Stats Program Strength

AE 87

Helluva Engineer
Messages
13,026
Okay, so a couple of people have raised questions about our offensive scheme again in the last week or so. By this time, I'd have thought that nonsense would've been put to bed.

As I posted in another thread, our average offensive efficiency (points per drive vs pwr 5 or BCS AQ), has been #6 in the nation, for teams playing more than 2 pwr5 or BCS AQ teams a year, over the past 5 and 8 years:
Let me put some numbers on the point made by @tech_wreck47:

From the 2008 to 2015 seasons, our offense has averaged 2.57 points/drive versus BCS AQ or Pwr 5 opponents. That's 6th ranked offense over the last 8 years. If we look at just the last 5 years (after Gailey's recruits for those who still bring that up), our offense has averaged 2.69 points per drive vs BCS AQ/Pwr 5, which is again ranked 6th over these years.

The offenses ranked higher than us over these periods are as follows:
2008-2015: Oregon (2.98), Baylor (2.82) & Stanford (2.82), tOSU (2.68), Bama (2.61)
2011-2015: Baylor (3.20), Oregon (3.18), tOSU (2.84), Stanford (2.77) and Bama (2.77)

Furthermore, others have made the case that our recruiting hasn't really gone down under Johnson.

So, the facts apparently don't really support your case beyond your initial claim ... you're sorry.

Over that same time period, we've been #50 on average defensive efficiency (ppd allowed vs pwr 5 or bcs aq).

I went ahead and calculated a differential for the 5 yr averages (Offensive PPDvPwr5 minus Defensive PPDvPwr5). Here's a ranking of the top 25 programs by this metric:
upload_2016-4-29_16-33-38.png


While we might find a lot to complain about in terms of W-L, which is of course the stat that ultimately matters most, I think that we still have reason for optimism.

Greater success will still hinge, in my opinion, on making significant improvement in defensive efficiency. If we get a D ranked in the top 40 in DefPPDvPwr5, that would statistically result in a top 20 team most years, given our typical offensive efficiency. That means we'd have a puncher's chance in head to head games.

A top 30 D would get us into top 10 category, and that's with an average O for us. If we combine a top 30 D with a 2009, 2011, or 2014 Offense, we'd be in the statistical range of a top 5 team. That's how close we are.

We don't need a VPI D or the CU D from the last couple of years. We need a D like Pitt's from a few years ago or Duke's from 2013.
 

furant

Jolly Good Fellow
Messages
351
Great analysis. I think that's pretty good company in that list. Thanks for taking the time.
 

jacketup

Helluva Engineer
Messages
1,551
Way to cherry pick a stat. Other posts by you prove that you have no credibility.

Using the stat of points per game, GT has only had one offense in the last 5 years ranked in the top 25, and that was obviously 2014 at number 12. The average ranking over the past 5 years is 43rd (with the benefit of rounding). Other offensive stats such as total offense are similar. FBS only.

As you acknowledge, the won loss record has not been pretty. If you look at all the stats, and not just cherry pick, you understand the symptoms. The disease is something else entirely.

Total defense has had a higher ranking 2 of the past 3 years than total offense. I use 3 years because that is the time Roof has been here, even though he is just now getting his recruits in.
 

tech_wreck47

Helluva Engineer
Messages
8,670
Okay, so a couple of people have raised questions about our offensive scheme again in the last week or so. By this time, I'd have thought that nonsense would've been put to bed.

As I posted in another thread, our average offensive efficiency (points per drive vs pwr 5 or BCS AQ), has been #6 in the nation, for teams playing more than 2 pwr5 or BCS AQ teams a year, over the past 5 and 8 years:


Over that same time period, we've been #50 on average defensive efficiency (ppd allowed vs pwr 5 or bcs aq).

I went ahead and calculated a differential for the 5 yr averages (Offensive PPDvPwr5 minus Defensive PPDvPwr5). Here's a ranking of the top 25 programs by this metric:
View attachment 1575

While we might find a lot to complain about in terms of W-L, which is of course the stat that ultimately matters most, I think that we still have reason for optimism.

Greater success will still hinge, in my opinion, on making significant improvement in defensive efficiency. If we get a D ranked in the top 40 in DefPPDvPwr5, that would statistically result in a top 20 team most years, given our typical offensive efficiency. That means we'd have a puncher's chance in head to head games.

A top 30 D would get us into top 10 category, and that's with an average O for us. If we combine a top 30 D with a 2009, 2011, or 2014 Offense, we'd be in the statistical range of a top 5 team. That's how close we are.

We don't need a VPI D or the CU D from the last couple of years. We need a D like Pitt's from a few years ago or Duke's from 2013.
Thank you for posting, sometimes on this board I get the feeling I'm the only one that feels this way, and it's so simple to understand imo but some how people just don't get it.
 

Milwaukee

Banned
Messages
7,277
Location
Milwaukee, WI
Way to cherry pick a stat. Other posts by you prove that you have no credibility.

Using the stat of points per game, GT has only had one offense in the last 5 years ranked in the top 25, and that was obviously 2014 at number 12. The average ranking over the past 5 years is 43rd (with the benefit of rounding). Other offensive stats such as total offense are similar. FBS only.

As you acknowledge, the won loss record has not been pretty. If you look at all the stats, and not just cherry pick, you understand the symptoms. The disease is something else entirely.

Total defense has had a higher ranking 2 of the past 3 years than total offense. I use 3 years because that is the time Roof has been here, even though he is just now getting his recruits in.

AE is a credible poster. Besides that I do agree with most of your post though.
 

tech_wreck47

Helluva Engineer
Messages
8,670
Way to cherry pick a stat. Other posts by you prove that you have no credibility.

Using the stat of points per game, GT has only had one offense in the last 5 years ranked in the top 25, and that was obviously 2014 at number 12. The average ranking over the past 5 years is 43rd (with the benefit of rounding). Other offensive stats such as total offense are similar. FBS only.

As you acknowledge, the won loss record has not been pretty. If you look at all the stats, and not just cherry pick, you understand the symptoms. The disease is something else entirely.

Total defense has had a higher ranking 2 of the past 3 years than total offense. I use 3 years because that is the time Roof has been here, even though he is just now getting his recruits in.
I think efficiency is what you need to be going by with GT because we eat the clock and take long drives, so of course a teams that gets 3 and 4 extra drives are going to have more yards and a higher score. Compare apples to apples please.
 

zhavenor

Jolly Good Fellow
Messages
468
Way to cherry pick a stat. Other posts by you prove that you have no credibility.

Using the stat of points per game, GT has only had one offense in the last 5 years ranked in the top 25, and that was obviously 2014 at number 12. The average ranking over the past 5 years is 43rd (with the benefit of rounding). Other offensive stats such as total offense are similar. FBS only.

As you acknowledge, the won loss record has not been pretty. If you look at all the stats, and not just cherry pick, you understand the symptoms. The disease is something else entirely.

Total defense has had a higher ranking 2 of the past 3 years than total offense. I use 3 years because that is the time Roof has been here, even though he is just now getting his recruits in.
Ok. Tell us why points per game is a better stat than point per drive.
 

AE 87

Helluva Engineer
Messages
13,026
Way to cherry pick a stat. Other posts by you prove that you have no credibility.

Using the stat of points per game, GT has only had one offense in the last 5 years ranked in the top 25, and that was obviously 2014 at number 12. The average ranking over the past 5 years is 43rd (with the benefit of rounding). Other offensive stats such as total offense are similar. FBS only.

As you acknowledge, the won loss record has not been pretty. If you look at all the stats, and not just cherry pick, you understand the symptoms. The disease is something else entirely.

Total defense has had a higher ranking 2 of the past 3 years than total offense. I use 3 years because that is the time Roof has been here, even though he is just now getting his recruits in.

I'll try one more time, more slowly.

Team A has the ball 20 times in a game and scores a TD on half the drives and punts the other half.

Team B has the ball 10 times and scores touchdowns on every drive.

Who has the better offense?

If you use pts per game, you would answer that they are the same. If you use pts/dr, you'd see Team B was twice as good.

Now, has that sunk in?

OK, in a game, both teams typically have the same number of drives. So, a team that is used to 20 drives a game will not score as many points when they have fewer drives.

Using pts per drive is the right stat, not a cherry picked stat.

Now, in football, the clock stops for incomplete passes, and pass plays are typically faster than rushing plays also. So, teams who primarily rush typically will have fewer drives in a game, especially if their D can't get off the field. That has been the situation at Tech.

I hope this helps you and others understand why the pts per game stat is the wrong stat to compare.
 

kittysniper101

Jolly Good Fellow
Messages
174
Both stats are pretty useless.

W-L trumps.

Ultimately W/L is what matters, but it doesn't do any good at telling how close you are to breaking through. I think AE has demonstrated pretty clearly that we're close to being a consistently strong team. Modest improvements to the defense would greatly increase our rank using his metric, and I would imagine have a similar impact on W/L given how many close games we've lost in past years.

I do worry about the offense if we can't prove that we are able to adjust to teams who truly sell out to stop the run and have talented defenses. Given that we have a way smaller sample size of this and it was a year marred by injuries and attrition I'll go with the larger sample size of years where our offense is the lone bright spot.

I would love to be able to isolate statistically the effect our offense has on recruiting and defensive performance, but that seems impossible. For now, anything we claim for that is total spit-balling whereas most facts say we have a pretty reasonable offense and a barely mediocre defense over the past 8 years.
 

awbuzz

Helluva Manager
Staff member
Messages
12,104
Location
Marietta, GA
W L matters...

Overall offense has been "going".

Other side of coin is defensive points per drive.

Combined they should correlate with W L record.
 

Augusta_Jacket

Moderator
Staff member
Messages
8,099
Location
Augusta, Georgia
Way to cherry pick a stat. Other posts by you prove that you have no credibility.

Using the stat of points per game, GT has only had one offense in the last 5 years ranked in the top 25, and that was obviously 2014 at number 12. The average ranking over the past 5 years is 43rd (with the benefit of rounding). Other offensive stats such as total offense are similar. FBS only.

As you acknowledge, the won loss record has not been pretty. If you look at all the stats, and not just cherry pick, you understand the symptoms. The disease is something else entirely.

Total defense has had a higher ranking 2 of the past 3 years than total offense. I use 3 years because that is the time Roof has been here, even though he is just now getting his recruits in.

Speaking of cherry picking a stat. Total PPG is just about useless. You get a lot of group of 5 schools that rate high there because they play no one. You know who usually ranks around 30th in points per game? Alabama. Ask them if they care.
 

CuseJacket

Administrator
Staff member
Messages
19,555
Way to cherry pick a stat. Other posts by you prove that you have no credibility.

Using the stat of points per game, GT has only had one offense in the last 5 years ranked in the top 25, and that was obviously 2014 at number 12. The average ranking over the past 5 years is 43rd (with the benefit of rounding). Other offensive stats such as total offense are similar. FBS only.

As you acknowledge, the won loss record has not been pretty. If you look at all the stats, and not just cherry pick, you understand the symptoms. The disease is something else entirely.

Total defense has had a higher ranking 2 of the past 3 years than total offense. I use 3 years because that is the time Roof has been here, even though he is just now getting his recruits in.
Since I've seen this hit and run from you before, without responding to the counterpoints each time, should I consider this a troll going forward?

If not, my opinion is that you're demonstrating a lack of understanding of the merits of ppdrive vs. ppgame. Ppdrive is not perfect, but imo it's far better than the stats you cherry picked.
 

AE 87

Helluva Engineer
Messages
13,026
Both stats are pretty useless.

W-L trumps.

It depends on what question you are asking. I put a caveat in my op about W-L, but it doesn't allow you to look at o and d separately. It also doesn't differentiate between close and blowout losses.

So yes, W-L answers the question of whether you have been a championship team but doesn't really help with the question of where you are on the road.

Think of it this way, if I didn't tell you the metric, but just gave you this ranking of the top 5 and top 10 teams over the last 5 years, would you agree that it seems reasonable?

Also, as a general rule, teams who on average outscore pwr5 opps by more pts will have won more games. So the stats aren't independent.
 

awbuzz

Helluva Manager
Staff member
Messages
12,104
Location
Marietta, GA
It depends on what question you are asking. I put a caveat in my op about W-L, but it doesn't allow you to look at o and d separately. It also doesn't differentiate between close and blowout losses.

So yes, W-L answers the question of whether you have been a championship team but doesn't really help with the question of where you are on the road.

Think of it this way, if I didn't tell you the metric, but just gave you this ranking of the top 5 and top 10 teams over the last 5 years, would you agree that it seems reasonable?

Also, as a general rule, teams who on average outscore pwr5 opps by more pts will have won more games. So the stats aren't independent.
Bottom left be there is a high correlation to W L record and the combined offensive and defensive PPD combination.

What is so difficult to understand????
 
Top