under construction

under construction

December 11, 2009

"Write These Down..."

Prior to the SEC Championship Game, Tom, as he now goes by, Luginbill (Photo) gave four game statistics "that coaches truly care about on a weekly basis."

Luginbill adamantly said to "take notes" and "write these down."  These statistics, according to Luginbill, were pivotal to winning and losing and whichever team, Alabama or Florida, "won" three of the four statistics, would win the SEC title game.  The four statistics:

1) Rushing attempts
2) Average yards per pass attempt
3) 3rd down conversions
4) Turnover margin

As it turned out, Alabama had the advantage in all four statistics and, sure enough, the Tide easily defeated Florida.

So, I decided to take the four statistics and apply them to each of Georgia's 12 games this season.

I thought the Bulldogs would undoubtedly defy Luginbill's stats since Georgia, in my opinion, was fortunate to go 7-5 this year.  In other words, I believed there were more games the Dogs won in '09 where these statistics did not favor them compared to losses, if any, where the four stats where to their advantage.

I thought this would be the case because, for one, Georgia won four games by seven points or less (South Carolina, Arizona State, Auburn, and Georgia Tech) but lost just two games by seven points or less (LSU and Kentucky).  More significantly, the Bulldogs, if I remembered correctly, had the advantage in turnovers--one of the four statistical components--in only two of 12 games (Auburn and Georgia Tech).

The numbers next to Georgia and its opposition corresponds to the statistics 1 through 4 above.  "Same" indicates the Dogs and their opponent had the same statistics for the particular category.

Oklahoma St.: 1, 2, 3, 4

S. Carolina: 1, 4

GEORGIA: 1, 2, 3
Arkansas: 4

GEORGIA: 1, 2, 3
Arizona St.: 4

Lsu: 1, 2
Same: 4

Tennessee: 1, 2, 3, 4

GEORGIA: 1, 2, 3
Same: 4

Florida: 1, 2, 3, 4

GEORGIA: 1, 2, 3
Tennessee Tech: 4

GEORGIA: 2, 3, 4
Same: 1

Kentucky: 3, 4

Georgia Tech: 1, 2, 3

In five of its 12 games, Georgia had a three or more statistical advantage over its opponent--all five were Bulldog wins.  In four of the games, the opponent had the statistical advantage--three of the four were victories by the opposition except the Georgia Tech game.

In eight of nine Georgia games, Luginbill's four-category, statistical determination was credible.  In the other three games, the Bulldogs nor their opponent had the advantage in three or more categories.

As I mentioned yesterday, maybe Luginbill learned a thing or two in the math courses he took during his one year at Tech.

I will add, however, if you add up the number of "won" categories for Georgia and its opponents for the entire season, the Bulldogs had the advantage in only 21, the opponent 24, and they were the same for three.  These results are not unforeseen because of Georgia's year-long turnover woes; however, they are somewhat surprising for a team that finished its season with a winning 7-5 mark.

1 comment:

JasonC said...

When I read the 4 stats, I thought Tech's offense is made for those stats.
As a option/flexbone team, they should win the first stat every time.
They don't try that many pass, so if they can connect on 50%, most of which are big gains, as was the case this year, they can check off number 2 as well.
The fourth stat, turnovers, is the only one that wouldn't favor them, since there is a high probability of a bad pitch, etc. But if they can take care of the ball, it should be easy for them to win 3 of the 4 categories.

I would suggest in the areas where there are splits- 2 categories for each, then you would want to explore the periphery of those categories. For example, if you won the turnover category, but after all of those turnovers, which you got on your side of the field, you punted, it doesn't help as much as getting points of turnovers.

That was one of the problems with UGA this year. Not only did we turn it over, we did it so many times on our side of the field which gave opponents easy points.

Another example of looking closer at the numbers would be in the South Carolina game. I didn't remember SC's run game being effective except for Garcia scrambling and that was the case. Maddox and Giles, who got the bulk of their carries, were largely ineffective. Ergo, does it matter if you run more if you don't do much when you do run it?

Interesting post, though. Thanks for crunching the numbers.