Don’t hold your breath: the super computers are back

 

 

 

 

• Where is football actually heading? Well, since you ask, Saudi Arabia.
• How the Premier League manages to have a feeling of continuity

By Tony Attwood

Clearly no one takes any notice of my ramblings – well, I knew that anyway, but I thought maybe some commentators would realise that the notion that a supercomputer might be used to predict football results is so childishly stupid that it can’t fool many people.

And then there was no apology for the gross inaccuracy of last year’s predictions.  Untold put Arsenal down as third, thinking that the club might get three blips in the season, rather than the two the club actually got (27 January to 15 February: three defeats and one draw, and then 9 April to 20 May, two wins, three draws, two defeats.).

But of course that is the difference between us and them – we do reasoning they don’t.  Indeed they didn’t apologise for errors last year. So before considering the predictions for the year ahead let’s remind ourselves of what the mythical supercomputer said one year ago. 

So on 25 June 2022, the Mirror claimed it had its own supercomputer shouting, “EXCLUSIVE: A Mirror Football supercomputer has predicted how the Premier League table will look next season” 

At least they have had the sense to kick out the imaginary machine and this year are claiming to have bought time on someone else’s machine.  But let’s look at last year’s prediction for the season just gone.

 

So to be clear, using a supercomputer, the most powerful type of computer in the world, which costs tens of thousands of pounds per hour to use, they got only one of the top four right.  Three of the clubs that they predicted would be in the top six were finally in the top six, but three more (Liverpool, Tottenham and Chelsea) were not.

Now an error rate of 75% in terms of just naming the clubs in the top four (irrespective of order) is a pretty spectacular failure when supposedly using the most powerful bit of predictive equipment in the known universe (ie Earth).  As is the fact that they didn’t actually get any of the top six in the right position.

As for the teams to be relegated this past season they predicted Nottingham Forest, Bournemouth and Fulham.  None of those were correctly predicted.   Fulham ended up in the top half of the league, Bournemouth 15th and Forest 16th.  A pretty poor show.

Now I am not laughing at the Mirror for getting these predictions wrong.   What I am laughing at is their insane insistence that they are using a supercomputer, and still got everything wrong.

So you might think that at this point, the Mirror would a) sack its football writers and b) stop the mindless gibberish about having access to a supercomputer,  and that any owner of a supercomputer would actually allow to be used for football predictions.  

But what they are doing is ignoring past failures and following them up with another set of predictions that they imagine someone somewhere might take seriously.  This time they are not claiming to have their own machine but to use one owned by King Casino Bonus.   (I simply do not believe that the casino has a supercomputer, but we’ll let that pass).

This time I’ve stuck my guesswork in at the end of their…. err… guesswork.  The justifications for my guesses are that the saga about who owns Manchester United with rumble on, while Newcastle will use their Saudi wealth before the end of the window, to great effect, given the manager they have.

I think Manchester City feel utterly secure seeing the way that football is developing and if they were to be deducted points for past misdemeanors I suspect they would then join an international super league.   But in reality, I can’t see them letting their trial end for at least another two seasons.

 

Mirror Pos 2023/24	Club	Pos 2022/23	Difference	Tony’s guess
1	Manchester City	1	0	1
2	Liverpool	5	+3	4
3	Arsenal	2	-1	2
4	Manchester United	3	-1	6
5	Newcastle United	4	-1	3
6	Chelsea	12	+6	9
7	Tottenham Hotspur	8	+1	8*
8	Brighton and Hove Albion	6	-2	7
9	Aston Villa	7	-2	5
10	West Ham United	14	+4	16
11	Brentford	9	-2	10
12	Crystal Palace	11	-1	13
13	Wolverhampton	13	0	17
14	Fulham	10	-4	11
15	Everton	17	+2	18
16	Nottingham Forest	16	0	14
17	Burnley	P		12
18	Bournemouth	15	-3	15
19	Sheffield United	P		19
20	Luton Town	P		20

 

*OK I may be biased; they could make 7th

Of course, there is no rationale for any of this.   Just as last time around there was no rationale for Liverpool being predicted to leap up to win the league rather than missing that target by a whopping 22 points.  And no apology.

But we must note that the Sun is still hated in Liverpool ever since its lunatic reporting in 1989, and so the Mirror it seems, for political gain is pushing Liverpool up in the predictions to curry favour.

Still credit where it is due.  While most papers were predicting Arsenal to finish fifth or sixth in 2022/3 the Mirror had Arsenal up to 3rd which is also what Untold predicted.

But no, there is no supercomputer.  Just an attempt to stay in with the Sun haters, and otherwise do the obvious.