PDA

View Full Version : Stats



Arch Stanton
26-07-2011, 02:49 PM
I reckon Hibs played pretty well for a good part of the match on Sunday.

Now, if anyone is sitting there thinking of telling me that the stats prove me wrong, then they can save themselves the bother, because they would be wrong.

If anyone judged that Hibs did badly then I'd sure be interested in how they came to that view - it's just people gong on about the stats that bother me. The stats certainly show Celtic were the better team, but that's about it.

The thing is, before we had TV pundits we never had "stats" like "possession" and "shots on target" (or more correctly, data that could be used statistically, if one felt in a pedantic frame of mind).

But because the pundits needed stuff to drivel on about and because statisticians like to count things they duly found things to count. And given how common 0-0 HT scores are, just counting the goals didn't really do it for them - hence the "shots on target" and the like - things that might show one team was performing better than the other - and to be fair this often helps make that comparison.

Without detailing all the problems with stats, here are some scenarios in ascending order (note!) according to the way stats are gathered.

1) Great cross in and the striker coming flying in just misses out on heading the ball into the net by millimetres.
2) Decent enough cross comes in and it hits the striker on the back of the head and goes flying over the bar.
3) Speculative cross come shot bounces off the strikers shoulder and then harmlessly into the goalkeepers arms.
4) Totally wild cross nowhere near any strikers comes off a defenders leg and into the goal.

I'm no statistician but I would say that as a general rule, the smaller a number is, the less you should think about quoting it as a stat.