Those of you who don’t like it when I go digging in the numbers may care to leave now…
As you know, I draw “inspiration” from far and wide. Today we’re going to borrow from ice hockey!
This article had some interesting quotes to draw me in:
However, it is only one month. The more important issue, whether your team has been good or bad, is: is this really my team? Is this the level of play I can expect for 82 games, in addition to the playoffs?
and then:
shot differentials, which contain much more data (there are 11 times more shots than goals in an NHL game), are more precise and therefore better represent what a team will be able to accomplish in the long run. Therefore, to determine if your team has been good or merely lucky, look at the shot differentials.
Aha. So I worked it all out. First, acquaint yourselves again with the league table.

My usual flannel about Goal Difference applies, but what we’re proposing to do here takes things to another dimension. By looking at shots and shots on target, both by each team and against each team, we get to see who’s really doing what. As the quote says, goals are what it’s all about, but are rare enough that they can distort things and give a false representation of how good a team is, and how well it is playing. The underlying metrics – if I may be so bold – can also be helpful.
Here goes:

Where to start? In short, the left most numbers are what each team has done while attacking in home games, away games, and in total, per game, then ranked. The second lot are what each team has had done to it, in terms of Shots and Shots on target.
So you can see for us that we’re much more menacing at home (63 shots, and 38 on target) than away (39 shots, 18 on target), we have put 102 shots on goal in total and 56 on target, averaging 10 shots and 6 on target per game. We rank 15th and 16th for these respectively.
Bear with me, I think it’s worth the effort.
The same explanations apply to the shots we’ve allowed in the next set of numbers.
But the key to all this is the per game totals. We have 10 shots (6 on target) per game and allow 12 shots (5 on target). This leads us to differentials, which is the number we want: we’re -2 for shots (we allow 2 more shots than we take) but +1 for shots on target (we get one more shot on target per game than we allow).
Where this gets really interesting is when you compare these number to the rest of the league:
- Chelsea are ominously good this year. It still beggars belief that people complained when we didn’t get near them. Yes, we might have shown more fight, but I’m fairly sure this is the best team in the league. The table, looking at overall differentials, is fairly bunched. Except Chelsea, who tower above the rest of us in terms of the shots they take and the shots they stop others taking. That there’s a winning formula.
- Man Utd aren’t too far behind, but while Chelsea have the best attack and the best defence, United hae the 3/4th best attack and the 2nd best defence. Close, but close enough?
- Were it not for the media I think people would be happy to acknowledge that Liverpool are a pretty good side in a bad run, rather than a bad side in need of reconstruction (their injury list is not inconsiderable; their reserves are not great). If they don’t panic things should go their way in time. That doesn’t mean they can recover this season, but next season really should be business as usual.
- Everton’s “indicators” suggest that things are broadly on track for them too, despite results so far.
- City and Tottenham, as last year, are there or thereabouts, but not there or thereabouts enough. Mark Hughes has clearly tightened City up (look at the shots v shots allowed – both lower than we’d expect), but now they need to push on to the next level. Indicators and recent performances suggest they’re not there yet. Same for Spurs.
- Wolves have made a decent fist of things so far.
- Villa play a strongly counterattacking game. Remember when we played them? Greening had 90 passes, about three times what he might normally expect. They clearly invite teams onto them and then kill them on the break, which makes them anomalous in this analysis.
- Portsmouth could be the opposite, attacking teams recklessly and being hit on the counter. Or they really have been unlucky – in game after game they seem to have got less than their performances deserve. Possibly a bit of both. The problem here is that if we start making exceptions for teams on style of play grounds we have to apply this to everyone, so probably best to say that this may be a factor but shouldn’t cast too much of a shadow over findings. But it may explain why Villa and Portsmouth diverge from what we’d expect to this extent.
- Blackburn, Bolton, Burnley, Hull and Stoke may be in trouble. Burnley we figure are overachieving in a Hull last year type way, but Stoke are surprising, being halfway up the league with a fair goal difference. But they are getting only four shots a game on target, while allowing nine at the other end. That’s a dangerous game. I suppose in their favour they are keeping things tight – only Fulham have fewer shots in their games – which allow for points nicking potential, but still. They look like an accident waiting to happen.
- Fulham… doing quite well, thank you. Our famed defence is only allowing 5 shots on target a game, a tribute to our midfield shape and defensive outlook. Against that, we’re only get 6 shots on target ourselves. Plus ca change, etc, but again, we are keeping things very tight, which is why we rarely seem to get turned over. Also worth noting the change (not shown here) in our early games and more recent matches – we seem a lot more open lately.
If I had to simplify this I’d divide the league into:
The top sides: Chelsea, Man Utd and Arsenal, with Liverpool off the pace but still better than anyone else
Other teams playing well: Everton, Portsmouth, Fulham, City, Spurs, Wolves (and let’s add Villa as noted above)
Bubbling under: West Ham, Sunderland, Wigan, Birmingham
In trouble: Blackburn, Bolton, Burnley, Hull, Stoke
As per the ice hockey article, I like how this conforms to standard pre-season expectations. Those teams in trouble are exactly the five you’d suggest, with Wolves outperforming expectations and Portsmouth not yet doomed by ownership issues.