A League 2 blog with an overwhelmingly pro-Torquay United bias
Next season’s fixtures were released yesterday and while it’s all jolly good fun seeing when clubs are playing their rivals, who they’ve got first and so on, I was wondering whether it was possible to extract any further information from them. Specifically, I was interested in whether I could get an early indication of how the season might pan out, using the relative difficulty of the fixtures to look for patterns.
Determining how ‘difficult’ individual fixtures are requires some way of comparing the strength of each club, but the transfer and loan market will be buzzing for some time yet, meaning that it’s still far too early to start ranking sides. However, I reasoned that I could have a stab at grouping the clubs into broad categories which give a rough indication of their prospects, which I did as follows:
Group A: Promotion contenders – Bristol R, Crawley, Dag & Red, Port Vale, Shrewsbury, Swindon
Group B: Play-off hopefuls – Accrington, Bradford, Gillingham, Northampton, Oxford, Rotherham
Group C: Mid-table makeweights – AFC Wimbledon, Aldershot, Crewe, Plymouth, Southend, Torquay
Group D: Relegation battlers – Barnet, Burton, Cheltenham, Hereford, Macclesfield, Morecambe
I’ve based these on last season’s placings, with a few very unscientific nudges where I’m aware that things have already changed. Please note that this is only a rough proxy done in a hurry, so don’t be offended if I’ve put your club in a group you don’t agree with. I’ll probably revisit this analysis nearer the start of the season when squads are settled and pre-season friendlies are underway.
Applying a rank to each group enables me to ‘score’ each club’s fixtures according to difficulty and produce the frankly hideous but hopefully straightforward diagram below (click to open a full sized version). You’ll see that I’ve ordered the clubs by their groups and colour-coded fixtures by group to make patterns easier to spot. In particular I’m interested in chains of tough (red) and ‘easy’ (green) matches:
Initially it looks like a jumbled mess, which is reassuring insofar as much as it evidences the randomness (and therefore fairness) of the fixture generators. However if you stare at it for long enough some patterns begin to emerge. For instance, in the top right you can see a fair bit of red, meaning that a lot of the promotion contenders are facing each other at the end of the season. This suggests that we’re likely to see some promotion or potentially title-deciding encounters in the final weeks. The last round of fixtures in particular looks like pitting the majority of clubs against others of roughly equal strength, so we could well be in for a few dramatic final day contests.
We can also read across a row to gauge the shape of that club’s season: let’s take Accrington (top row of ‘group B’) as an example. They start with a fairly mixed succession of games in August and September, so you’d expect them to have found their level after these first two months. In mid-October they begin a torrid run of fixtures in which they’ll face 5 promotion contenders in 7 games, so one would expect them to go into a bit of a tailspin here. Conversely, December and January look positively benign by comparison: at one point they face five ‘minnows’ in a row, so this should allow them to storm back up the table. Things then continue in a reasonably random fashion for the remainder of the Spring, but their last 3 games are all against ‘Group A’ heavy hitters, made all the more dangerous by their likely involvement in an end-of-season promotion scrap. If Stanley harbour any ambitions of repeating last season’s play-off placing then they’ll need to be safely entrenched in this zone by mid-April, as they’re unlikely to get much joy from their final few matches.
Anyway, you get the idea. This isn’t a particularly robust approach, and as soon as matches start getting rearranged due to cup ties and weather it’ll be rendered virtually useless.Further limitations include the arrogantly naive assumption that clubs can be lumped together in equal-sized ability groups (although we happily do this to schoolchildren) and that it doesn’t differentiate between home and away games, which can make a huge difference – just ask Crewe fans.
Self-deprecation aside, I hope you find this interesting. As always I’d welcome any comments or observations about how this could be further developed, as well as anything you’ve spotted in the data.