tazthenomad
Club Member
Using this stat to predict wins by averaging (or saying that you would have to have an average of 10 or better to indicate you are playing "playoff caliber" football sure doesn't seem to work very well. Our average of all the scores for our 14 games is a prediction of 7.86 wins for the season. I'm guessing that we will have at least 9 wins but more importantly if we do make the playoffs and the 10 wins that the statistic average for the season will not catch up.
Secondly, if you want to use a time-series to see if we have improved since the first half of the season and use this statistic you will see that for the last 4 weeks in a row we have been above the average for the first 8 weeks (7.55 predicted wins). One of the standard rules for spotting a shift in a process is a "run" of 5 or more points either above or below the average. We are on the verge of statistically saying that we are improved over the first half of the season if we have another statistic above 7.55 for this week's game. I am not sure I think this is a good model, but if you accept it as a measure of how good a team is then those of us who see improvement will be able have statistical justification for our opinion as well.

Secondly, if you want to use a time-series to see if we have improved since the first half of the season and use this statistic you will see that for the last 4 weeks in a row we have been above the average for the first 8 weeks (7.55 predicted wins). One of the standard rules for spotting a shift in a process is a "run" of 5 or more points either above or below the average. We are on the verge of statistically saying that we are improved over the first half of the season if we have another statistic above 7.55 for this week's game. I am not sure I think this is a good model, but if you accept it as a measure of how good a team is then those of us who see improvement will be able have statistical justification for our opinion as well.
