Re-examining our situation

[mad dog]

Tina would be proud...

There has been a lot of question, at least in my mind, about just how bad the last five games have been. I needed a concrete way to ascertain if things really were as bad as they have ever been. I had a suggestion PMed to me last night, and I decided to run with it.

First, I cast aside all subjective factors, including media coverage, expectations, and sensitivity to the athletic fee vote. Next, I examined the most basic component of this kind of competition, that being 1) points scored, 2) points given up, and 3) the differential between the two. Then I selected a few years for study. I chose the consensus worst year for North Texas football, which was the 1972 campaign which registered exactly one win, and then the last four years, which offer all of Dodgeball to this point and two of the worst years of Dickeyball. My findings and raw data are attached if you want to take a look.

The data suggest two conclusions. First, this season, through five games, is objectively the worst in the past four years, and decidedly worse than even the 1972 season. Second, there seems to be a phenomenon at place where the offense and defense has doubled in what looks remarkably like point inflation.

Through five games, our point differential average is -39.2. This means that we have been outscored, on aggregate, by 39 points so far this season. If this average holds through the rest of the year, it would be 19 points higher than the next closest season average which, coincidentally, was last year. By comparison, the 1972 point differential was -18.7. It is also worth noting that, in 2006 (the year Darrell Dickey was fired), our total season differential was -12.5. For grins, I went back and took a peek at the 2003 team. That squad put up 28 points per game, and allowed 21, for an average differential of +7.

Now, if we decide to only look at the first five games , the same distribution holds, but the difference decreases. Through the first five games, the differential was -32.2 in 2007 and -28.2 in 1972. This likely reflects a higher concentration of talented teams in out of conference play earlier in the year.

When that differential is broken down, a very telling phenomenon becomes apparent. Between 2006 and 2007 (Dickey's last year and Dodge's first), offense doubled from 12.8 points/game to 24.8. Unfortunately, the defense also nearly doubled, from 25.3 to 45.1 points/game. So, while the points scale increased, there was little net change in terms of overall effect, and that was reflected in the win column (3 wins in 2006, 2 wins in 2007).

The problem we're facing now is that the defense is as bad as it was last year through five games (52.2 ppg this year, 51.6 last year), but the bottom has fallen out of the offense. At this point last year, the team was scoring 19.4 points per game. This year, we're only scoring 13.

Historically speaking, the team's offensive and defensive numbers are better at the end of the year than through the first five games. However, the improvement hovers around 5 ppg offensively and 3 ppg defensively. If we apply both that to our current figures, our year end numbers are 18 ppg on offense and 49 ppg on defense for a net projected differential of -31. A final year end score of -31 would be 11 points higher than the -20.3 we ended with last year.

So what does all this mean? From the data compiled, it looks like this season is exactly what we think it is. It is disastrous on a historical scale, and getting worse. The good news is that the offense is probably good enough to win us a few games if the defense had a pulse. But, until the defense can find some way to hold their opponent to five touchdowns or less, it really won't matter how good the offense is.

[jimmyjames_99]

Good analysis! No question we have problems.

Just thinking outloud, if you take away the points allowed the defense had no control over (i.e. punt/kick-off returns, interception run backs or even drives that started within our 20 yd line) would the defense look better statistically than those other teams?

[letsgiveacheer]

As someone who attended everyone of those home games, I can recall what a dismal season we had. The irony was that after Rod Rust was fired, he later (very briefly) became a NFL head coach! Probably the only one in our history. As these things go, it is hard to make any sort of anecdotal comparisons except we have been there before.

My favorite cheer from 1972:

Give em hell, Rod Rust, give em hell...

Give em hell, Rod Rust, give em hell...

Give em hell, Rod Rust, give em hell...

Ah hell, give em Rod Rust...

I wonder how the Wheat Shockers are doing this season?

[island eagle]

As someone who attended everyone of those home games, I can recall what a dismal season we had. The irony was that after Rod Rust was fired, he later (very briefly) became a NFL head coach! Probably the only one in our history. As these things go, it is hard to make any sort of anecdotal comparisons except we have been there before.

My favorite cheer from 1972:

Give em hell, Rod Rust, give em hell...

Give em hell, Rod Rust, give em hell...

Give em hell, Rod Rust, give em hell...

Ah hell, give em Rod Rust...

I wonder how the Wheat Shockers are doing this season?

Rod Rust was a decent coach, but a lousy recruiter. When he ran out the talent he inherited when he took over as coach, his record went south.

[Cerebus]

Your statistical method is flawed.

You are forgetting strength of schedule. A 10 point loss to LSU should be weighted differently than a 10 point loss to Sul Ross.

You also can't compare 5 game versus a season since our strength of schedule is usually so front loaded.

I am not saying I disagree with your hypothesis, I am saying you haven't proven it.

[mad dog]

Strength of schedule is subjective, and rather than getting bogged down in who was better and how you quantify that, it was left out.

As far as front loaded OOC schedules, I have broken down OOC and Conference losses in my data, which support the same conclusion.

[Cerebus]

Strength of schedule is subjective, and rather than getting bogged down in who was better and how you quantify that, it was left out.

Really? Because ranking services like Saragin have been using mathematical models to try and gauge that for a long time.

Sure, they aren't a perfect, but SOS is too big to overlook. You have to try and quantify it somehow.

As far as front loaded OOC schedules, I have broken down OOC and Conference losses in my data, which support the same conclusion.

I was talking specifically about this comparison:

Through five games, our point differential average is -39.2. This means that we have been outscored, on aggregate, by 39 points so far this season. If this average holds through the rest of the year, it would be 19 points higher than the next closest season average which, coincidentally, was last year. By comparison, the 1972 point differential was -18.7. It is also worth noting that, in 2006 (the year Darrell Dickey was fired), our total season differential was -12.5. For grins, I went back and took a peek at the 2003 team. That squad put up 28 points per game, and allowed 21, for an average differential of +7.

Like I already said, I don't disagree with you hypothesis, but your numbers don't prove it.

This is in the vein of cum hoc ergo propter hoc.

[GoMeanGreen1999]

Tina would be proud...

There has been a lot of question, at least in my mind, about just how bad the last five games have been. I needed a concrete way to ascertain if things really were as bad as they have ever been. I had a suggestion PMed to me last night, and I decided to run with it.

First, I cast aside all subjective factors, including media coverage, expectations, and sensitivity to the athletic fee vote. Next, I examined the most basic component of this kind of competition, that being 1) points scored, 2) points given up, and 3) the differential between the two. Then I selected a few years for study. I chose the consensus worst year for North Texas football, which was the 1972 campaign which registered exactly one win, and then the last four years, which offer all of Dodgeball to this point and two of the worst years of Dickeyball. My findings and raw data are attached if you want to take a look.

The data suggest two conclusions. First, this season, through five games, is objectively the worst in the past four years, and decidedly worse than even the 1972 season. Second, there seems to be a phenomenon at place where the offense and defense has doubled in what looks remarkably like point inflation.

Through five games, our point differential average is -39.2. This means that we have been outscored, on aggregate, by 39 points so far this season. If this average holds through the rest of the year, it would be 19 points higher than the next closest season average which, coincidentally, was last year. By comparison, the 1972 point differential was -18.7. It is also worth noting that, in 2006 (the year Darrell Dickey was fired), our total season differential was -12.5. For grins, I went back and took a peek at the 2003 team. That squad put up 28 points per game, and allowed 21, for an average differential of +7.

Now, if we decide to only look at the first five games , the same distribution holds, but the difference decreases. Through the first five games, the differential was -32.2 in 2007 and -28.2 in 1972. This likely reflects a higher concentration of talented teams in out of conference play earlier in the year.

When that differential is broken down, a very telling phenomenon becomes apparent. Between 2006 and 2007 (Dickey's last year and Dodge's first), offense doubled from 12.8 points/game to 24.8. Unfortunately, the defense also nearly doubled, from 25.3 to 45.1 points/game. So, while the points scale increased, there was little net change in terms of overall effect, and that was reflected in the win column (3 wins in 2006, 2 wins in 2007).

The problem we're facing now is that the defense is as bad as it was last year through five games (52.2 ppg this year, 51.6 last year), but the bottom has fallen out of the offense. At this point last year, the team was scoring 19.4 points per game. This year, we're only scoring 13.

Historically speaking, the team's offensive and defensive numbers are better at the end of the year than through the first five games. However, the improvement hovers around 5 ppg offensively and 3 ppg defensively. If we apply both that to our current figures, our year end numbers are 18 ppg on offense and 49 ppg on defense for a net projected differential of -31. A final year end score of -31 would be 11 points higher than the -20.3 we ended with last year.

So what does all this mean? From the data compiled, it looks like this season is exactly what we think it is. It is disastrous on a historical scale, and getting worse. The good news is that the offense is probably good enough to win us a few games if the defense had a pulse. But, until the defense can find some way to hold their opponent to five touchdowns or less, it really won't matter how good the offense is.

Nice work and yes, as Tasty said, Tina would be proud. I wonder though, if the defense this year isn't better than it was last year but only appears to be worse since our offense seems to turn the ball over so many time, and in bad places. The more you turn it over in the redzone and on the opponents end of the field the harder it is on the defense.

[eulesseagle]

FLAWED-

TWO DIFFERENT STYLES OF OFFENSE.

[JesseMartin]

If I had a stats package at home I'd ask for your numbers--I'd like to see a multivariate analysis that also includes YPG (defense, offense, and ST). I think that level of regression might shed some light on very specific problems.

I forgot to mention--somebody was talking about Sul Ross on here EARLY this year--I passed it while driving to AZ! Middle of nowhere, but it did look like a very pretty campus.

[BeanCounterGrad'03]

I forgot to mention--somebody was talking about Sul Ross on here EARLY this year--I passed it while driving to AZ! Middle of nowhere, but it did look like a very pretty campus.

Ah yes...Alpine, TX...quite familiar with that town (they were in our district in high school)

[ADLER]

With all due respect, as I am grateful to every player that suits up for North Texas, but as far as players that could garner interest from the NFL, Dickey inherited a gold mine compared to what he left. Other than Jamario, those type of players stopped arriving at North Texas after the 2001 signing class. Was that due to coaching staff changes, changes in recruiting philosophy, complacency, time constraints brought on by the New Orleans Bowls, or possibly something else?

[LongJim]

If I had a stats package at home I'd ask for your numbers--I'd like to see a multivariate analysis that also includes YPG (defense, offense, and ST). I think that level of regression might shed some light on very specific problems.

There is some correlation in the following stats with attrition (for whatever reason, God help us) on the OL and DL, which is where UNT (IMO) can trace most of its recent problems back to. Check out the stats for 2005, 2006, 2007, and 2008 for the following statistics:

Stats

Sacks

TFL's

Sacks allowed

3rd down conv %

3rd down conv %, defense

Rushing D

Passing D

Scoring D

Interceptions

Each of these speaks to what is ailing UNT right now.

[JesseMartin]

There is some correlation in the following stats with attrition (for whatever reason, God help us) on the OL and DL, which is where UNT (IMO) can trace most of its recent problems back to. Check out the stats for 2005, 2006, 2007, and 2008 for the following statistics:

Stats

Sacks

TFL's

Sacks allowed

3rd down conv %

3rd down conv %, defense

Rushing D

Passing D

Scoring D

Interceptions

Each of these speaks to what is ailing UNT right now.

I still have nothing to run it through. Doing it by hand would take FOREVER.