Click here for Fink Tank’s predictions here
WELL that didn’t take long. The season hasn’t started and I’m furious already.
Honestly, how much more of this am I supposed to take? You listen to one radio programme and they tell you that such and such a player is tired because he has been playing in the World Cup. Then you turn over and hear the same player being talked about. Apparently he is rusty because of the close season. Aaaaaaaaagh.
And then there’s Chelsea. One newspaper that I was “enjoying” informed me that Chelsea were certain to retain their title. They then opined that Liverpool would give Chelsea a good run for their money. Now either this punditry was brilliant (such fantastic clairvoyant skills that they know Liverpool will do very well but in an agonising finish will fall just short) or it was asinine. Take your pick. I have.
So let me explain in words of one syllable (well not actually one syllable, but you get the point):
Advertisement
1. It is impossible to predict who will win the Premiership. All you can do is say how probable different outcomes are.
2. When you have done the calculations you will find that none of the outcomes will be 100 per cent likely (hence the absurdity of “Chelsea are certain to retain their title”) and none will be 0 per cent. I once made the error of informally referring to a side’s chances of winning a trophy as being zero when in fact their chance was 0.3 per cent. Unfortunately, I was referring to Greece and the tournament was Euro 2004. I have not repeated this mistake.
3. When you produce a table of the most likely positions for sides to finish up in, your table may be the best possible estimate, but it is still likely to be completely wrong. This is not just because teams change in unpredictable ways, it is also because results, even between equally skilled teams, are randomly distributed and you won’t find out exactly how that unfolds until the season finishes.
Right. I am glad I’ve got that out of the way. Now let me give you the Fink Tank’s best estimate of the season ahead.
Every result (as well as data on shots) for the past four seasons has been inputted into the computer model run by Dr Henry Stott and Dr Ian Graham. Properly weighted, this data allows us to rank teams and calculate the probability of one side beating another on a given ground.
Advertisement
We then used the average data for each team last season and simulated the entire season to work out how many points each side were likely to get. Using the average allows us to guard against distortions caused by a few games at the end of last season but has the disadvantage that a team deteriorating towards the end of the season (Manchester United) are rated a little more highly than perhaps they should be, while one showing an improvement (Chelsea, interestingly enough) are mildly underestimated.
You can see the table for yourself, although please remember that it is simply the best possible estimate — the final table will definitely look different.
Here are some highlights: the title race is wide open with Chelsea’s chances only 35 per cent; Arsenal are real contenders and are as likely, indeed slightly more likely, to win the title as either Manchester United or Liverpool; Reading look a good bet for a mid-table finish (although it is always hard to rank such teams properly in the first few weeks); and Iain Dowie will have his work cut out at Charlton Athletic.
Another prediction — my fuming has only just begun.
finktank@thetimes.co.uk