Huddersfield Town, Crystal Palace, Swansea City and Stoke City
I want to talk about three teams this week. These three are all bottom of the list for the difference between the rate at which they convert their shots to goals and the rate that their opponents do.
Anything between around -3% to 3% is just stuff you shrug your shoulders at and say it doesn’t mean much, but anything outside these boundaries – either hot or cold streaks – begins to look like either significant luck or significant quality (or a lack of either).
After 28 games, Stoke City have a conversion difference of -4.4%, Huddersfield has one of -5%, and sad, sad Crystal Palace’s is -6%. So, if all of these clubs were scoring every 1 in 10 (about league average), their opponents would be scoring 1 in 6 and a half, or thereabouts.
It’s worth noting that this conversion rate difference lies on top of the shots that teams take and concede – Huddersfield and Stoke’s shot stats put them legitimately in the clump of teams who could be fighting relegation while Crystal Palace’s are fairly decent.
The former both get outshot in terms of shots on target by about 1.5 per game (well, Stoke by 1.7, Huddersfield by 1.3) while Palace is outshot by 1.
But if you look at a concept called ‘optimal shots’ – non-headed shots in the center of the box, the most dangerous area of the pitch – Palace come out in the black, behind only the top 6, while Stoke and Huddersfield are the third and fourth worst teams in the league.
The graph of Palace’s rolling conversion difference tells the story of their season.
These graphs – unscientifically – began life as a way to spot teams who might be riding a high or a low in the past few games and would be expected to revert back to the mean soon and therefore make a decent story.
So, in the course of 3 games, anything between -10% and 10% as a conversion difference average is nothing much. You can see how all of the lines waver about there. The peaks and troughs are interesting, and Palace took an age to get up to normality. Their conversion difference since Game 15 is just -2.4%, definitely within normal levels.
But Stoke and Huddersfield are a much different story:
Stoke are just sad
They’ve rarely had a run of three games where things have gone their way, which is a killer not just in terms of the cold points, but also in the momentum that good runs can build (not wanting to get too focused on the intangibles).
Take Swansea, a poor shot side (and showing it with their league position), but their season-long conversion difference is (after a new manager bump) -1.4%.
Even before Carvaljal, Swansea had a couple of early runs where a couple of results saw them get the rub of the green in terms of their and their opponents’ shot conversion, before a sustained period of scoring at a slower rate than their opposition. Note: this is likely because they are a bad team.
Also, note how the recent dip in rolling conversion difference coincides with people commenting that the shine is beginning to come off Carvaljal’s reign. The graph, it works.
But, finally, we turn to Huddersfield, who these pages have covered before.
Almost the complete opposite to Swansea’s sustained period of slightly-negative conversion difference, Huddersfield’s season seems to have operated solely in extreme swings, from a magnificent start to numerous horrendous troughs.
What does this mean for the rest of the season?
Who knows, to be honest. But I have a theory, which I haven’t yet been able to investigate, that perhaps spikes and troughs in conversion differences, match-to-match, aren’t quite the same in value to a consistently middling, or low, level of conversion difference.
Statsbomb’s Ted Knutson has spoken before about the value of wins for teams at the bottom of the table (as opposed to playing for a draw), and maybe it’s the same with the luck of the green. You want the three points one way or another, right?
Discover an easy way to bet on Premier League, with the best prices and massive limits! Check out our hassle-free Skype Betting.