These templates are designed to provide a compact preview of a given match. They have a similar “look” to the match timelines but are a bit bigger due to smashing together compact versions of two visualisations: each club’s E Ratings and their relative positions on the attack and defence scatter graphics.

I’ve broken an example into sections to break down what they’re intended to do and how to read them:

E Ratings (with ranks)

The first section shows the current strength rating of each team – split by attack, defence and overall – and how this has changed over the past 30 matches. The numbers all show the current rating in “expected goals” (the calculations are explained here) and the brackets after them are ranks within the division. The lines show the progress over time, with the straight horizontal line showing the division’s long-term average for comparison.

In the example shown, Barnet have an attack rating of 1.21 (15), which means that they would be expected to score 1.21 goals against the average team and that this ranks them as the 15^th best attacking side in the division at the moment. We can see that their attack rating is below the average line but has recently dropped from being above it, suggesting that their attack is less potent than it was around 15-20 matches ago.

The defensive ratings work in the same way, but show the number of goals each team would be expected to concede: therefore low is good and high is bad. Finally, the overall rating is simply the attack rating minus the defensive one, i.e. the “expected goal difference” in an average match. This can be used as an overall strength rating, and it shows Rovers having improved steadily this season to the point where they are now ranked as the third strongest side in the division.

Probabilities

As many others have done with their own ratings, I have plugged the E Ratings into a match simulation model which generates a probability for each of the three main outcomes: a home win, a draw or an away win. These are to be taken with a pinch of salt as the model has relatively few inputs – things like line-ups, injuries etc aren’t factored in, for example – but in testing they’ve worked well enough to justify their inclusion and they give a useful indication of how difficult the match should be for each side.

We can see in this example that Bristol Rovers’ strong ratings this season make them marginal favourites despite being the away side. A quick glance at OddsChecker shows that the bookies are even more bullish (giving Rovers closer to a 50% chance of winning), perhaps due to factoring in player absences or placing a greater emphasis on recent performances than the E Ratings do.

Performance

I also wanted to include a more traditional and transparent measure of each team’s performances, so I’ve squeezed minimalised versions of my attack and defence scatter graphics in. These are structurally identical to the large stripey ones, but only the two featured clubs are coloured in and the “best” quadrants are highlighted in green.

In this example, the attack chart shows that Bristol Rovers have taken far more shots than Barnet this season but have been less clinical in front of goal. The defence chart shows that Rovers have been among the better performers defensively – they’re in the “good” quadrant – while Barnet have the division’s second leakiest back line despite allowing a relatively average number of shots.

I’ve chosen to keep these plots as showing the current season’s record only, as otherwise it would give a misleading picture for teams who were recently promoted or relegated.