Predictions 2015-2016 (Round 7)

Since the first six fixtures of the season have been played, my prediction system will now change to its permanent form where the predictions are based on the current table position of each team.  I expect that that will improve the accuracy of my predictions which has been relatively poor so far this season.  (On the other hand, home teams have not done that well this season and that is  also responsible for my poor performance since my system favors the home team.) 

That being said, my system also faded in the second half of last season which suggests to me that, perhaps, the system would work best if it were based entirely upon recent performance.  Very possibly, my system would be most accurate if I only considered the relative table positions of each club over the past X number of games.  Of course, the problem would then be deciding what X should be.  If X is too low, the unbalanced schedule and the limited sample size would overly impact the predictions.  If I chose too large an X, it would not be much different than simply relying on the whole season. 

On the other hand, the whole purpose of my prediction system was to test whether a simple mechanical prediction system could outperform, or at least perform comparably to, the experts.  Last year it did.  This year, so far it has not.  If I were to start using a system that required me to look at the last 15 matches for each club, it would go from a simple system that is easy to use to one that requires a lot of work which sort of defeats the whole purpose.  Of course, it still would be work that could be done by anyone without any special expertise.  I will probably experiment with this system in a few weeks and, if I have time, I might go back and see how the new system would have done with the games this season.  However, I am not anticipating a lot of spare time for this sort of thing so it may never happen. 

This past week I picked two games out of ten correctly with no correct scores for 20 points.  Lawrenson picked four out of ten correctly with one correct scores for 70 points.  His celebrity guest picked three correctly for 60 points.  Merson picked three out of ten correctly with no correct scores for a total of 30 points.  Yet another bad week for me.

My current total for the season is 26 out of 60 with for 350 points.  Lawrenson has picked 25 out of 60 for 460 points.  Merson has picked 29 out of 59 for a total of 530 points.

When Lawrenson and Merson make their new picks I will link them here and here--unless I forget or get too busy yet again.

PREDICTIONS

My system’s picking rules can be found here.  Based upon those rules here are my predictions for the next round of games:

Tottenham-Man City       0-1

Leicester-Arsenal             2-1

Liverpool-Aston Villa       1-0

Man U-Sunderland          1-0

Southampton-Swansea   0-1

Stoke-Cherries                  1-1

West Ham-Norwich         2-1

Newcastle-Chelsea          1-1

Watford-Palace               1-1

West Brom-Everton        1-1