The Stat Man: Examining Sunderland’s performance since two watershed moments in the season

When this article of mine was published last week, I had someone on Twitter request that I look at Phil Parkinson’s data in isolation from Jack Ross. This is something I have done before but not recently and not with this specific set of analyses.

If you want a full explanation of some of the charts that follow, please read here. I will give some high level descriptions as we move through the data but I won’t explain as exhaustively as I did previously.

---

Ross vs Parkinson

In each of the charts that follow, I have included three data points for Sunderland, all from this season. Sunderland 19/20 represents the season as a whole (both Ross’ and Parkinson’s games), Sunderland (Ross) 19/20 and Sunderland (Parkinson) 19/20 have a green and red border respectively. What the data points represent should otherwise be fairly obvious.

If it was hoped that the switch from Ross to Parkinson would result in a shift towards the top-right corner of the plot (that’s where automatic promotion happens), then I’m afraid the opposite has happened. The attacking side of Sunderland’s game has declined both in the number of chances being created and the number of goals being scored. In fact, that the Parkinson data point is touching the red ellipse suggests we are more relegation worthy than promotion chasers (at least when only looking at the attacking side of our game).

Next we take a look at defending.

As nice a bloke as Jack Ross was, he will forever be remembered for 1-1 draws. While the attacking plot shows Ross having achieved about 1.40 goals per game, the defending plot shows his Sunderland also conceded about 1.30 per game. With a result of 1.4 for and 1.3 against, you can understand why we saw the dreaded 1-1 rather too often.

On the other hand, the position of the Parkinson data point suggests a significant improvement in the defensive side of our game since his arrival (this is something I have mentioned often in my previous articles). Since taking over, the new manager has taken a team conceding 1.3 per game and got them conceding less than 0.7.

Ross had us performing better as an attacking force and Parkinson has overseen a transformation of the defence.

But which manager has the better performances overall?

Again, please take a look here for a full explanation of this chart if you need it. Briefly, the closer a data point is to the bottom left corner of the plot, the better the overall performance.

It is clear that the Parkinson data point is closer to ‘perfection’ than that of Ross. Quite a lot closer actually. Using some help from my old Greek mate, Pythagoras, I can state that Parkinson’s Sunderland have performed 24.4% better than Ross’ Sunderland. However, unfortunately for us, every one of the automatically promoted teams from the last three seasons (including Parkinson’s Bolton) were better.

---

The Resurgence

It won’t have escaped your notice that results were grim from October 19th (following Parkinson’s underwhelming appointment) to December 26th. In that period, Sunderland dropped from 6th to 15th in League One. Thus ‘achieving’ the lowest league position in 141 years of a great football club.

The horror with which each appalling result was greeted was intense to say the least. The fans simply could not understand why the situation was being allowed to continue. For many, setting out to get a draw at Gillingham was unforgivable and the final straw. For others, that moment arrived during the 0-0 home draw against Bolton on Boxing Day.

Boxing Day did indeed prove to be a watershed, but not in the way most expected. When Sunderland travelled to Doncaster three days after the Bolton debacle, few predicted a very good performance and an even better result. That is what they got though. And so began an unlikely, unexpected and very welcome upturn in form that led us to where we are as I type - pondering the intricacies and probabilities of grabbing an automatic promotion place with at least a modicum of reality.

So what do the attacking, defending and overall performance charts look like if we only look at events since Boxing Day? Let’s find out.

Although results have very much taken a turn for the better, the attacking part of our game has only shown a small improvement. Sunderland - even during the good run we have been on lately - still suffer from a chronic lack of creativity in the final third. However, you will note that the data point representing Sunderland since Boxing Day is well above the dotted line, thus indicating that we are converting a greater than average proportion of our shots on target into goals. Good, but we need better.

Now let’s turn our attention to the defensive side of the game.

Last week I mentioned that Wigan Athletic had a “freakishly good” defensive performance during their promotion campaign in 2017/18, but, on this evidence, Sunderland have conceded even fewer goals per game than the Latics’ average since Boxing Day (0.56 per game to be precise). However, we are still a little way behind Wigan for shots on target given away so there is, even now, room for improvement.

Finally we examine the overall performance.

For the first time in this article and the one from last week, we see Sunderland right in among the automatically promoted teams. Again using a little Pythagoras, I rank the teams plotted above in order of ‘closeness to perfection’ as follows:

But for a truly dreadful start to Parkinson’s time in charge of Sunderland, that saw them hit relegation form for over two months, we might have been in the automatic places already.

As it stands though, we have to hope that not too much damage was done in that late 2019 period. If Parkinson can pull off automatic promotion from 15th in the league on Boxing Day, then it is he who will deserve a pint from every one of us rather than Richard Hill...