24-25 season,  predictions

How are teams performing compared with last season?

Ok, it’s very early doors, and some teams have only played five league games, but let’s have a quick look at how all of the teams are performing in the league, compared with what might be expected based on last year. To do this I have taken the number of wins each team has recorded in the league, and subtracted from this the predicted score based on last year’s match ups. As I detailed here, I’ve got two methods for calculating this. The first is just straightforwardly taking the average number of goals for and against when two teams played each other, separated out for home and away. The second more complicated, but not necessarily more accurate method is to look at the average number of shots each team took for and against each opponent, and then multiply this by their season average conversion rate.

To take an example, last season when Sheffield played Belfast at home they took an average 50.67 shots and faced an average of 36.67. Over the whole season against all opponents they took an average of 60.17 shots per game and scored an average 4.2 goals. This gives them a shot conversion rate of 60.17/4.2= 14.33, ie they scored on average every 14 shots. So, for Belfast then we can divide the number of shots they take by the conversion rate to give expected goals of 3.55. We do the same for Belfast, who have an expected goals of 2.35 in this match-up. With rounding and the prediction would be 4-2. It was actually 3-2 after shootout, so the actual score prediction wasn’t correct, but the overall outcome was.

Still with me? So we have these two methods of comparison with last year. They don’t differ wildly but the second one might even out some of the more freakish results. We can now look and see how many wins each team has compared to what might be expected. If we subtract expected wins from actual wins, that gives a rough indicator of performance against expectation. It’s fabulously early to do this, so for example Sheffield have only played 5 games, so one errant result gives them a score of -1. But let’s do it anyway and then revisit later in the season.

To absolutely no-one’s surprise, Coventry are the team performing way better than last season. I mean we knew that, but both predictive methods had them winning only 1 game from their first 6 and they have won 5. This gives them a “doing better than expected score” of 4 (5 actual wins – 1 expected win). It is of course easier to do better than last season if you were really crap last season, so for Sheffield a doing better score of 0 would be fantastic as it’d probably seal the league again.

Anyway, here are those scores:

TeamBetter than expected (Av Goals)Better than expected (Expected Goals)
Sheffield-1-1
Coventry44
Nottingham00
Manchester-11
Guildford20
Dundee01
Cardiff01
Glasgow-1-2
Belfast1-1
Fife-2-1

So depending on which method you take, Nottingham are doing about as well as last year, which will be a cause for concern if it’s the same after 10 games. Glasgow and Fife fans I regret to inform you that even though current form looks bad, it’s even worse than might be expected. But it’s early yet so the sample is low (and uneven) so I wouldn’t draw much from these yet and declare the league is going to Coventry or other teams should be pressing the panic button, but consider it an early whiff of future direction.

Leave a Reply

Your email address will not be published. Required fields are marked *