Expected goals

[EIM]

I'm not riled. I just think it's a load of wank.

[imp]

I'm riled and also think it's a load of wank. Loads of wank rile me - it's just the way I roll.

[MrLeam]

Ok, rather begs the question of why you think it is a "load of wank"

I might be wrong, but I'm guessing its primarily because you're not interested in that type of thinking/analysis of football. That's absolutely fine - people should be totally free to like football in whatever way they choose. That's kinda my point though, in that the anti-expected goals stuff comes across as some purist crap about it adding nothing to the game, being job creation, about people wanking about it over their pints etc when it should be about letting people who find it adds to their understanding and enjoyment do that.

Of course, your dislike could be for other reasons. You might just think the statistics are actually incorrect (in which case tht's a different discussion).

[Ray de Galles]

I can’t say I’m very interested in (or even understand) this Expected Goals thing but I think Mr Leam has a good point.

The amount of people chuntering on about how it’s the final death knell of football is ridiculous. It appears to be a minor thing that is very easily ignorable if you’re not in to it and football has approximately 837 things more worth getting angry about.

[DCI Harry Batt]

It isn't the final death knell. That would be a stupid claim. Not one I've seen made, mind. But if anyone did say it, that would be stupid.

[Ray de Galles]

That’s how your previous post on the thread reads. You can add substitute “as if” for “how” in front of “it’s the final death knell of football” if you want. That’s how you’re carrying on.

[DCI Harry Batt]

I guess, unless you read the words in it, or something, maybe.

[Ray de Galles]

See my edit, written before your subsequent post.

[DCI Harry Batt]

I'm already done with football, I'm sad it's dead, but since it is you are welcome to this shitty xg crap. How on earth you can read that as "the final death knell", I have no idea. But just to be clear, once again, I'm happy to say that it isn't.

[imp]

It doesn't add a thing to my enjoyment or understanding of the game. Just like possession stats. Just like 'packing' stats. Just like almost all stats. I'd rather be hearing or reading about something else. It's superfluous minutiae for nerds, and part of a long-running trend to pretend that football is a science. I can ignore it until people start talking about it like it's something important, then it's annoying.

[Nefertiti2]

we understand you don't like it. Other people do.

[ad hoc]

I haven't read anything about it, and I won't because it just sounds so annoying, but is it basically "chances"? Like you used to get commentators talking about chances and half chances? So now some bloke in his bedroom (and it must have been a bloke, and he almost certainly was in his bedroom - either in his mum's house or in his University halls of residence), has thought to himself "What exactly is a half chance?" Why don't we have a quarter chance? Or a 5/16th chance? I know I'll put together an algorithm to actually work out what fraction of a chance each chance actually was."

And here we are. That about it?

[Janik]

It's trying to be a bit more rigorous than that, but essentially yes.

It all stems from much the same place as the analytics Brentford are apparently using for player recruitment (it wouldn't surprise at all to hear they use this one, in fact), which is based on the idea that the people's instinctive 'feel' of the game includes various inbuilt biases, rules of thumb and just plain inaccuracies. Much like Baseball had before Moneyball came along. Which was also roundly dismissed with the American equivalent of 'a load of wank' when it first appeared, and is now such standard practice that no-one thinks to mention its existence. Everyone uses Moneyball these days.

Moneyball was indeed invented by a mathematically minded bloke sitting in his bedroom. So even if this particular method of analysis wasn't created like that, one could still make the claim based on the inspiration.

[DCI Harry Batt]

I can ignore it until people start talking about it like it's something important, then it's annoying.

Indeed. The idea that this stuff is ignorable is bizarre.

[ad hoc]

. Everyone uses Moneyball these days.

Except fans. Fans don't sit there and calculate moneyball stats, they watch and enjoy (or not if they're losing) the game.

[Janik]

There is no need. Once (if) it's established, it becomes a coaching/team development aid. Some people might like to know how their team is going about selecting which players to sign or include on the teamsheet, but that is not a concern for everyone anymore than it was prior to the existence of the stat.

As with Moneyball, it will either prove useful to people in the industry, or it will swiftly wither and die.

[Rogin the Armchair fan]

The reason it's not so good in football as in some other sports is that football not only has too many variables but is also a much more fluid team sport. Moneyball worked in baseball because baseball is essentially one on one - pitcher v batter. Analysis of particular player stats in football is much more problematic. For example in goals per 90 minutes played Peter Crouch is the best striker in the league this season.

[imp]

Everyone uses Moneyball these days.

So is everyone successful? Or, less facetiously, has the game become more competitive as a result? Has the overall quality of baseball improved?

Also, as ad hoc points out, if this is something genuinely useful to coaches, then good luck to them. The Germans have been deep-analysing stats for the national team for 15 years now. But for me to appreciate a good performance by Germany, I don't need to know the details of all their homework.

[Snake Plissken]

The other problem is that the stats stuff is in its infancy, so not only is it not explained well, then it isn't as accurate as it needs to be. There was a decent article in the NY Times about Burnley and how they defy the xG stat. Basically, they give away chances that score high xG but the defence is in positions which block all possible routes for that shot to goal. The reason I started wondering about whether it was deliberate was because I'd got used to Tom Heaton being almost prenaturally able to stand in the one position where an opposing forward heading a comparatively open opportunity. Then Nick Pope started doing it as well and I figured there had to be something more to it.

This discussion is much like the discussion in (ice) hockey. The argument goes that the game is too fluid, too random for stats. The stats nerds seem to be trying to impose significance on moments of chance, instinct or skill. I understand that baseball and cricket are games that offer discrete moments that can be statistically analysed but for football or hockey, my preference is to just leave it be.

[ursus arctos]

Given the sheer mindlessness of the large majority of what passes for “football commentary” in the UK, the idea that analytics are ruining anyone’s enjoyment of the sport is rather risible.

[Snake Plissken]

I've been crying out for some proper tactical analysis. But this isn't it. Basically it just gives the same old tired voices the ability to go "well, dey should be doin more wiv that levul of posseshun".

[cantagalo]

It's trying to be a bit more rigorous than that, but essentially yes.

It all stems from much the same place as the analytics Brentford are apparently using for player recruitment (it wouldn't surprise at all to hear they use this one, in fact), which is based on the idea that the people's instinctive 'feel' of the game includes various inbuilt biases, rules of thumb and just plain inaccuracies. Much like Baseball had before Moneyball came along. Which was also roundly dismissed with the American equivalent of 'a load of wank' when it first appeared, and is now such standard practice that no-one thinks to mention its existence. Everyone uses Moneyball these days.

Yes indeed, but some like Brentford and Brighton use it better than others such as Fulham.

Personally, I find the xG stats an interesting discussion point and a useful tool for analysis. Also anything that pisses off the old-school ‘load of wank’ pundits has to be a good thing. The positive stats were one reason why no-one at Brentford panicked when we failed to win any of our first eight games this season. We knew it would come good eventually.

[ursus arctos]

All of the analytics types I know would tell you that possession is not at all useful metric.

In fact, possessions' inadequacy as an analytical tool was a major impetus in the development of xG (which no one thinks is perfect or even fully formed).

Or, less facetiously, has the game become more competitive as a result? Has the overall quality of baseball improved?

The answer to the first question is undoubtedly yes; the answer to the second is a question of personal aesthetics.

[treibeis]

Just like 'packing' stats.

When I play, I rarely give the ball away. On the other hand, I rarely play a forward pass. So my passing rate is nigh-on 100%, whereas my 'packing' rate is as near to zero as makes no difference. So, statistically, I'm simultaneously brilliant and shit. I think.

[blameless]

It should be noted that the gridiron team who have bought most heavily into the Moneyball ethos are the Cleveland Browns, who have won exactly once in their last 30 games and are considered a byword for how not to do things in professional sports.

It's like taking a cricket statistic and expecting it to use it to create a successful football or rugby team.