Much has been made of the apparent inflated price Liverpool have paid for Jordan Henderson, believed to be in the region of £16m.
Let us step away from the price tag for a moment, particularly as he has yet to kick a ball for his new club club, and consider the most telling statistic about Henderson’s 2010/11 season, that of his chance creation (tweets from OptaJoe here and here):
Only four players created more goalscoring chances in the 2010/11 PL season than Sunderland’s Jordan Henderson (82).
The four to create more scoring chances than Henderson were Malouda (117), K Davies (91), Brunt (86) and Downing (85).
Chance creation gives us a much better idea of a player’s contribution to his side than assists, for the simple reason that it’s not dependent on attacking competency. Season after season, assist charts are dominated by players from top teams, since they’ve had the likes of van Nistelrooy, Henry, Drogba and Ronaldo to ‘complete’ the assist.
Chance creation – a pass that leads to a shot on goal – is hardly a perfect statistic in itself (what about dangerous crosses that finds no one?), but it is much better than sight-based biases in player assessment.
Is creating an abundance of chances a formula for success, and worth the millions that Liverpool have spent? On first glance, perhaps not (click to enlarge):
Whilst there is a downward trend in chances creation with respect to league position, there appear to be plenty of exceptions, namely Man City, West Brom, Blackpool and West Ham.
Clearly, creating chances in itself is not a prerequisite for success, if relegated West Ham can tally only 8 fewer chances than champions Manchester United all season.
The issue returns to attacking competency. If two teams have equally good attackers, does the team that creates more chances collect more points over the course of the season?
To test this, a model with a measure for attacking ability needs to be formed. I’ve chosen to use chance conversation rates; that is total assisted goals over total chances created. This ranges from 14.5% with Manchester United to 5.8% with West Ham.
Holding constant attacking ability, along with other possession and defence-based variables, chance creation has a significant effect on total points accumulated over the season. A 1% increase in chances created leads to approximately a 0.4% increase in points accumulated over the season.
The brains behind Liverpool’s transfer policy, Damien Comolli, will have an even better idea of the returns on chance creation. He is of course good friends with Billy Beane, the man whose revolutionary method of valuing baseball players has now been immortalised in film after Michael Lewis’ bestselling book Moneyball. Simon Kuper’s recent FT column reveals how data drove Liverpool’s purchases of Andy Carroll and Luis Suarez; to call the article ‘fascinating’ would do it a disservice.
Of course, ‘adding’ chances to Liverpool won’t yield returns unless those other variables – attacking competency and passing and defensive ability – are equal to that of their rivals. Over the course of the 2010/11 season, they fell short of the top four in all these aspects, but the numbers are skewed by Roy Hodgson’s tenure at the club.
Comolli and Dalglish must believe that in Carroll and Suarez they have a partnership as clinical as any other in the division, and that the team can also hold their own in other aspects of the pitch. If not, then the additions of Jordan Henderson, and potentially Stuart Downing and Charlie Adam, may well be money foolishly spent.
It may be a simple and obvious message – creating more chances for good strikers leads to a higher league position – but it’s an important element to Henderson’s game that has been forgotten in amongst debate over his transfer fee.
This post, along with other articles by the author, is available on 5 Added Minutes.
Very good post mate interesting take on the chance creation throughout the league. Especially liked the bar graph. Very informative and tells you that West Ham created more chances that Liverpool FC and still went down!
Yes essentially the idea is that we shouldn’t be creating chances willy-nilly, we need to have the personnel. It sounds simple but West Ham’s demise shows it isn’t a simple numbers game.
Great point and well shown in the stats.
Good article. I was having an argument with someone recently who thought Maxi was rubbish because he had a low number of assists.
I too said that Maxi had a high percentage of chance creation which is a bigger contribution than assists to a team and it appears based on your arguments I was right. Nice. Keep it up!
Thanks – and yes exactly. Perhaps Maxi has played a number of games alongside Ngog and not Torres/Suarez (don’t know the numbers), and that’s an important consideration.
Great example is Leighton Baines, who made 11 assists this season but only 1 in December/January – Cahill was away from the entire month of January on international duty.
good points here but surely stating the obvious, it doesn`t matter how many chances you create in a game if you don`t put em in the onion bag. How many managers of relegated teams continually say (after another defeat) “at least we are creating chances time to worry is when we don`t.”
Not sure that’s what I’m totally getting at – the chart suggested that it doesn’t really matter, but when considering other factors it turns out it *does* matter how many chances you make, given equal quality strikers with rivals.
Admittedly that last part is a little stating the obvious, but Comolli will know that had Liverpool created in the region of 50-60 more chances over the season (according to my very rough model), they’d have finished ahead of Spurs. Knowing the value of each chance becomes imperative to player valuation.
Cracking post, but felt like I wanted to know more. Would love a follow up or an addition to this.
Looks like more work for Mr Chaudhuri!
Cheers, will have a think as to where I can go with this 🙂
So have you had a thought? Would be good to get a follow up part! 🙂
Will perhaps look at chance creation exclusively from set pieces in the wake of the purchase of Charlie Adam, and its value compared to open play chances. More than open to suggestions though.
A bit with accuracy/chances created from set-pieces would be interesting, with quite a few regular takers in the squad now.