Why predicting football is about looking at trends, not individual games

By Tony Attwood

The Athletic in its Daily Football Briefing for 31 October 2022 is running a piece which begins, “One year ago yesterday, Tottenham Hotspur were beaten by Manchester United.”

They then add that the defeat was “so humiliating that Daniel Levy knew he had to put into motion his well-guarded plan. Nuno Espirito Santo was out as head coach and Antonio Conte was in, just five months after saying thanks but no thanks to taking over at Spurs.”

On the day one year and one day ago Arsenal beat Leicester 0-2 away and Tottenham lost 0-3 at home to Manchester United.

The result left Arsenal in sixth and Tottenham in eighth, two points behind Arsenal.  And you will recall that by the end of the season, Tottenham had made up that two points gap and in fact reversed it, ending up two points ahead of Arsenal, and thus clinching fourth spot to Arsenal’s fifth.

According to the Athletic, “It felt like a transformative moment, landing one of the best coaches in the world — and in many ways, it has been. Conte instantly improved Spurs and gave the club a sense of upward momentum they had not had for years. The players were fitter, hungrier and sharper. The football ideas clearer, the defence better organized, the attacks more coordinated. Conte driving Spurs up into fourth place will always be a triumph of coaching.”

And yet now, one year on from the incident which led to Tottenham changing managers, Arsenal are five points ahead of Tottenham, with Arsenal having a game in hand as well.  Additionally, Arsenal have a nine-goal better goal difference than Tottenham.

Tottenham were full of optimism at the end of last season, boosted by unending positive publicity from the media. The manager (who had been tipped to leave) stayed, and the media bought into the notion of another £150m equity being injected into the club (although they might have wondered about the astounding debt the club now has, and the reason why no naming rights sponsor can be found for the ground and the manager still wanting to leave).

But those issues were pushed aside.  “Tottenham on the up” was the message all the way.

And as a comparison of points after 10 games one year ago, and the same for this season shows, Tottenham are indeed on the up, as they have eight points more than they did after 10 games a year ago.

 

Pos 2020/1 Team Pts 2021 after 10 Pts 2022 after 10 Change Pos 2021/2
1 Chelsea 25 20 -5 4
2 Liverpool 22 16 -6 7
3 Manchester City 20 23 +3 2
4 West Ham United 20 11 -9 13
5 Manchester United 17 19 +2 5
6 Arsenal 17 27 +10 1
7 Brighton and Hove 16 15 -1 8
8 Tottenham Hotspur 15 23 +8 3

 

But the real point about this is that while Arsenal has made great leaps forward this season being ten points better off than one year before, other clubs have made some progress but not as much.

And of course there has to be a counter to this for while Arsenal, Tottenham and Manchester City have gained 21 points between them over last year at this stage, other clubs have slipped.  Liverpool are down six, Chelsea down five, West Ham down nine.

This is one reason why we do insist on looking at longer-term trends as well as the actual league table early in the season.

We have long since argued that the first three games of last season were an aberration caused by covid, and the League’s bizarre failure to allow Arsenal to postpone matches in the way that Liverpool were allowed to later in the season.

So if we take the last 35 games of last season we see the top of the league looking like this…

 

club  P W D L F A GD Pts
35 27 6 2 89 25 64 87
35 26 7 2 88 25 63 85
35 22 3 10 61 39 22 69
35 19 10 6 70 32 38 67
35 19 5 11 66 40 26 62

 

That shows the direction of travel of the clubs.   It is not a perfect guide for it doesn’t give a clue as to what went wrong at Liverpool, but most particularly shows that it was those much publicised opening three games where Arsenal were playing with very much a patched-up team that scuppered much of the season.

Indeed a forward line of Pepe, Balogun and Martinelli is one that clearly was not up to the task.  Pepe has now been shipped out on loan, Balogun is still only 21 and is now with Reims, and starting to show some serious goal-scoring form in the French league, while Martinelli was in the early stages of learning how things worked at Arsenal, and a season in which he scored six in 29 games (as opposed to five in 12 so far this season).

Thus we argue one should look for meaningful trends, not, as the Athletic does, one off shots like “One year ago yesterday, Tottenham Hotspur were beaten by Manchester United.”  The statement is true but doesn’t really tell us much.

Last season Tottenham got 1.87 points per game.  This season it is 2.0.   They are improving.  But just not as fast as Arsenal who last season had 1.82 points per game, but this season are getting 2.58.

So why have Arsenal improved at a much faster pace?  A prime reason must be because Arsenal have had a manager with a clear-cut plan which was to reform the club step by step over a three-year period.

In the last two-thirds of 2020/21 Arsenal were the second-best team in the league.   In the last nine-tenths of 2021/22 Arsenal were the third-best team in the league.   In the first third of this season, Arsenal are the best team in the league.  That is progress as a result of planning.

In 2019/20 Arsenal were the most carded club in the Premier League.  This season we are 14th.  This again is the result of planning, month by month, year by year.  Sometimes a manager has to go but Tottenham have had 12 permanent managers this century.  Arsenal has had three.  There’s the difference.  

Leave a Reply

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