Today is Super Sunday, the biggest day in sports in America, and I've got football on my mind. The interesting twist is that I'm talking about the kind of football where people actually use their feet to kick a small round ball.
Instead of heading over to watch the Superbowl at a friend's house with 50 (that's right, 50) other people, I decided to stay at home and watch it on a 50 inch plasma instead. So I swapped a person for an inch on my television - sue me. Actually, I'm not feeling that great and I preferred the surround sound to come from the television instead of other people and their little children.
Anyway, as the Steelers and Cardinals come out on the field, I've got this nagging feeling that Tottenham could go down this year. I know I've said it before, but this time I actually did a little research. I took the teams that are in the bottom 10 places in the league and listed out the remaining 14 games (about 1/3 of the season) for each in an attempt to predict where teams will finish at the end of the year. I took 2 different approaches, and both of them have some serious shortcomings. However, short of me predicting what will happen in 141 games (14 * 10 plus Blackburn's 1 game in hand), this is the best way I could think to do it.
Approach #1
I used current position in the table and previous head to head result. To simplify it, if the team in question is playing a worse team (position in the table) and in the first meeting beat that team, then I gave them 3 points for a win in the upcoming game. The reverse is true too. If the team in question is playing a better team (per their position in the table) and previously lost to that team, then I gave them 0 points for the loss in the upcoming game. There were various situations that resulted in giving the team in question 1 point for a draw...such as tied on points and the first game was a draw, or if they previously beat a better team or lost to a worse team (i.e. inconsistencies in the position versus previous result). If the previous meeting ended in a draw, the 3 points would go to the better team in the table. If the teams were tied on points in the table, the 3 points would be awarded based on the result in the first meeting. I know it sounds a bit confusing, but I got it all sorted out in my head.
Approach #2
I used past results to predict future results. So, if a team in question beat a future opponent in the past, the assumption is that they will beat them again (or draw or lose). The only drawback to this approach is that each previous game was either played home or away and that will flip when they meet a second time (i.e. each team plays each other team home and away). Also, this doesn't take into account current form.
Results #1
16. Tottenham - 32 pts
17. Stoke City - 32 pts
18. Blackburn - 29 pts (relegated)
19. Boro - 27 pts (relegated)
20. West Brom - 26 pts (relegated)
Results #2
16. Tottenham - 39 pts
17. Sunderland - 38 pts
18. Boro - 38 pts (relegated)
19. West Brom - 37 pts (relegated)
20. Blackburn - 34 pts (relegated)
So, depsite the obvious shortcomings, the results bascially confirm why I'm worried. If everything stays status quo, obviously the teams currently in the bottom will stay in the bottom. However, some of the other teams level on points with Spurs (Newcastle, Portsmouth) climb the table while Spurs remain in the bottom. As you can see, there isn't any room for error. Good thing the games aren't played in paper.