Revealed - The NUFC Difference Makers

Turns out there are stats that can tell us whether results are better or worse when a particular player is in the team. Matthew takes a look at what they reveal.

It's very tempting. We've all done it. It carries just the right level of apparent expert knowledge, mixed with slightly left-field, unfashionable insight. When a player’s absence has coincided with a disappointing run of results, it's a conversational move that's almost always destined to land well.

“You know what, I think we've really missed Joelinton/ Willock/ Longstaff/ Trippier/ Billy Askew...”

Best of all, it's also impossible to disprove. There is, by definition, no evidence to show what would have happened if that player hadn’t been unavailable. Which is particularly irritating for me as I'm inherently sceptical of just about everything anyone presents as football wisdom with no evidence to back it up.

But now we need wonder no more! Kind of. Because those clever statistical people have given us (OK, just me, at a loose end, on a day’s leave) a new set of figures to play with. For every player in the top five leagues in Europe, you can now find out with one click whether their team performed better or worse with them on the pitch.

And so, without further ado, here's the list for NUFC in 2024-25, ranked from the largest positive “effect” on results to the largest negative “effect” (of course, it's not really a causal effect, just a correlation).*

Interesting, right?

Now before you get overexcited and crown Matt Targett as our player of the season,** there are some people we sadly need to exclude from the list. Basically, if a player didn't miss many matches (Bruno, BDB) or didn't play many (Miggy, Osulaaaahhhh) then the sample is far too small to produce anything meaningful, when adjusted per match. For that reason, I'm going to exclude anyone who played or missed fewer than six full matches.

This takes out the outliers at the top and bottom of the list and a few in between. If we ignore the boring figures in the middle, we end up with two interesting lists of players.

First, there are five players whose presence on the pitch was somehow worth at least half a goal extra per match last season. In order from the top: Murphy, Dubravká (!), Tonali, Isak, Joelinton.

Then, there are four players whose presence on the pitch apparently cost us at least half a goal per match: Longstaff, Pope (!), Gordon (!!), Willock.

What does all this tell us? I reckon, five key points stand out.

1. The Purples earn their plaudits

Now this isn't surprising, but it's clear that our best players are also our most important players. And actually that's reassuring because it means the numbers confirm what we see with our eyes. Sandro, Alex, and big Joe are our most influential players, along Bruno who we knocked out of the list for statistical reasons (he was good, but not THAT good).

In addition to those elite players, the numbers confirm that two less elite players had brilliant seasons - Murphy (he really needs an affectionate nickname) and the Colossus of Blyth. In fact, whichever way you slice it, cheeky Jakey was our biggest difference maker by far last season.

2. Dubs over the Pope? Really?

We all remember the season before last, right? When the Slovak Yashin stepped in for the injured Pope and we promptly started shipping goals like early-era Jossy’s Giants. Well, in the season just finished our results with our two keepers were completely reversed. We were getting on for a goal a game worse off with Pope in nets than with Dubs.

The explanation? God knows! Oh, you want a more informed view than that?

Well, it could just be a case of our old friend “reversion to the mean”. The stats skewed to one extreme last year through random other factors and this year we've got a correction the other way. In fact, take the stats over two years and there's no difference between the two keepers, although you begin to see why Eddie might want another for his vast collection.

Dig into the data and other possibilities emerge. In 23-24, whether or not it fits with your perception, the stats show that Dubravká’s shot-stopping was still top notch and at least as good as Pope’s. The main difference between the two keepers was their starting-position, with Pope much further forward and able to sweep up more frequently, allowing the back four to play higher up while protecting them from being exposed for pace. This season, that difference in position was much less stark (16.0 vs 15.3). That might have played a role in shifting outcomes this season.

Remember, though, that these stats include both goals scored and goals conceded. And, in fact, the biggest difference for our keepers this season came not in the goals they let in, but in the goals we scored with them in the team (1.6 per match with Pope, 2.3 with Dubs). Maybe the latter's relative confidence with his feet enabled us to build attacks more effectively. His average passes were much shorter than Pope's, for example, more accurate, and more frequent.

Or maybe it's just the random way the fixtures fell. Or the mysterious man on the grassy knoll. Or an alien conspiracy led by David Icke.

3. The time has passed for Willock and Longstaff

In all the matches I saw last season when either Willock or Longstaff played, it was obvious that we were a less effective team with them on the pitch. This is not to judge them as people or to wipe away in an instant the contribution they made to where we are now, but these stats do seem to confirm the eye test.

Little Joe runs and runs but with little end product and at the expense of team control and structure. Always a confidence player, Longa looked shot, a victim last season of the excellence of our midfield trio, Eddie's reluctance to rotate, our good injury record, and the lack of additional fixtures to get games in his legs.

4. What's the story with Anthony Gordon?

Answers on a postcard. Maybe it's just a statistical anomaly. Interestingly (if you like this kind of thing), our xG stats were better with him on the pitch than off it. So maybe it's just a question of luck. Or his dodgy finishing? Either way, this is confirmation of a disappointing season if you measure it by the direct impact on results.

5. No supersubs

And finally... Because they played so few minutes, it's right to exclude Miggy, Osula, and Wilson from our analysis. But even then, what the stats do show is that these attacking substitutes hardly ever positively impacted matches when they came on. In fact we were a combined 7 goals worse off with them on the pitch.

Last season, we lacked difference makers to come off the bench. Let's hope that changes in the season to come.

Oh and let's also hope Matt Targett’s agent isn't reading and gets carried away with terms for a new contract.

Matthew Philpotts

---

*In case you're wondering, the figure is arrived at as follows: 1. Work out our goal difference when that player was on the pitch, then average it out across the number of 90 minutes they played; 2. Do the same for when the player wasn't on the pitch; 3. Take the second number from the first. The answer is the number of goals better or worse we performed per match with that player on the pitch. All courtesy of fbref.com.

**Targett played 21 minutes in two matches, weirdly both against Aston Villa, during which time he had a grand total of 17 touches and made 14 passes, no tackles, one clearance, and one interception. He didn't have a shot. But while he was on the pitch, we scored a goal - Joelinton in injury time at home to Villa - and we didn't concede. So statistically, his presence on the pitch made a greater difference per 90 minutes than anyone else!

---

---