Defense: Is 2006 better than 2003? - cougarguard.com — unofficial BYU Cougars / LDS sports, football, basketball forum and message board

pelagius · 12-13-2006, 06:52 PM

There has been a lot of talk about the 2006 defense. Is it the best BYU defense since Bronco has been at BYU? If you just look at the raw numbers in terms of points allowed then the 2006 defense looks very good:

Thus on average the 2006 defense has allowed about 10 points per game less than the 2003 defense. The difference in the medians is smaller but still prety large: 5.5 points per game. However, BYU played a much tougher SOS in 2003. Here's a quick comparison of the average Massey year end rank of the opponents:

Code:

year  rank
2003  92.6    
2006  47.6

Thus the raw graph above is really not a fair comparison. In order to form a better comparison I made a matched sample: I matched games in 2003 with games in 2006 based on the opponents year end ranking. I got 7 good matches:

Code:

year team         allowed rank year team         allowed rank
2003 Georgia Tech      13   34 2006 Arizona           13   37
2003 Stanford          18   77 2006 Wyoming            7   74
2003 Colorado St       58   49 2006 Tulsa             24   53
2003 Wyoming           13   88 2006 New Mexico        17   85
2003 Boise St          50   14 2006 TCU               17   23
2003 Notre Dame        33   46 2006 Utah              31   52
2003 Utah               3   24 2006 Boston College    30   24

I then redid the average and median points allowed using the matched sample:

The differences are smaller now. The 2006 defense still performed better on average (although the difference is statistically insignificant) but the medians are very similar. If you forced me to pick I would give an edge to the 2006 defense mainly because of consistency. The 2006 defense never had the disaster game(s) that the 2003 defense did even when playing good teams (maybe that has to do with the switch to the 3-4).

Notes:

(1) The sample sizes are small here so the analysis doesn't have a lot of statistical power.

(2) I am using year end rank; clearly it would make sense to use an offensive rank for the sample matching but I don't have a good measure of that.

(3) Taking into account homefield advantage is probably a good idea. I think I should redo the numbers doing a +3/-3 road/home adjustment. I'll try to do that later.

Indy Coug · 12-13-2006, 08:33 PM

Please take into account that BYU allowed 8 TDs in 2003 on turnover and kick returns and 0 TDs in 2006. That doesn't even begin to account for the numerous very short fields allowed by BYU in 2003 for both of the above as well. Again, those things happened very rarely in 2006.

BYU's 2003 scoring defense is extremely skewed relative to their other major defensive measurables.

pelagius · 12-13-2006, 09:11 PM

Quote:

Originally Posted by Indy Coug

Please take into account that BYU allowed 8 TDs in 2003 on turnover and kick returns and 0 TDs in 2006. That doesn't even begin to account for the numerous very short fields allowed by BYU in 2003 for both of the above as well. Again, those things happened very rarely in 2006.

BYU's 2003 scoring defense is extremely skewed relative to their other major defensive measurables.

All very true but it doesn't really change things that much. The matched sample includes 5 of those scores because the following games aren't included:

USC: 1
SDSU: 2

Still, there are 5 other occurences (Boise St = 2, Georgia Tech = 1, Wyoming = 1, and Col St = 1). On the other hand, the 2006 defense gave up some meaningless touchdowns at the end of some of the games in the matching sample. Still, you make a good point. Taking out those 5 scores for the matched sample leads to the following:

Code:

        Mean    Median
2003   21.86        18
2006   19.86        17

The median is unchanged and the average is now a few points lower for 2003.

pelagius · 12-13-2006, 10:19 PM

Also, if I do a homefield advantage adjustment +3/-3, keep Indy Coug's scores off turnovers adjustment, and take out the overtime score for Boston College, I get the following for the two years:

Code:

Year    Mean  Median
2003   20.14      18
2006   15.43      17

jay santos · 12-13-2006, 10:32 PM

Quote:

Originally Posted by pelagius

There has been a lot of talk about the 2006 defense. Is it the best BYU defense since Bronco has been at BYU? If you just look at the raw numbers in terms of points allowed then the 2006 defense looks very good:

Thus on average the 2006 defense has allowed about 10 points per game less than the 2003 defense. The difference in the medians is smaller but still prety large: 5.5 points per game. However, BYU played a much tougher SOS in 2003. Here's a quick comparison of the average Massey year end rank of the opponents:

Code:

year  rank
2003  92.6    
2006  47.6

Thus the raw graph above is really not a fair comparison. In order to form a better comparison I made a matched sample: I matched games in 2003 with games in 2006 based on the opponents year end ranking. I got 7 good matches:

Code:

year team         allowed rank year team         allowed rank
2003 Georgia Tech      13   34 2006 Arizona           13   37
2003 Stanford          18   77 2006 Wyoming            7   74
2003 Colorado St       58   49 2006 Tulsa             24   53
2003 Wyoming           13   88 2006 New Mexico        17   85
2003 Boise St          50   14 2006 TCU               17   23
2003 Notre Dame        33   46 2006 Utah              31   52
2003 Utah               3   24 2006 Boston College    30   24

I then redid the average and median points allowed using the matched sample:

The differences are smaller now. The 2006 defense still performed better on average (although the difference is statistically insignificant) but the medians are very similar. If you forced me to pick I would give an edge to the 2006 defense mainly because of consistency. The 2006 defense never had the disaster game(s) that the 2003 defense did even when playing good teams (maybe that has to do with the switch to the 3-4).

Notes:

(1) The sample sizes are small here so the analysis doesn't have a lot of statistical power.

(2) I am using year end rank; clearly it would make sense to use an offensive rank for the sample matching but I don't have a good measure of that.

(3) Taking into account homefield advantage is probably a good idea. I think I should redo the numbers doing a +3/-3 road/home adjustment. I'll try to do that later.

Thanks for the analysis. My favorite stat to use for defense and offense power is my adjusted scoring stat. It's scoring D adjusted for SOS. I simply take points allowed (or scored) and subtract points scored (or allowed) by our opponents.

BYU finished #28 in that this year.

2003 was a crazy year due to the stuff Indy refers to. I analyzed that D from a number of angles and called it about the #25 D in the nation.

The two D's were pretty equal, I would say. However, it's amazing how well this D did considering the expectations.

I call this a great coaching job by Bronco on the D.

pelagius · 12-13-2006, 10:43 PM

Quote:

Originally Posted by jay santos

T

The two D's were pretty equal, I would say. However, it's amazing how well this D did considering the expectations.

I call this a great coaching job by Bronco on the D.

Agreed on both counts. To me the big question is how much the switch to the 3-4 contributed to the success this year.

Goatnapper'96 · 12-13-2006, 10:46 PM

that if I joined a golf foursome with you it would be nipple hardening fun, but in all honesty this might be the gayest, and not gay ha ha if I can steal a Slingblade quote, thread I ever read.

Somebody save it with a picture of Paris Hilton almost nekkid and a Santa coughing out "Ho Ho Ho!"

However, if the final conclusion is similarity between the performance and effectveness of the 2003 and 2006 defense, I agree wholeheartedly. Also if the conclusion is that it was a phenomonal coaching job by the defensive staff with the expectations, I say amen and hallelujah!

Now David Nixon might just be the answer and the solution to our droopy knockers of the past!

Here is to Mike Waters, David Nixon and hot chicks from the Great State of Texas!

pelagius · 12-13-2006, 11:06 PM

Quote:

Originally Posted by Goatnapper'96

that if I joined a golf foursome with you it would be nipple hardening fun, but in all honesty this might be the gayest, and not gay ha ha if I can steal a Slingblade quote, thread I ever read.

That's premature; you haven't seen my upcoming post on the "Secret Gospel of Mark."

creekster · 12-13-2006, 11:16 PM

Quote:

Originally Posted by pelagius

That's premature; you haven't seen my upcoming post on the "Secret Gospel of Mark."

I'm lost; will the upcoming post be NH fun or gay-est?

pelagius · 12-13-2006, 11:37 PM

Quote:

Originally Posted by creekster

I'm lost; will the upcoming post be NH fun or gay-est?

LOL.

12-13-2006, 06:52 PM	#1
pelagius Senior Member Join Date: Nov 2006 Posts: 1,431	Defense: Is 2006 better than 2003? There has been a lot of talk about the 2006 defense. Is it the best BYU defense since Bronco has been at BYU? If you just look at the raw numbers in terms of points allowed then the 2006 defense looks very good: Thus on average the 2006 defense has allowed about 10 points per game less than the 2003 defense. The difference in the medians is smaller but still prety large: 5.5 points per game. However, BYU played a much tougher SOS in 2003. Here's a quick comparison of the average Massey year end rank of the opponents: Code: year rank 2003 92.6 2006 47.6 Thus the raw graph above is really not a fair comparison. In order to form a better comparison I made a matched sample: I matched games in 2003 with games in 2006 based on the opponents year end ranking. I got 7 good matches: Code: year team allowed rank year team allowed rank 2003 Georgia Tech 13 34 2006 Arizona 13 37 2003 Stanford 18 77 2006 Wyoming 7 74 2003 Colorado St 58 49 2006 Tulsa 24 53 2003 Wyoming 13 88 2006 New Mexico 17 85 2003 Boise St 50 14 2006 TCU 17 23 2003 Notre Dame 33 46 2006 Utah 31 52 2003 Utah 3 24 2006 Boston College 30 24 I then redid the average and median points allowed using the matched sample: The differences are smaller now. The 2006 defense still performed better on average (although the difference is statistically insignificant) but the medians are very similar. If you forced me to pick I would give an edge to the 2006 defense mainly because of consistency. The 2006 defense never had the disaster game(s) that the 2003 defense did even when playing good teams (maybe that has to do with the switch to the 3-4). Notes: (1) The sample sizes are small here so the analysis doesn't have a lot of statistical power. (2) I am using year end rank; clearly it would make sense to use an offensive rank for the sample matching but I don't have a good measure of that. (3) Taking into account homefield advantage is probably a good idea. I think I should redo the numbers doing a +3/-3 road/home adjustment. I'll try to do that later.

12-13-2006, 10:19 PM	#4
pelagius Senior Member Join Date: Nov 2006 Posts: 1,431	Also, if I do a homefield advantage adjustment +3/-3, keep Indy Coug's scores off turnovers adjustment, and take out the overtime score for Boston College, I get the following for the two years: Code: Year Mean Median 2003 20.14 18 2006 15.43 17

12-13-2006, 10:46 PM	#7
Goatnapper'96 Recruiting Coordinator/Bosom Inspector Join Date: Jan 2006 Posts: 3,412	Gentleman, I have no doubt that if I joined a golf foursome with you it would be nipple hardening fun, but in all honesty this might be the gayest, and not gay ha ha if I can steal a Slingblade quote, thread I ever read. Somebody save it with a picture of Paris Hilton almost nekkid and a Santa coughing out "Ho Ho Ho!" However, if the final conclusion is similarity between the performance and effectveness of the 2003 and 2006 defense, I agree wholeheartedly. Also if the conclusion is that it was a phenomonal coaching job by the defensive staff with the expectations, I say amen and hallelujah! Now David Nixon might just be the answer and the solution to our droopy knockers of the past! Here is to Mike Waters, David Nixon and hot chicks from the Great State of Texas! __________________ She had a psychiatrist who said because I didn't trust the water system, the school system, the government, I was paranoid," he said. "I had a psychiatrist who said her psychiatrist was stupid."

12-13-2006, 08:33 PM	#2
Indy Coug Senior Member Join Date: Jan 2006 Location: Between Iraq and a hard place Posts: 7,569	Please take into account that BYU allowed 8 TDs in 2003 on turnover and kick returns and 0 TDs in 2006. That doesn't even begin to account for the numerous very short fields allowed by BYU in 2003 for both of the above as well. Again, those things happened very rarely in 2006. BYU's 2003 scoring defense is extremely skewed relative to their other major defensive measurables.