Football League fixture lattices
You may remember from last year that I developed a simple algorithm to visualise each team’s fixture list for the coming season, which I recently refreshed. It works like this:
- Quantify each club’s relative strength using a simple proxy measure: in this instance I’ve chosen Bet365′s current promotion odds (appended at the foot of this post). Even if I don’t fully agree with them, ‘market forces’ should be less controversial than my own opinion.
- Factor in how home advantage affects each team. I do this by taking the ratio of points earned at home vs. away last season.
- Combine these two numbers to assess the relative difficulty of each fixture for each team. Each team’s strength calculated in (1) is obviously more significant and hence is more heavily weighted than the home vs. away rating calculated in (2).
- Rank each team’s fixtures in order of difficulty and colour code them, allowing fans to pinpoint particularly winnable or tough spells for their team.
However I was concerned about the readability of the original heat maps, which as well as being a bit garish also relied on people being able to discern subtle shades of colour between red, yellow and green. Where the graphics in this post differ from the original heat maps is that I’ve instead grouped each team’s fixtures by relative difficulty as follows (these can also be found in the key for each image, so scroll past this if you’re in a hurry):
- Their easiest 5 matches are shown as large dark green circles;
- The next 10 easiest matches are shown as smaller, lighter green circles, giving a total of 15 ‘green’ matches (as close as I could get to one third of their league matches);
- The next 16 matches are judged to be of average difficulty, and shaded in even smaller grey circles;
- Mirroring the approach for the easier games, the next 10 most difficult matches are in slightly larger light red circles;
- This just leaves 5 of the 46 remaining fixtures – the 5 toughest – which if you haven’t already guessed are denoted by large dark red circles. This gives a total of 15 ‘red’ games to mirror the 15 green ones.
Without further ado, here are the visualisations for each division, each of which can be clicked to load a full-sized version in a separate tab.
To give a few quick examples:
- Middlesbrough have an easy start, with 4 of their first 5 games being ‘green’, including 2 of their top 5 most winnable. However Brighton‘s is even easier: they don’t have a ‘red’ match until mid-October.
- By comparison, Barnsley have a horrible start, with 3 of their toughest matches in their opening 6 games, although at least this gets them out of the way early. Leeds also face a tough beginning to the season, with 4 ‘red’ fixtures in their opening 5.
- Burnley may find themselves slipping down the table as the season draws to a close, with 6 of their final 10 games being ‘red’.
Again a few illustrative examples:
- Bournemouth and Walsall have horrible starts, having to wait until their 9th and 10th games respectively for a relatively easy encounter.
- Brentford on the other hand have little excuse for not sitting near the summit at the start of October, as 8 of their first 10 games – including all of their first 6 – are green.
- Bury on the other hand could expect a late surge up the table, as their last 6 games are all green, including 2 of their 5 easiest.
As usual, a few sample observations:
- Wycombe have a rollercoaster run-in. Their 11 matches from late January until late March contain 7 ‘red’ games (including 2 of their toughest 5) and only 1 ‘green’, but then 5 of their final 6 fixtures are green. They could well start fall away, seemingly out of play-off contention, only to surge strongly upwards as the season nears its climax.
- Barnet and Burton have testing starts, with only 1 ‘green’ game between them in the opening 9 fixtures. As you’d expect, these two sides are compensated by a lot of green towards the end of their campaigns, with the Bees also having a very favourable run of games from the start of November through until Christmas.
Quick plug: you can now find the blog on Facebook at http://www.facebook.com/experimental361