clock menu more-arrow no yes

Filed under:

Don’t abuse Westwood and Gardner with wrong stats

New, comment

Just because it fits what you believe, doesn't mean it's true

Cardiff City v Aston Villa - Sky Bet Championship Photo by Athena Pictures/Getty Images

Villa slumped to a bad defeat away at Cardiff and fans were understandably angry. The players most criticised were Ashley Westwood and Gary Gardner, both taken off at half-time.

In the aftermath of the match, an astonishing stat about the the midfield pairing’s first-half performance began to circulate:

The statistics seemed absolutely incredible. In fact, a few who were at the game said it didn't fit with what they'd seen.

But it fit with Westwood and Gardner having a bad performance, and the figures came from WhoScored.com, which is normally a pretty good source of statistics, so it spread very quickly across Twitter and then beyond - the Birmingham Mail did an article on it, which has now been removed.

Well if it seems incredible, maybe it's because it's not true. The stats looked wrong because they were wrong, and they've now been replaced by new ones at WhoScored's summary of the match.

See the full stats here - https://www.whoscored.com/Matches/1085112/LiveStatistics/England-Championship-2016-2017-Cardiff-Aston-Villa

So Gardner had 41 touches, Westwood had 30 touches for a combined 71, rather than 9.

I'm making no argument about Gardner and Westwood's performance here. By all accounts the pair were poor and here at 7500toHolte we've talked a number of times about the need to strengthen Villa's midfield.

But if we're going to use statistics to assess players, we have to engage with them critically. In this case the stats were just plain wrong but seemed to confirm what everyone already knew, so were spread quickly.

This is the danger of lazy statistics use - such and such a player is good or bad because of this one stat, which can be inaccurate, or not tell the whole story, or be from just one match which is not enough to draw conclusions.

Real statistics use in football is still at an early stage and I absolutely encourage anyone interested in this stuff to delve into the world of football analytics - start over at Statsbomb.com and branch out from there - but let's not blindly repeat something just because it fits what we already think.