The Stat Man: Is the result all that matters?

When the fixtures came out for the current season, many remarked how tough a start Sunderland had. With the exception of the home season opener against Bristol Rovers, the first five games seemed challenging with Oxford, Peterborough, Charlton and Blackpool in successive weeks.

Against that backdrop, the current seven points on the board after three games is probably more than many of us might have expected (yet undoubtedly demanded). The league position is healthy enough too, with Sunderland sitting in fifth place behind three sides with 100% records and Doncaster with their superior goal difference.

If you read my article last week, you will know that I agree with you as you mutter ‘it’s much too soon to read anything into the league table’. But you will also know that I believe it isn’t too early to start assessing performances. A point I will return to shortly.

---

The result is all that matters, man!

Saturday afternoon and much of Sunday was notable on #SAFC Twitter for fans apparently dividing into two distinct camps.

On one side were the fans who believed all that mattered was the result against Peterborough and any criticism of the quality of the performance was out of order. Then on the other side of the argument were those who were also thrilled with the result but equally wanted to make their concerns about the quality of the performance known.

This early in the season, sides have played so few games that randomness and luck (both good and bad) cloud the league table in a cloak of misleading variance. And then there is the imbalance in the quality of teams played. So while Sunderland’s current position in the league table is healthy enough - and good results are of course welcome - performances definitely matter.

On December 18th 2018, Sunderland were sitting 3rd in League One on 43 points having won 12, drawn 7, and lost just one game. On that day, a prominent and respected ‘stats type’ (and betting analyst) took to Twitter to point out that the league position was false and there was no underlying performance data to justify the ‘lofty heights’ of third in League One (yeah I know...). He signed off his tweet with “*runs for cover”.

Photo by Ian Horrocks/Sunderland AFC via Getty Images

What followed were some exchanges with Sunderland fans that ranged from anger to ridicule. One of my favourite replies said that this was “proof that stats mean little“. The responses prompted the analyst to further tweet:

Occasionally there are outliers, that’s natural. But all reasonable logic suggests they’ll regress to normality at their current rates - it’s not science, it’s just fact that if you’re giving up more quality opportunities than taking yourself, you’ll eventually come a cropper.

And...

I’m not the only one saying this. They could be Real Madrid and they’d still struggle to maintain their current results by turning in those sorts of figures.

And...

Unless you’re Messi, the team matters more than individuals. Of course, higher budget players will over perform but I keep going back to it, they can’t/won’t do so consistently at this rate. To be reliant on moments of magic etc is concerning, to me.

The person tweeting all this was @MarkOHaire - give him a follow if you’re on Twitter.

I mention that pre-Christmas 2018 Twitter skirmish because - for the opening 20 ish games of that season - I, and several others, had been saying the same things. Results were good but performances were poor and that wasn’t sustainable.

So it came to pass that results converged - over the remainder of the season - to the level the performances predicted they would. “Proof that stats mean little” anyone?

Results matter in the short term but performances matter far more in the longer term. It is the latter that will determine the outcome of the season, not the scoreline at the final whistle following a fortunate 1-0 home win against Peterborough.

---

League One performances so far

Having emphasised the importance of performance, I now want to follow up on what I began last week. If you didn’t read that article, it might be helpful if you do for a better understanding of what follows. Briefly, however - when viewing the wheels - black, green and red are good while white and grey are bad.

Here is how every League One side is looking so far this season:

Some notable good performances on display above come from Blackpool on the top row, Hull and Ipswich on the second row, and Portsmouth and Sunderland (highlighted in red) on the bottom row. Conversely, Bristol Rovers on the top row and Rochdale on the bottom row look poor.

I’ll return to the bigger picture shortly, but first I want to take a closer look at Saturday’s game against Peterborough. Below I show the performance wheels for the two sides for the individual game (remember the charts above show the whole season).

Peterborough looked very dangerous in the early parts of the game. They were unlucky not to be one or two goals ahead after fifteen minutes having, at that stage, already had four shots on target. As you can see from their Goals and Shots ‘spokes’, however, not only did they fail to score, they even failed to generate the number of shots one might expect from a promotion chasing side. It was noticeable that they faded somewhat after that bright start and created too few shooting opportunities for the remainder of the game.

That Peterborough were able to threaten is reflected in Sunderland’s defensive half of the wheel. They looked less solid than in the previous two games as is represented by the small SOT (shots on target) spoke that shows Lee Burge was being tested.

The biggest area for concern in the Sunderland wheel is at the top right. Namely, the Goals, Shots and SOT spokes. Anyone who read my articles last season will know I moaned regularly about us having ‘relegation levels’ of shots on target and, as a result, not scoring enough goals to be a realistic promotion candidate. In the Bristol Rovers and Oxford games, we looked promising in that respect, having got eight and four shots on target respectively. But it was back to earth with a bump on Saturday with us managing just two shots on target and one of those being a penalty.

To the right of 6 o’clock is SPA (accurate smart passes). A smart pass is “a creative and penetrative pass that attempts to break the opposition’s defensive lines to gain a significant advantage in attack“. I note with interest that Peterborough are ‘elite’ for this attribute while it is non-existent for Sunderland. Not doing much quality smart passing is not unique to Sunderland in League One, but it is important - seven out of the eight promotion winners of recent seasons scored very high for SPA. The remaining one of the eight was Phil Parkinson’s Bolton Wanderers. Even they built smart passing into their game, just not at the elite level of the others.

Below I show the wheel for Sunderland’s season as a whole in more detail.

Overall, this is healthy. However, I draw your attention to the absence of an offensive spoke for SPA, thus illustrating that the absence of such passes extends beyond just Saturday’s game.

More concerning for me is the top right portion. There are signs that we are trending towards the same issue that plagued us last season - relegation level attacking threat.

---

Performance league table

Last week I presented you with a league table based on performances to date. It basically converted the wheels for each side into numbers and ranked them from best to worst. I’ve done the same here but this time taken I have taken into account the quality of opposition each side have faced and the numbers have been adjusted accordingly.

Good news again for us as we cling on to top spot even when taking into account the quality of the opposition faced. The visit of Blackpool on October 10th looks like it could be a tasty encounter!

But what is going on with Portsmouth? A look at the League One table will tell you’re they’re in the relegation places, yet they are third here. And then there is Lincoln who are second in the league table yet 16th here. Doncaster too. “Proof that stats mean little”? We’ll see.

It is of course to be welcomed the Sunderland top this performance table, even though the season is just getting started. It’s important not to read too much into just yet though. Things will settle to something we can believe in more firmly in the next few weeks.

But even at this early stage, I confess I’m a little concerned that Peterborough found us so easy to defend against. That is an aspect of the game that killed us last season and could again without some changes in our attacking play or personnel.