- Joined
- Oct 18, 2009
- Messages
- 61,115
I would be interested to know you reaction to your ypc rush, ypa pass, and YPP ratios if you exclude 2009.
Why exclude 2009? (curious)
I would be interested to know you reaction to your ypc rush, ypa pass, and YPP ratios if you exclude 2009.
Why exclude 2009? (curious)
Hey I got a great idea, let's pick and choose games and report stats that are statistically even less valid than looking at the whole period of 2008-2011 games.My point has been that the data cyp provided doesn't support statistically significant improvement for teams with extra-prep compared to teams without extra prep when normalized by to their average rush and pass defense.
In 2009, the only two teams who had extra prep were apparently Miami and Iowa. Now, our rush offense typically outperforms an opponents yds/carry so the ratio of actual yds/carry to a D's ave yds/carry is typically over 1, often well over 1. It is less than .9 only twice: Miami's Home games in 2009 at .69 and in 2011 at .68. The 2009 game is calculated as 1 of 16 for the extra-prep stat and 1 of 32 for the non-extra prep. The 2009 Iowa game was also in Miami and they crushed our passing game compared to their expected Pass-D. So, I thought it would be interesting to see how the numbers stacked up without 2009.
As I've said, I think the normalization of the GTypc/DEFypcAv and GTypa/DEFypaAv are interesting statistics. If you take this ratio from the no-extra-prep games as the baseline, then you can compare the ratio from the extra-prep games as a %change. (Remember, imo, the differences are insignificant since they typically fall within a half st dev of the no-extra-prep ratios). Still, here's what you get.
Teams with Extra Prep compared to those without:
2008-2011
Rush: 1% Worse
Pass: 27% Better
YPP: 10% Better
2008, 2010-2011
Rush: 7% Worse (4% Worse w/o 2011 Miami)
Pass: 18% Better (20% Better w/o 2011 Miami)
YPP: 4% Better (6% Better w/o 2011 Miami)
2010-2011
Rush: 15% Worse (11% Worse w/o 2011 Miami)
Pass: 11% Better (14% Better w/o 2011 Miami)
YPP: 2% Worse (1% Better w/o 2011 Miami)
Now, it should also be noted that 2008 was the only year in which we averaged more passes in the games for which the other team had more prep than the games in which they didn't. So, it's not simply the fact that with more prep they are forcing us to throw it more and worse.
I still contend that the data does not support the assertion that teams with extra prep will perform better against us.
Hey I got a great idea, let's pick and choose games and report stats that are statistically even less valid than looking at the whole period of 2008-2011 games.
Also your idea of somehow deducing statistical significance from standard deviation is way off. You can definitely have statistically significant differences under 1 or half standard deviation.
Since I have done more analysis, here is an update.
To summarize the different stats presented, with extra time against Tech in the period of 2008-2011 BCS games:
1. GT offense does 11% worse in the total offense (total yards) stat after adjusting for opponents' defense.
2. GT offense does 9% worse in the yards per play stat after adjusting for opponents' defense.
3. GT win-loss record is worse. (23-8 vs. 4-11)
3a. GT win-loss record is worse when GT is favored to win. (18-3 vs. 2-6)
3b. GT win-loss record is worse when GT is not favored to win. (4-6 vs. 2-5)
4. GT win-loss record is worse against the spread. (18-11 vs. 5-9)
5. GT scoring differential is worse after adjusting for the Vegas spread differential. (doing 3.39 points better than the line vs. doing 7.66 points worse than the line)
https://docs.google.com/spreadsheet/pub?key=0Ar8cqnkh36RRdENBQk1fRTYyRDFIVTdOa1Q0cXNrbGc&gid=3
My reasoning is slightly different, that we have an advantage with our unconventional offense and the advantage gets lessened with extra prep against the offense.Hey, I got an idea. Let's just use common sense. It goes like this:
Does it help a team to have extra prep time for an opponent? Sure it does. Is that true for any team? Pretty much. Does it help a team a bit more to have extra time when the opponent's system is rather unconventional? Sure it does.
See how easy it is?
:wink:
So has anyone run these number for other teams? In other words, it's pretty well established that extra prep time is an advantage against us, but is it any more/less of an advantage against other teams?
Agreed 100% and thank you for your earlier posts too.This has been covered by others but I think it bears repeating.
Those protesting any advantage in prep time versus other teams are viewing the assertion as validation that teams have "figured out" the triple option. The idea that there is a "blueprint" to stop it has been put out in the media several times. But the idea that we have an advantage with shorter prep time is NOT the same thing as anyone figuring it out.
I'm fact, my last statement is the whole point. We have an added advantage with short prep time that we lose with longer prep time. The whole thing supposes a competent DC who can use that time effectively to prepare better and athletes who can execute the plan. And it still comes down to execution on both sides. It is a tendency, not a rule. We can win even when other teams have the extra time and we can lose if they do not have the extra time. Conflicting anecdotes do not disprove the overall tendency.
We just have a little extra advantage with short prep time that we lose with longer prep time. With longer prep time we are more on an even keel with all the other teams, not at a disadvantage. The same would be true for any offense that is significantly different than the offenses that most other teams run.
If suddenly all of the NCAA switched to a similar offense as ours the difference in advantage due to prep time would probably diminish or even disappear. Yet some teams would still be more or less successful and there would still be counter examples.
P.S.
The other factor mentioned in the past was the penalty teams pay for preparing for us excessively. Lots of teams who are more successful against us fall off in the following week, presumably because they lost time prepping for a more traditional offense.
good start sam, could you share list of games and the W-L records? Google docs is free.
btw, those numbers look very low. it seems like pretty good bet to bet against those teams.
Here’s the website I used to get the wins, loses, and pushes against the spread: http://www.statfox.com/cfb/gamelog~teamid~georgia+tech.htm . Here’s the link to google docs: https://docs.google.com/spreadsheet/ccc?key=0AjQT-u2MpDhOdExVbExFRDRRbVN2UVc2SjdZY2FucUE , I hope this works I’ve never loaded anything onto google docs.
On the website, you can find each year’s score and against the spread (ATS) column under the “Game Log” link under the teams names.
I had to set up a matrix showing teams who had extra time and who they played. Then I counted the wins (W) and loses (L) as opposites for the opponents to see how they fared against the spread and teams with extra time (I didn't count pushes). In my first numbers I must of missed something, but here’s a corrected spreadsheet. Also, it’s only as good as the data I put into it, it is possible I could have missed a game here or there.
To summarize, win percentage against the spread vs. teams with extra time, ACC games only 2008-2011:
ACC 43%
VT 55%
UVA 25%
GT 44%
UM 57%
UNC 29%
DU 30%
CU 40%
FSU 29%
NCSU 71%
WF 60%
BC 20%
MD 43%
This has been covered by others but I think it bears repeating.
Those protesting any advantage in prep time versus other teams are viewing the assertion as validation that teams have "figured out" the triple option. The idea that there is a "blueprint" to stop it has been put out in the media several times. But the idea that we have an advantage with shorter prep time is NOT the same thing as anyone figuring it out.
I'm fact, my last statement is the whole point. We have an added advantage with short prep time that we lose with longer prep time. The whole thing supposes a competent DC who can use that time effectively to prepare better and athletes who can execute the plan. And it still comes down to execution on both sides. It is a tendency, not a rule. We can win even when other teams have the extra time and we can lose if they do not have the extra time. Conflicting anecdotes do not disprove the overall tendency.
We just have a little extra advantage with short prep time that we lose with longer prep time. With longer prep time we are more on an even keel with all the other teams, not at a disadvantage. The same would be true for any offense that is significantly different than the offenses that most other teams run.
If suddenly all of the NCAA switched to a similar offense as ours the difference in advantage due to prep time would probably diminish or even disappear. Yet some teams would still be more or less successful and there would still be counter examples.
P.S.
The other factor mentioned in the past was the penalty teams pay for preparing for us excessively. Lots of teams who are more successful against us fall off in the following week, presumably because they lost time prepping for a more traditional offense.
Maybe we should use Texas Tech under Mike Leach as a control? His offense was a little unconventional, so it should show the same pattern.
No. I am not seeing it as a evidence of "figuring it out." My protest is that the data doesn't support the claim.
Only an idiot would cite, or be persuaded by, W-L or total yards as evidence of a team being able to prepare for our "unconventional offense." W-L and Total Yards are not independent from Defense. Therefore, they are not rationally a good indicator of team's prep "against our uncoventional offense" for which cyp offered them.
Then, once you look at the actual offense vs defense statistics, you see that they don't really support the claim either. For the last two years, teams with an extra prep did 15% worse against the rush than teams without it. Would any reasonable person conclude that extra-prep now hurts you against our rushing attack? It's true that passing yrd/attempt seem to indicate an improvement for teams with extra-prep, but the stdev for the ypa ratio is huge (almost twice the stdev for rush-ratio), and the difference between the ypa ratio for teams without prep and teams with prep is statistically small.
Moreover, it's not a trend. If you look at the ypa ratio for the individual years, you'll see that teams with extra prep had a 34% improvement in ypa ratio, in 2009, 54% improvement, in 2010, they did 18% worse, and in 2011, they had a 28% improvement. That's not a trend.
Philosophically, I'm sympathetic to the theory just as Cyp has articulated it. However, I don't think the data that he has presented has supported the theory.
My reasoning is slightly different, that we have an advantage with our unconventional offense and the advantage gets lessened with extra prep against the offense.
Six of one, half dozen of the other. Some things are so evident that they are simply warranted without extensive proof. IMO, this is one such item.