What do the stats say about Reading?

Reading FC finished the year in third place with 85 points. Their record was 26-7-13. On the surface, those look like great numbers. However, once you dive a little deeper, things start to get weird. Reading scored 68 goals this season. A respectable number that puts them 6th best. The conceded 64 goals however. Only 7 teams conceded more goals than that. You can usually estimate where a team will finish on points based on their goals scored and conceded. Doing this for Reading says they should have finished about 20 points worse than they did. To highlight just how unusual Reading’s season is, no team in the last 20 years has been promoted with a GD of less than +9. No team that has ever given up more than 62 goals has ever been promoted. How exactly did they manage such success when metrics and history says they shouldn’t?

Shots and Luck

First, lets look at their luck. Since goal scoring is rare, getting a few lucky or unlucky breaks on shots can have a huge effect on your goal scoring numbers. One way to examine this in greater detail is by shots. Total Shots Ratio (TSR) compares the number of shots taken, with the number conceded to give an estimate of team quality. A great team will have a TSR over .60, a poor team will be well below .50. For illustration, Newcastle had a TSR of .612 this season. Fulham’s TSR was .576. What was Reading? A very poor .443.

So, far we aren’t finding any answers. One of the flaws in TSR is that it can’t factor in the quality of the shot. For example, if you have three shots from inside the six yard box on an empty net; you are likely to score more often than the team that is taking 20 shots from 30 yards away. Expected Goals (xG) is the best way to deal with this fact. Every shot is assigned a value based on how likely it is to score. Sum up those values, and you get the number of expected goals. Was Reading taking more great chances and conceding more bad chances? Not according to expected goals. Reading had an xG of 58 and were expected to concede 76. Somehow in trying to figure out how Reading succeeded with such a poor goal differential we’ve only made them look worse.

Team Style

Like Fulham, Reading is a possession oriented team. They were second in the amount of possession in the Championship and second in pass completion. Only Fulham beat them in those two categories. Unlike Fulham though, this possession doesn’t lead to a lot of chance creation. Where Fulham led the league in shots per game, Reading were tied for 19th. This tells us that Reading is very adept at holding the ball, but aren’t that dangerous when they have it. This style can be very good for killing off games. It just so turns out that Reading were the masters of killing off games this year.

It’s been said many times that goals change games. It may be a cliche, but it’s very true. And no goal is more important than the first goal. On average, the team that scores first has a winning percentage of 76.7%. Now, it’s a little higher for a team at home versus a road team. And a better team will win more often than a poor team after scoring first, but that gives us a good baseline of what to look for. Fulham this season scored the first goal 24 times. In those 24 games they were 17-5-2, good for a 81.3% winning percentage. Reading scored first 27 times. Amazingly they went 25-1-1 in those games. That’s a 94.4% winning percentage. That rate is probably not sustainable, but their style does suggest they do have an advantage to killing off games.

This is a double edged sword however. In games where they conceded the first goal, they only managed 2 wins and 3 draws across 17 games (they also had three 0-0 draws). This winning percentage of 20.6% is actually worse than you would expect on average, though to be fair twelve of those games were on the road where it is harder to come back. Fulham managed to go 5-8-8 in games where they conceded first. While that’s not great, a 42.9% winning percentage is better than most teams.

These numbers do a bit to explain why Reading looked so bad by TSR and xG. A team that’s trailing almost always shoots more than one who is winning. They might not be high quality chances, but even enough low quality chances can add up to a big xG number. Reading didn’t underperform their xG scored by very much, but they did vastly overperform their xG allowed. Perhaps playing with a lead so often allowed teams to generate a lot of low quality shots that added up to make things look closer than they really were.

Blowouts

Reading was the third best team in the league at winning one goal games. Their win percentage in goals decided by one goal or that ended in a draw was 71.7%. Only Huddersfield Town (73.5%) and Brighton Hove Albion (74%) were ahead of them. Reading only played in one more one goal game than Fulham. They were just much better at winning or drawing those games. They had 10 more wins, and 7 fewer draws. Where Reading struggled was in games decided by more than one goal. They went 8-8 with a -9 GD. Fulham on the other hand went 14-3 with a +27 GD in games decided by at least two goals. Readings style doesn’t make it very easy for them to come back when down by multiple goals, and some of their results suggest they weren’t that interested in trying when the deficit got too big. Those blowouts may have impacted their shot number and xG numbers quite a bit and could be making them look worse than they really are.

What does this all mean?

By a lot of metrics Reading looks like the outlier in the playoffs. They have the worst defense, they give up the most shots, they shoot the least, their GD is second worst, and their xG difference is off the charts bad. Digging a little deeper though, we do see some reasons for their success. They are good at scoring first, and they are just about perfect once they get a lead. If Fulham can score first in these matches, it’s going to make it hard for Reading to come out on top. It’s tempting to look at Reading’s xG numbers, TSR, goals conceded, and xGD, and think Fulham should be the big favorite. But some of the underlying data suggests that Reading is a little bit more than smoke and mirrors and won’t be the easy out that Fulham fans are hoping for.