Big Sam's cross-heavy West Ham and the drawbacks of stat-driven tactics

Though it’s banal to write out, it’s worth repeating: advanced statistics in football, or indeed in any sport, have a wide number of uses and applications, much wider than some fans and pundits assume.

For example, you can use data analysis to aid in player recruitment. You can use it as a diagnostic tool to see beyond scorelines and subjective impressions how your team is really doing. Or, you can go whole hog and use them to change the way your team plays. Finally, you can use all of these approaches simultaneously.

The key, however, is to maintain perspective, to see the game as a complex whole in which correlation may not necessarily imply causation, and to remember that focusing too much on one seemingly positive element of the game may cost you in ways you cannot foresee.  

With that in mind, 2013-14’s West Ham United offers an interesting case study, particularly in light of today’s caveat-ridden “backing” by club owners of manager Sam Allardyce in which they demanded “more entertainment” for next season.

Big Sam’s Love of Clean Sheets and Crosses

Before we go on, a warning: most of what we know about club analytics is through a glass darkly, and West Ham are no different. While there is some interesting information on the public record over Allardyce’s tactical preferences, much of this information is vague and incomplete.

However, we can glean some clues. This Daily Mail profile from earlier in the season quotes the club’s head of performance analysis David Woodfine:

'We've got benchmarks and one that we do look at is knockdowns,' adds Woodfine. 'Every time we play the ball forward to the front, we're looking at whether we win the flick-on, we win the ball that gets headed down (the second ball) because it's a great opportunity to gain possession in the opposition's half so we're constantly looking at how well we do that.

Knockdowns are often the result of crosses, and this is another feature of West Ham under Allardyce which Michael Cox noted for ESPNFC earlier in the season. While noting that crosses tend to be among the least efficient ways to score goals, Cox points out that Allardyce specifically bought players with the ability to make the most of those “second balls” or knock downs, particularly Kevin Nolan:

Nolan's shot conversion rate was actually quite low last season, but by having so many attempts per game, from such a close range, he's seemingly guaranteed a few goals. At Bolton, Allardyce tried to find the most frequent positions of the 'second ball' - and Nolan, one of his players from those days, remains a master of it.

Finally, we also know Allardyce values clean sheets, as he relayed on Sky Sports earlier this season.

The Results for the 2013-14 Season

First, all numbers below are cited from whoscored.com, Michael Caley’s EPL advanced stats, Ben Pugsley’s Sortable Tables, and James Grayson’s end of season data dump. For an advanced stat primer, please read this.

West Ham finished 13th in the league on 40 points, with 40 goals scored (one less than Sunderland) and 51 conceded (same as Spurs).

The club recorded a Total Shots Ratio of .382, the second worst in the Premier League. This means West Ham were outshot most of the time, in a way that in most circumstances would have seen them relegated. That’s generally very poor, but perhaps we should view this ratio in context.

For one, West Ham enjoyed a 1089 PDO, with a 32.79% shooting percentage and 75.36% save percentage. These numbers, particularly the save percentage, probably saved them (they tied for 5th highest total clean sheets in the league at 14) but they tend to be more driven by luck than talent.

That luck showed up in other ways, too: West Ham’s opponents got the second lowest percentage of total shots on target in the league, a figure which varies a lot from year to year. Moreover, their opponents scored ten fewer goals against them than Expected Goal stats indicated (49 to 59). It would be tempting to reverse engineer some cause for these numbers, but they do tend to vary a lot from season to season.

As for Allardyce’s tactics, If Cox is right about his emphasis on more chances in the dangerous areas from crosses, we should see that reflected in the stats. Indeed, West Ham averaged 25 crosses per game, second highest in the league (Moyes’ Man United was first, of course). They were also third highest in the league in percentage of crosses leading to shots from the “danger zone”, the area directly front of goal with the highest average conversion rate.

Finally, West Ham also enjoyed the second highest percentage of total shots from the “danger zone” in the Premier League. There is good evidence the club relied on their Downing/Jarvis/Carroll/Nolan dynamic, with the duo sending in crosses to ping off Carroll’s head to Nolan’s feet, but perhaps to a fault.

Verdict

West Ham’s strategy “worked” insofar as the side did create more dangerous chances per total shots. But it came at the cost of the goals the club might have scored had they otherwise employed a more mixed approach in attack.

Analyst Ted Knutson put it well on his old blog last year he wrote, “I honestly don’t know why any smart manager, and especially an analytically savvy one, would choose [a cross-heavy approach] for their team.”

West Ham’s analysis isn’t necessarily “wrong,” in that the team really is generating good chances and putting them away. But it seems to come at the cost of build up play, through balls, longer range shots, generally the big tent approach most football clubs employ to score goals.

Using advanced stats to craft a particular tactical approach isn’t necessarily a bad thing, but perhaps more emphasis should be put on possible opportunity costs.