The appliance of science: Fulham fans show considerable bias in predicting the rest of the season

Have you heard of the wisdom of crowds? The thinking is that a series of quite well informed judges will be better able to predict something than a single expert. This might be demonstrated by a 'guess the weight of the cow' competition: take the average of twenty farmers' guesses and you'll probably be closer to the cow's real weight than if you ask one cow weighing expert for his opinion.

For Fulham fans it has been two weeks since we had anything to jump to conclusions about and frankly this is making many of us nervous. ARE WE STAYING UP OR NOT?

To find out I asked a number of these same Fulham fans to run the BBC's predictor machine. In this you are given the fixture list for the remainder of the season and asked to fill in the scores you expect for each game. The BBC then works out the league table based on your predictions. 

For the purposes of this exercise I only bothered reporting on the bottom half of the table. Here are the average points totals from the 19 completed predictions:

Middlesbrough     40
Newcastle     37
Fulham         36
Sunderland     34
Wigan         32
Bolton         31
Reading     30
Birmingham     29
Derby         12

Hmmm. What is going on? Not only are Fulham staying up, but they are doing so quite comfortably. This suggests considerable optimism in SW6.  And why not?  These points can be achieved without having to do anything too crazy. We'll just have to beat the likes of Birmingham and Sunderland at home, get some points from Reading and Derby away, and scrounge up a couple more points (through valiant 0-0 draws I expect) from other, harder games. It's not impossible.

At this point history throws a spanner into the works. Here are the points totals of the 17th best team in recent Premierships:

2006-7: Wigan 38
2005-6: Portsmouth 38
2004-5: West Brom 34
2003-4: Everton 39
2002-3: Bolton 44
2001-2: Sunderland 40
2000-1: Derby 42
1999-0: Bradford 36
1998-9: Southampton 41
1997-8: Everton 40
1996-7: Coventry 41
1995-6: Southampton 38

According to our predictions eight teams will end up with fewer points than Wigan did last year and Portsmouth did the year before that! This is plausible on one level - as the big teams get better the smaller teams will find it harder to get points - but also suggests that we are collectively underestimating the ability of teams to pick up points in unlikely places as the going gets tough. So while Fulham may well manage put some points on the board, our direct competitors may find more. The Times' own Fink Tank predictor agrees, and sees Fulham finishing second from bottom with 32 points.  It will be tight though and a couple of strange games here and there can change everything.

So who do we trust? The cold blooded simulator or the one eyed wisdom of 19 Fulham fans? Who knows. It's going to be a terrifying few weeks.

Comments

[wassarra]

shiiitttt...my foot fulham will come in bottm 2...

[P. Thompson]

The mismatch stems from the fact that the response from each group of fans seems to have been aggregated only in respect to the points their own team will pick up. Because each result distributes two or three points between two teams, you need to include all the predictions from all the groups of fans into the aggregate calculations, so their predicted results count for the opposing team as well as their own.