sunderland man's impressive chance creation figures some important context.
"We looked at his data thoroughly" said Comolli. "He created chances for fun".
There's certainly no doubt about that statement. Henderson created eighty-two goal scoring chances in 2010/11, which was the sixth most in the division. He dropped to joint-eleventh on a chances per game basis, but that's still very impressive considering he was just twenty years old at the time. Comolli continued:
"You break down chances created, when looking at a midfield player. You break it down: chances created in run of play, chances created from free kicks. We're looking at Jordan Henderson, all these fitness tests which was very good, then we realised how many chances he created, disregarding the set plays."
This is where I start to question Comolli. 38% of Henderson's chances in 2010/11 were from set plays; strip those out and he created fifty-one in open play. This equates to one every sixty-three minutes that he played. Yet the most creative players that season were providing their team mates with opportunities in open play every thirty-five minutes or so; approaching twice as often, in other words. If we look at players who created at least twenty-five opportunities in 2010/11, then Henderson was ranked fiftieth for frequency of open play chances created in the season before he joined Liverpool. Far from poor, but hardly an earth shattering record either. Thirty-three players bettered Henderson's clear-cut chance tally of seven too (which includes set plays; open play only CCC data isn't available unfortunately), and in percentage terms, just 8.5% of the Mackem's chances were clear-cut. Of the forty-five players to create at least seven CCCs that season, only one (Florent Malouda) had a lower proportion of top quality chances than Henderson did. Having assessed Henderson's chances through the prism of my chance quality system, I think perhaps Comolli might have been better off looking elsewhere if chance creation was the attribute he was looking to add to the Liverpool team. A reminder of the zones I use to assess chance quality:
Chart by @natefc of Oh You Beauty
Let's have a look at how many chances Henderson created in each of the six zones, and how many assists he would be expected to provide with these particular goalscoring opportunities.
I don't like to use assists when assessing the creative abilities of a player, as they rely on the player who receives the key pass converting the opportunity in order to be counted, and that is outside the hands of the chance creator. I would therefore usually ignore the fact that Henderson only got four assists in 2010/11, and focus on the eighty-two chances he created overall.
The most interesting thing for me here then is that my system would expect Henderson to have got 4.8 assists from open play in 2010/11, so the fact that he actually got four suggests that the system could be pretty accurate.
But perhaps more importantly, it shows that the quality of chance that Henderson was creating would only expect to create a goal every eight games or so; that doesn't seem like a lot to commit £16-£20m of your transfer budget to (depending on what the transfer fee for Henderson actually was of course).
I must stress, this article is not meant to be critical of Jordan Henderson, who I think is a fine player and should play every game for Liverpool (as indeed he has this season). Rather, this questions the methodology that Damien Comolli used when assessing which creative talents to buy for Liverpool in the summer of 2011.
One final point: I wonder if chance creation charts like the one below lead Comolli to convince Dalglish to play Henderson on the right of midfield at Liverpool. It's all idle speculation of course, but whatever the reason I don't think the decision did the former Sunderland man any favours at all. Perhaps Comolli should've used a chance quality system to find a creative right-winger instead?
Recent posts you might like: Solid Foundations - Whether Liverpool finish in the top four or not this season, I think they're well set for another challenge next year, and here's why. PLCQ: Twenty Game Round Up – My project to determine the quality of chance that each Premier League team has and allows is just past half way, so here are the findings so far. Liverpool’s Chance Champion – A look at which Reds have created the most chances in 2013/14, and who has found the best areas most frequently. How Many League Goals Can Suárez Score in 2013/14? – This features a forecast table, which is updated after every match. LFC Pass Combination Heatmaps 2013/14 – A look at which players have been most involved pass-wise, and who they’ve linked up with in every league match this season. Please check out my other articles, and follow me on Twitter, Facebook or LinkedIn. Thanks.