Stat manipulation for tomorrows game (Long)

Got to thinking last night, so I decided to look at some stats (against my own vows) for the game this weekend, and noticed some interesting trends that not only show that both teams are incredible, but that LSU has played a better group of teams than bama. The “stats” show a wash for the most part, but the argument that statistical rankings based purely on points/yards/etc. is a poor way to analyze data. As an engineer that must analyze data daily to optimize manufacturing lines, I know that you must be able to normalize the data, and relate it to a baseline to make heads or tails of it. So here is my analysis to show that both teams are good and that tomorrows matchup SHOULD be one for the ages. I know I will be bashed for the length, so bring it on..

Scoring Defense
bama-7
LSU-12
Scoring Offense
LSU-39
bama-39

Opponent Total offense rank-average (average scoring)
LSU-75 (28.3)
bama-95 (23.2)
Opponent Total defense rank-average (average scoring)
LSU-50 (23.3)
bama-51 (23.5)

Points scored against opponents above their average (average of all games)
LSU - 16
bama - 16
Points allowed against opponents LESS than their average (average of all games)
LSU – (-17)
Bama – (-16)

Assumptions you can make from these stats/laymans terms:
So, even though LSU is allowing 5 more points per game, it is against a better group of teams that put up more points (5) against their opponents than bama’s opponents.

Against the opponents “D”, which are ranked very similar (50 vs. 51), both teams are putting up just as many points (39pts vs. 39pts). So total Differential in PPG vs. PPG allowed normalized for opponent quality is within a half of a point (32.7LSU vs. 32.2bama)

Now taking it even further, AT HOME, bama’s opponents have been ranked 92 on offense (23.5ppg), and 56 on defense (25.2). LSU when playing away from Tiger Stadium has against an average Offensive ranking of 49 (33.3 ppg), and Defense ranking of 43 (22.5ppg). Average HOME win for bama was 40-5, and average ROAD win for LSU was 36-15. Again normalizing the DIFFERENTIAL point total against the opponents AVERAGES allowed and scored, there is only a slight bama edge of 32.4 vs. 31.6. This indicates that LSU is just as potent on the ROAD, as bama is at HOME.

And of course between our common opponents (uf/ut) the results are similar again. 32/33 point differential with LSU scoring slightly more points 40 vs. 39, and both teams allowing only 9 ppg.

Bottom line, you can twist stats pretty much any way you want to make them look good for your purpose. I am sure if we wanted to break it down into SEC games/Night games/Wearing whatever uniforms/Moon Phases, we could spin the results either way. So who is the better team? On paper, with adjustments for variables, and normalized we are looking at two VERY good teams that are very equal when put side to side, and considered as a TEAM. This is the problem with statistics unfortunately. It utilizes averages/margins/deviations/etc., but does not take into consideration individual matchups, weather, atmosphere, and other personal factors. So based on all of this, I don’t think I will bet on the game, but it will be a heck of a game to watch. That being said.. I can't wait for tomorrow night!! GEAUX TIGERS!!!