Bias for bravery

A couple of months ago, noted Ralf Rangnick sceptic Michael Cox questioned whether data had made a dint on football’s on-pitch tactics.

I found this interesting. As I wrote shortly afterwards, maybe that’s right and, what’s more, maybe there’s a clear direction that cold, calculating data would push things.

In fact, maybe it already has.

I had a theory - and if you’ve taken enough notice of this post’s title, you’ll know it too - and I started listing things that played into it. More of the list than I’d expected were already on-pitch trends.

Across the board, I suspect that data analysis would reveal more risk-taking is better. Ian Graham, Liverpool’s former director of research, tells an anecdote in his recent book, How To Win The Premier League, about Brentford owner Matthew Benham. I used it in the previous piece - it’s about advising Brentford to Always Be Attacking - but I didn’t use the best line from it:

In other words, coaches are so affected by risk-avoidance that you need a caricature-level opinion just to drag their Overton window towards the optimal strategy. (Or so Benham believes).

After this had marinated in the back of my mind a little, my little grey brain cells made a link to goalkeeper cross-claiming. Analyst and data scientist John Harrison has previously published some work around shot prevention by claiming crosses: measuring how active and accurate claimers can save their team potential goals before shots even happen. And I remembered that David Raya (at data-savvy Arsenal) is particularly prolific in this department. (In fact, Harrison’s models have suggested that shot-prevention was a key part in Raya being an upgrade on Aaron Ramsdale’s performances in the men’s team).

Of course, we can’t talk ‘risk’ in goalkeepering without mentioning Manuel Neuer’s sweepering antics. Neuer didn’t invent keeper-sweepings by any means, but he’s a big reference point, and the position of goalkeepers in general play is a key point of analysis nowadays. Neuer was such an influence because of how successful he was.

Where else can teams play with more risk? High defensive lines, trying to keep the ball when under pressure, more take-ons, working the ball into better shooting locations, playing an attacking style even when leading, substituting as soon as necessary. (Any others?)

A lot of these are pretty familiar to modern football, even if analytics hasn’t driven take-up.

This doesn’t mean that these risk factors are a one-way street though. A push for high lines opens up a new risk-taking option for the in-possession teams: adventurous passes in behind the defensive line. But perhaps the highest line achievable is always the optimum.

At Amazon - a company even more successful than Matthew Benham’s Smartodds - they have a leadership principle called ‘bias for action’. The two beliefs which underpin this are that speed is extremely important to businesses, and that many decisions can be reversed if they turn out to be incorrect.

Risk avoidance serves an important purpose in business as in life as in football. But maybe the most effective path is to be poked a little bit more into the path of bravery.