In partnership with CBSSports.com
You have no favorite boards.
As many of you know, I track the effectiveness of predictions for all UK games, and this season is no different. Through 9 games, the variance of actual margins from the game predicted margins have been -4, 32, -3, -15, 5, 3, -37, 18, and -8. The outlying high and low values are not unusual, in fact occur in almost all years with all teams per my experience. The team is defined by the grouping of games in the middle of those outliers, e.g. the -4, 03, 5, 3, -8, -15, and 18.
What is unusual with this year's data is how these variances occur, home v away. Nearly every other year, the negative and positive variations are equally distributed between home and away games, but this year, the home variations have been 32, -3, 5, 3, and 18, and the away variations have been -4, -15, -37, and -8.
This data describes a team that outperforms its average at home and under performs its average on the road. This raises some questions:
Is that an earmark of a young team?
Is this simply an anomaly within the data that is a matter of happenstance, and not an indicator of any reliable trend?
Today's Vegas line is 7 points. Sagarin places it at 4 1/2 points. My model makes this a 12 point game. Whether it is 4 1/2, 7 or 12 is not the issue, but the average home variation this season has been 11 points. If the outliers are not used for the home calculation, the average variation is 8 1/2 points at home. If the answer to this question is that the home-away difference is real and due to the nature of this particular team, then today's game is clearly in play for this team.
One last point, I do not see any significant correlation between the game to game variations and the pattern of injuries that this team has sustained. Take that for what its worth.
My hope is that this young team continues its growth today, and beats Vanderbilt by doing what it has done so far this season at home, out perform its own tendencies at Commonwealth Stadium.
I'm impressed by your knowledge of stats... you're like the Rain Man of sports. lol
Seriously, it's interesting how one becomes so addicted to numbers, but I can see how it pays off at times...your predictions are amazing!
Keep up the great work Professor
Interesting stuff. One question. You are probably taking this into account but do these numbers change if you perform the same analysis on the opposition? In other words, does this even matter? If this is solid you may be my new betting line.
Stats don't lie do they Professor? Good stuff
In football, I track these numbers for UK, all UK non-conference opponents, and all SEC teams. I believe the fundamental analysis is solid. I do the same with basketball, and my confidence in the basketball model is very high, been doing it longer, and much larger database available for each team, each season etc.
I don't recall seeing such a difference, home v away for UK football in the years I have been tracking the football numbers. I can't tell if it is simply some weird incidence of events, or whether this team is really that much better at home v away from Commonwealth.
As for the numbers being honest. Numbers are numbers. They are what they are. Some people use numbers to prove subjective points of view. Others use numbers more objectively. I believe I fall in the latter group, but those who don't have any respect for the numbers group everyone in the former group in an attempt to discredit the numbers they oppose.
247Sports In partnership with CBS Sports