After 14 games, year on year comparison

[J-Bay Blue]

Ah, the old statistical argument. You have to take into account which teams you have played and whether it was home or away. MANUre had the softest start of anyone. Go through each team and see how the result varies from last season.

Actually, his statistical inference is correct - there is less variance / a lower standard deviation this year - so far.

By using proper statistical techniques, you CAN draw some good inferences and make some reasonable predictions.

For example, run a corelation matrix and do multiple regression to see which are the most important factors in determining success.

Of course, many factors are inter-related - any six year old can work out that there is a positive relationship between goals scored and victory, and a negative relationship between goals conceded and losses.

The thing is that not all these factors are equal, by any means.


Bookies and betting companies understand these issues very well - large parts of probability theory derived from the gambling fraternity.


One factor that is VERY important if you want to finish top 4 or top 1 is goals conceded, and here we are far better than we were lat the same stage last year - the fact that we have conceded the same fewer amount of goals this year compared to last as we have scored fewer goals is not neutral - it is very positive.


Bottom line - we are in much better shape than we were this time last year.<br /><br />-- Thu Nov 25, 2010 10:51 am --<br /><br />

Ah, the old statistical argument. You have to take into account which teams you have played and whether it was home or away. MANUre had the softest start of anyone. Go through each team and see how the result varies from last season.

Actually, his statistical inference is correct - there is less variance / a lower standard deviation this year - so far.

By using proper statistical techniques, you CAN draw some good inferences and make some reasonable predictions.

For example, run a corelation matrix and do multiple regression to see which are the most important factors in determining success.

Of course, many factors are inter-related - any six year old can work out that there is a positive relationship between goals scored and victory, and a negative relationship between goals conceded and losses.

The thing is that not all these factors are equal, by any means.


Bookies and betting companies understand these issues very well - large parts of probability theory derived from the gambling fraternity.


One factor that is VERY important if you want to finish top 4 or top 1 is goals conceded, and here we are far better than we were lat the same stage last year - the fact that we have conceded the same fewer amount of goals this year compared to last as we have scored fewer goals is not neutral - it is very positive.


Bottom line - we are in much better shape than we were this time last year.

*ends Stats 101 lecture, awaits a torrent of abuse / piss take from the usual suspects