For many years, I watched League One football. I grew to love its quirks, its oddities, its annual visit by Oldham Athletic, even the way it acts as high and low watermark for different levels of clubs. Leeds, Norwich and Aston Villa could never go lower, one suspects, and have used it as a springboard, while Stevenage and Rushden and Diamonds would be unlikely to get higher, and see it as an achievement in and of itself. I’m a big fan and, even though Huddersfield Town have escaped its zombie-like clutches, it remains my favourite league and, to me, probably the most interesting English league.
This might well end up being a long one, as I compare Bradford and Brentford, whose twinned fortunes have been intriguing – after 11 games, Bradford led the Bees by six points – some seven games further on, that figure is reversed, with the Londoners two wins clear of the Bantams. I was tempted to pair Wolves and Peterborough, but that might wait for a while, and this can be incorporated within it. Anyway, bear with me. You might want to come back to this.
If we start by looking at some very basic metrics to get a handle on the differing fortunes between the two sides, I’ll start with a bit of Points Per Game work. It will tell us the story of the season so far, from which any conclusions have to be drawn, of course.
Here’s the whole season in PPG format, and it illustrates the change of positions in recent weeks after both had a fairly positive start. There’s reasons for that, which we’ll get to later. Its more obvious when that’s taken to a 6-game form PPG; Bradford’s low ebb sinks far below Brentford’s who, even when not winning games, tend not to be losing them.
Obviously, we’re at the peak of a Brentford run at the moment (they’ve won five from six, most recently against Peterborough) whereas Bradford have been spluttering a bit, as you can see.
The elephant in Bradford’s room is the influence of Nahki Wells. Undoubtedly a predatory striker, the Bermudan has 40 goals in 60 starts (26 sub appearances). However, he missed a spell of four games for the Bantams after suffering an injury in a challenge with Shrewsbury’s Dave Winfield. That run started in late September, which coincided with a dip in form for his team, as illustrated below. Bradford still don’t seem to be quite at the same level since his return, but are certainly better than in his absence.
So, missing your best player makes you win fewer games. Who knew?
The other approach I want to bring to this is the tale of the two side’s opposition. By 46 games into the season, you’ve played every team ranked 1-24 twice, and, while the table doesn’t lie, you might well have played them at different times during the season.
To wit, I’ve grouped the eighteen games so far in sixes, and produced a little ‘opposition position at time of game v. points per game’ graph, which should illustrate (very roughly) whether either team has been particularly good or bad against teams who are playing well. It isn’t an exact rubric, by any means, but is at least indicative.
From that very basic graph, it looks as though Brentford had the tougher start, but have found their feet against slightly easier opposition – there is very little to choose from the two teams, though, in that sense, that isn’t evidenced by the form we saw earlier. So what is going on?
End of Part One
What IS going on, I asked at the end of part one. Well, hopefully this will go some of the way to explaining it.
Firstly, if I want to elaborate on the point of Bradford’s missing Nahki Wells, its worth looking at the accuracy and clinicality of both teams’ attacks. These are fairly common things to do, but do give an idea of how teams go forward.
Specifically to the Nahki Wells issue, I’ve taken the liberty of including individual match figures for Bradford in faded claret and amber behind the two cumulative lines; the claret bars being the games Wells featured in, and the amber ones being the ones he didn’t. That’s fairly revelatory, isn’t it, particularly in the clinicality graph (below) – Phil Parkinson’s side are obviously pretty reliant on Wells for goals and don’t produce efforts of particularly high quality when he doesn’t play; no shock when considering his strike partner is often James Hanson, who is more battering ram than fox-in-the box (not that there’s anything wrong with that, but it doesn’t guarantee goals).
The two sides look fairly evenly matched between themselves in those metrics – though Bradford are falling, a little, in the latter (as Wells’ red-hot form from the early season looks to have faded). That’s all well and good but we can go a little further.
If you’re familiar with the work of Statsbomb (and if you’re not, you should be) you’ll be aware of the ShDom statistic for measuring teams’ efficiency and effectiveness. I hope I’m not completely
mis-representing the fine work of Ted Knutson (find him as @MixedKnuts on Twitter), but that measures teams relative strengths across games within the season.
I’m on shaky ground here, but my information is good, so it should all pan out.
So, then, if you’re not familiar with it, it works as the shots a team takes divided by those that are taken against them. Simple enough in practice, but because shots are the main way of scoring goals, illustrates the likelihood of winning games (at admittedly a basic level given the small sample size of, at this stage, 18 games) for each team given their general performance.
As such, the value of ShDom is based around the number 1. 1 would represent a team having exactly the same amount of shots as their opponents, and, as a result, the bigger the number the better. All indications are that it is reasonably accurate, and gets moreso as time goes on. The freak results end up being eaten up by the logic of size. So how does that graph look, then?
That’s interesting, isn’t it? Brentford sit around 1.8, with Bradford only a little over 1. That would suggest, to me at least, that this current swing in position is likely to be maintained (and further suggests it might be worth investigating Wolves and Peterborough) as, with Bradford’s ShDom figure being so consistently low, they’ve been propped up with the goals of Nahki Wells and, if his ratio drops and – he’s in a bit of a dry spell at the moment – indications are that it might, they’ll fall away towards midtable.
There’s a little more to do with this as some figures leapt out at me as I was working it out. There’s a stark drop-off between Bradford’s home ShDom and their away tally. Brentford, obviously, are better at home than away though their 1.54 away figure is decent enough, to see Bradford drop below 1 (to 0.80) on their travels in not positive news.
What can I conclude from all this, then? Nothing I wasn’t half-certain of before, to be honest. Bradford’s start flattered them, as they’re a decent side bolstered by Nahki Wells. Brentford took a while to get going, but now they are, will take some stopping.
Addendum: For added indication of Bradford’s reliance on Wells, he’s taken 15.3% of their total shots, yet managed to accrue 33.3% of their shots on target (having missed 4/18 games so far, that’s a massive amount).
*Bantams, of course, are birds. I needed a catchy title.