|
KCFR Statistics Page = March 10, 2006
Many college football fans may find the following
stats interesting and useful. Some of you I may
have to explain them further if you wish. Many
of you won't give a hoot or may think that I am
mad. Whatever, I hope that someone out there can
get something out of them. Let's start chewing.
*************************************************
Following is a six-season study of season average
stats for 58 BCS teams from 2000 through 2005.
These teams are the current members of the ACC,
Big 10, Big 12, Pacific 10, SEC plus Notre Dame.
That is a total of 348 sets of season averages,
which include conference playoffs and bowls.
Each statistic category of 348 samples has been
sorted freely for a 6 year mean average result.
Then each category is assigned an adjusted max
and adjusted min. Season results for each team
category is then converted to 1.000 (best) down
to 0.000 (worst) for comparison to the team's
final W/L percentage record. The difference of
of the two categories is a positive deviation,
and measures which categories are consistently
closer to W/L Pct.
Category * 6yrMean * AdjMax * AdjMin * DevAvg
---------------------------------------------
W/L Pctg * 000.563 * 1.0000 * 0.0000 * 0.0000
ScorMarg * +3.634 * +25.13 * -24.02 * 0.0753
ScorPctg * 000.536 * 00.744 * 00.267 * 0.0715
NetYdPct * 000.515 * 00.623 * 00.376 * 0.1061
Turnover * +1.509 * +18.50 * -20.35 * 0.1551
OffPts/G * 27.540 * 40.967 * 10.266 * 0.1202
RushYd/G * 157.829 * 247.85 * 42.016 * 0.2000
PassYd/G * 223.951 * 321.91 * 97.925 * 0.2089
DefPts/G * 23.906 * 10.893 * 40.646 * 0.1144
RushYd/G * 143.160 * 61.942 * 247.65 * 0.1422
PassYd/G * 216.499 * 145.65 * 307.65 * 0.2053
---------------------------------------------
Yards per Rush Attempt and Pass Attempt were
compiled but not listed here as their deviation
results were very close to their Yards per Game.
As you can see, Scoring Percentage is closer in
correlation than Scoring Margin. Scoring Pctg
is a team's points vs total points in the game.
After all, isn't a 20-0 win more impressive than
a 50-30 win? Both are 20 point margins. Score
margin average is a close second in correlation.
Net Yard Margin is 3rd followed by Defensive Pts
Allowed. I use Net Yard Margin in my predictive
formula, and when combined with a Turnover score
the result is very close in correlation with the
W/L Percent. I believe that those who use the
Scoring Margin heavily in their formula should
convert to Scoring Percentage X Avg TotPts/Game.
RETURNING STARTERS, OFFENSE, DEFENSE, COMBINED.
-----------------------------------------------
#R QB #S Points Yards OFFENSE
11 Y 2 +5.633 +41.6
10 Y 7 +1.843 +00.8
9. Y 19 +4.828 +16.2
9. N 3 -0.289 +19.0
8. Y 38 +2.510 + 8.5
8. N 8 +0.381 -26.3
7. Y 55 +1.108 + 5.4
7. N 20 -0.690 - 9.6
6. Y 39 +1.098 +13.9
6. N 29 -2.883 -19.2
5. Y 17 -2.818 + 3.2
5. N 24 -3.412 -26.7
4. Y 7 -0.886 +19.6
4. N 14 -4.141 - 1.1
3. Y 2 -4.600 + 4.5
3. N 4 -0.800 +41.7
2. N 1 -4.400
1. Y 1 -3.700
--------------------------
ALL Y 187 +1.326 +6.55
ALL N 103 -2.356 -8.64
--------------------------
Average No of Ret Offense Starters = 6.56
#R = Returning Starters
QB = Quarterback return Yes or No
#S = number of samples in study
Points = increase or decrease in offensive pts
Yards = increase or decrease in offensive yds
The above results show a clear indication of
the importance of the quarterback returning.
The number of starters relates somewhat to the
diffences in scoring, much less in yards gained.
It would appear to be about 0.5 points per the
number of starters difference from the 6.5 avg.
DEFENSIVE STARTERS. These results are about the
same (minus the QB study of course). Six or
seven returning starters showed very little
change over the next year. Eight or nine show
about 70% of the teams improved. Four or five
show 70% of the teams were worse.
COMBINED STARTERS RETURNING.
19 Starters = +7.30 Scoring Margin
18 = +4.48
17 = +3.51
16 = +3.50
15 = +3.23
14 = +2.14
13 = +0.77
12 = -0.56
11 = -1.01
10 = -7.67
9 = -7.57
8 = -6.04
6 = -13.40
187 samples from above with quarterback return
= +2.00 points average. 103 samples with no QB
= -3.50 points average. The same study as above
was done with conference games only assuming
that equally rated schedules would produce a
fairer result. The output was about the same,
with the returning quarterback enjoying +1.63
points per games vs -3.31 for new quarterback.
|
