Can we really predict the outcome of the league table from earlier in the season?


By Tony Attwood

Predicting how things (or indeed people) will turn out is something that has been part of human existence for as long as we have been in societies.   Of course, if we really could know the future the whole of society would change since by and large there would be no point in putting any effort into anything.  After all if we know who would win this season, there wouldn’t be much point in competing, or watching.   I mean, for the last game of the unbeaten season, we were all pretty sure Arsenal would beat already relegated Leicester, but we still wanted to be there to see it happen.  And at half time we thought it might not.

If we all felt that we knew what would happen, then I suspect a lot of the crowd numbers in the Premier League would decline.  Yes, we might enjoy watching TV dramas in which we know what will happen (because it’s a series and the format always repeats), but we still want to see HOW the whole thing works out.

But just how predictable is the Premier League?   Here is the the league table at the start of 2023 showing the number of points, and then again at the end of the season.   Because different clubs have always played different numbers of games we’ve translated this into points per game, so that in the final column we can see which clubs had a better 58% or so after the world cup and which ones had a worse 58% of games in that period.

And there was something of a surprise.   Because the top five teams one year ago all had a worse performance after the world cup compared with before.   Their points per game worsened by anything from 0.03 to 0.35 (in the case of Arsenal).   PPG = points per game.


Pos Team P Pts PPG end pts END PPG PPG +/-
1 Arsenal 16 43 2.69 84 2.34 -0.35
2 Manchester City 16 36 2.25 89 2.21 -0.04
3 Newcastle United 17 34 2.00 71 1.87 -0.13
4 Manchester United 16 32 2.00 75 1.97 -0.03
5 Tottenham Hotspur 17 30 1.76 60 1.58 -0.18


Now at this stage we have not undertaken an analysis of all the clubs in the league, and quite evidently if the top five all declined, some of the teams below them must have improved.

But a quick look back at the end of last season shows us something interesting.  The top four clubs around this stage of the season were all in the top four at the end of the season.   The club dropping out of the top five was Tottenham who were replaced in fifth position by Liverpool.   Tottenham in fact sank right down to eighth.

Of course not too much is made of findings like these because they reduce somewhat the excitement of the league – if the position at this stage in the table is going to be pretty much the position at the end of the season but with the top clubs all performing a little worse.

I mean if you were running a newspaper column which was there to generate excitement rather than explore what is actually happening, would you run the story that in the latter stages of the season most f the clubs performed worse than before?   Probably not.   That is not the sort of story the media wants.

But this does suggest something rather interesting.   That is that the clubs that finished near the top of the league had a format and approach, which they stayed with.  What could upset that approach, as we have seen before is the sudden change in terms of injuries or an attempt by one of the clubs to throw everything into making that one extra push to move up the table.

That “one extra push” can sound like a good idea, but in fact, it normally ends up as being a change to a less successful system.   The best results generally come from clubs which are staying as they are.   This explains why so often as we have discussed earlier. what happens is not a general slip backwards (such as winning t io and drawing two in four games, rather than winning 3 and drawing one in four games) but a sudden crash – as we saw with Arsenal last season.

And what is so interesting is that the media constantly tell us that the league table does not truly “take shape” until 10 games of the season have been played.   But this season after ten games Tottenham H were top of the league and the media were full of that being their best start of a season since the Premier League was invented – conveniently forgetting that measures of the future are unreliable after just ten games.

Indeed as the Athletic pointed out recently a study looked at match data from 1995 to 2017 “found that the team who topped the Premier League after matchweek 10 had a 77.3 per cent chance of finishing in the top three by the end of the season.”


6 Replies to “Can we really predict the outcome of the league table from earlier in the season?”

  1. Off topic but half time atthe Emirates and our Women are three one up against Chelsea. Beth Mead with the opener in the 6th minute, Chelsea equalising with an offside Sam Kerr, clearly inthe sightline of Zinsberger. Naturally ignored as an Arsenal player involved. Illstead with a header from a Catley corner and Russo with the third. So far so good.

  2. Final score 4-1. Russo with a penalty. Rotten officiating to the end, some in Arsenal’s favour but we are now level on points with Chelsea and have narrowed the goal difference to 2. If only Blackstenius had taken a couple of her gold plated opportunities at the end of the game we would be top.

  3. Same as Andrew off topic , I watched the Sp##s v Bar Codes this afternoon , how did Romero stay on the pitch with a cowards tackle above the ankle where the player was on the floor . Sky kept replaying the incident along with Neville saying virtually how is he remaining on the pitch , VAR got involved and didn’t even send the referee to the monitor , nothing makes sense anymore to me .

  4. Steve Vallins

    Not only that but at the same time 2 players came head to head in a manner that would without doubt of seen an Arsenal player sent off. NOTHING.

    I wish it didn’t make sense, unforetuneately, as far as I’m concerned it all makes perfect sense.

    One rule book for Arsenal, one for everybody else. Been like this for years and it’s just getting worse.

  5. seismic

    It’s amazing the amount of big decisions that have gone against Arsenal when that Gillette character has been involved, be it as referee or VAR.

    The first major decision I remember was a VAR call in a Palace match some years ago when he disallowed a late winner for a foul that literally nobody else could see.

    He has been involved in many more controversial decisions against us over the years, including at least 3 this year alone, most of which have at one time or another been highlighted on here by either myself or others.

    I have talked on here about a theory I have had for years where there can be cheating without cheating. It’s based on the fact that so many decisions in football are subjective. Foul, no foul? Red card, Yellow card? Hand ball, not hand ball? They are all subjective.

    Even what you would think of as an objective decision, such as offside, can be open to ‘interpretation’. At what point was the ball played? What camera angle do you use to place the lines? When offsides are being given for millimetres these things can be crucial. There is absolutely nothing to stop an unscrupulous VAR operator winding back and forth, changing angles, as often as he wants before selecting the right angle for any particular outcome. This could explain why it sometimes takes an eternity for them to ‘put the lines up’. They keep looking until they find the freeze frame they want.

    Now my argument is, with all this subjectivity, all these 50/50 60/40 40/60 calls, some should go for you and some against. And if they did, well perhaps you could take it. But that simply doesn’t seem to be the case with us. Of course we get VAR calls and refereeing decisions that go our way, but I believe we get a massively disproportionate amount of 50/50 calls go against us. And if all was fair and above board that should not be the case.

    The problem is, as I say, they are subjective decisions, and by their nature subjective calls can go either way. In other words, it would be very hard to accuse anyone of cheating, or even bias, on the basis of one 50/50 call going against you. But personally I believe for us, (and possibly other teams as well), these subjective 50/50 calls go against us much much more often than 50% of the time.

    So my question is, is it cheating without cheating when such a disproportionately high ratio of these tight calls go against us? I think it is. At the very least there seems to be some sort of unconscious bias at play.

    And the thing is, this Gillet Character seems to constantly err on the side of our opponent. Not occasionally. Not 50% of the time. But every single time.

    He is a Liverpool fan. Lets not forget we are direct rivals. There is history between Arsenal and Liverpool, and some still bear a grudge for Anfield ’89.

    And consider this, because he is a Liverpool fan he is not allowed to officiate in Liverpool matches. Why would that be? It’s for a reason. It’s because there is a real risk of bias, either conscious or subconscious, in favour of Liverpool.

    As such, why is he allowed to referee the team that will go level with the team he supports if, just for example, they should score a late equaliser at Aston Villa? He cannot officiate Liverpool because he may be biased towards them, but isn’t being biased against their direct rivals just as possible?

    Of course it is, and as such it is not a stretch to suggest he could be biased against Arsenal, either consciously or subconsciously. It is not conspiracy theory. It is a real possibility, especially when he has a history of giving so many contentious decisions against us.

    Personally, and this is just my opinion, I think Gillette is a cheat, plain and simple. But as I say, just my opinion.

Leave a Reply

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