More Interesting Stats

JTS

JTS

Jolly Good Fellow
Joined
Aug 10, 2003
Messages
1,969
GT has only won 2 games during the Gailey era when the opposing team has had more rushing yards.

So far this season:

- 1.03% of our passing attempts have resulted in TDs which is the worst percentage of 119 D-IA teams.

- 1.55% of our passing attempts have resulted in INTs which is the 11th best of 119.

- On offense, we have gained 5.81 yds per pass attempt which is 92nd out of 119.

- 0.98% of passing attempts against us have resulted in INTs which is 117th out of 119.

- 2.45% of passing attempts against us have resulted in TDs which is 10th best of 119.

- On defense, we have allowed 6.11 yds per pass attempt which is 47th of 119.
 
- 1.55% of our passing attempts have resulted in INTs which is the 11th best of 119.
Click to expand...

Yay for a good passing stat!

- On offense, we have gained 5.81 yds per pass attempt which is 92nd out of 119.
Click to expand...

That's wretched. I mentioned earlier that that stat has the highest correlation with superbowl wins of all stats.

- 0.98% of passing attempts against us have resulted in INTs which is 117th out of 119.
Click to expand...

That is clearly not Tenuta Football.
 
It cuts both ways:

GT ranks #1 in the nation in kickoff return defense (15.62 ypr).

GT ranks #1 in the nation in net punting (40.7 ypp).

GT ranks #2 in the nation in turnovers given away with 6 (uGA is #1 with 5).

GT ranks #2 in the nation in tackles for loss, averaging over 9.5 per game. Of the 454 plays the defense has played, 177 have gone for negative yards or no gain!

GT ranks #5 in the nation in sacks with 3.86 per game.

GT ranks #10 in kickoff returns (25.9 ypr).

GT ranks #13 in the nation in scoring defense @ 16.3 ppg.

GT ranks #13 in the nation in rushing defense (82.6 ypg), and currently yeilds the fewest rusing yards per game of any team since 1948.

GT ranks #35 in punt returns (11.6 ypr). (editorial note: maybe our "playmakers" ain't so bad after all. ;) )

GT has won 17 of last 18 and 23 of last 25 games when leading at the end of the 3rd quarter.
 
GTKyle said:
Tenuta football is hide your crappy coverage secondary with blitzes. That stat is pretty much textbook Tenuta ball.
Click to expand...

Unless it isn't.

2006: T-52 with 13 interceptions
2005: 6th with 21
2004: T-75 with 10
2003: T-56 with 13
2002: T-79 with 11

While interceptions doesn't necessarily make a pass defense, saying that it's classic Tenuta defense to not pick some passes off is somewhat off-base.

That said, I do agree the secondary has always been the weakest feature of the defense. But at least they were better at picking off passes than they are this year.
 
GTKyle said:
Tenuta football is hide your crappy coverage secondary with blitzes. That stat is pretty much textbook Tenuta ball.
Click to expand...

That stat us unsuccessful Tenuta football. Successful Tenuta football actually hides the crappy coverage. Hehe.

Although this weekend should help that stat a bit.
 
beej67 said:
That's wretched. I mentioned earlier that that stat has the highest correlation with superbowl wins of all stats.
Click to expand...

Almost certainly the best correlated stats involve scoring. Perhaps it is the best-correlated stat that does not include scoring?

I looked at this a few years ago, and for the college game, the best-correlated stat to winning percentage not related to scoring was Yardage Differential (Yards Gained per Game minus Yards Allowed per Game).

Yards per Pass Attempt was 2nd-best correlated, however.

See:

http://www.geocities.com/mm42/2000summary.htm

and

http://www.geocities.com/mm42/2000season.htm
 
mm42 said:
Almost certainly the best correlated stats involve scoring. Perhaps it is the best-correlated stat that does not include scoring?

I looked at this a few years ago, and for the college game, the best-correlated stat to winning percentage not related to scoring was Yardage Differential (Yards Gained per Game minus Yards Allowed per Game).

Yards per Pass Attempt was 2nd-best correlated, however.

See:

http://www.geocities.com/mm42/2000summary.htm

and

http://www.geocities.com/mm42/2000season.htm
Click to expand...


Why not do turnovers? This would have to be #1 I would think which is why I tend to support Gailey up to this point. I think he gets how to win at GT; he just doesn't get how to win in genreal in college. Risk-aversion will win you quite a few games but only enough to be called mediocre.
 
thwg said:
Why not do turnovers? This would have to be #1 I would think which is why I tend to support Gailey up to this point. I think he gets how to win at GT; he just doesn't get how to win in genreal in college. Risk-aversion will win you quite a few games but only enough to be called mediocre.
Click to expand...

Turnover margin is on there, about 10 spots down.

Also, looking that this makes me even more angry that the BCS told the computer guys to remove margin of victory from the calculation.
 
ASimPerson said:
Turnover margin is on there, about 10 spots down.

Also, looking that this makes me even more angry that the BCS told the computer guys to remove margin of victory from the calculation.
Click to expand...

interesting, i guess this is the reason Gailey is seen as mediocre. TO would have much more importance in the pros where everyone team has very similar talent.
 
thwg said:
interesting, i guess this is the reason Gailey is seen as mediocre. TO would have much more importance in the pros where everyone team has very similar talent.
Click to expand...

Yeah. In college turnovers don't matter much unless you're picking up a ton of them or you're committing a ton of them. Outside of that, you need to score some points.
 
I'm afraid this is just baseless rationalization.

For the NFL season so far, turnover margin correlates only marginally with winning percentage, at .50

Beej's Yards per Passing Attempt, for instance, has a higher correlation at 0.62

I will look at some past seasons if I get a chance, but turnovers are just such a small fraction of the game that it's not surprising to me they aren't very highly correlated with wins.
 
I'm not going to try to get into the mathematics of it, but I think TOs matter more depending on the style of play a team uses. I equate it to basketball. If you're running a ball control, set piece offense TOs will kill you. But if you're running a wide open, pressure attack they don't matter as much. Think Lethal Weapon 3. We turned the ball over quite a bit in spurts, but it didn't matter because we were going to get many more chances to make them up.

Chan plays a more conservative, ball control style of game so TOs are killers in that we simpy aren't going to strike quick to up the # of chances we have. But a WVU or Oregon or somebody can make up for more mistakes because of how quick they can strike. Neither wants TOs, but one can deal with them better.
 
ncjacket said:
I think TOs matter more depending on the style of play a team uses.
Click to expand...

A plausible hypothesis, but I believe nothing until I see the numbers to back it up.

When I started looking into those correlations, it was my firm belief that a good defense was superior to a good offense, and that the stats would bear me out. Turned out I was wrong, the stats seem to be either neutral or slightly tilted towards offense as indicators of success.
 
Over the past 5+ seasons, Scoring Defense is slightly more correlated with winning percentage (0.85 to 0.83).

If you just look at the Top 25 teams in winning percentage over the past 5+ seasons, correlation leans even more towards Scoring Defense (0.49 to 0.43).
 
Last edited:
JTS said:
Over the past 5+ seasons, Scoring Defense is slightly more correlated with winning percentage (0.85 to 0.83).
Click to expand...

I was referring to the non-scoring stats. Scoring rolls a lot of stuff together. Scoring defense includes not giving up TDs or field position on special teams, not giving up TDs by the other team's defense, ball control by your own offense that limits the other team's scoring opportunities, etc.

I also don't think the difference between .85 and .83 is statistically significant.

Non-scoring stats are more "pure", i.e. yards per passing attempt is really a direct measurement of how well your offense is performing with minimal influence from these outside factors. It would be better if the NCAA counted sacks against passing attempts instead of rushes, though.
 
You must log in or register to reply here.
Back
Top