Upon Further Review: Manny Diaz as DC

[#1cane]

I judge Manny by the eye test. How does Miami look against good offenses? Not very good. On the other side Miami usually looks good defensively against the weaker teams (Duke last year) padding statistics making the defense look statistically better. The attacking style works against less talented teams but gets burnt against high powered offenses.

 

[Genghis Cane]

I judge Manny by the eye test. How does Miami look against good offenses? Not very good. On the other side Miami usually looks good defensively against the weaker teams (Duke last year) padding statistics making the defense look statistically better. The attacking style works against less talented teams but gets burnt against high powered offenses.

It looked pretty good against ND in 2017.

 

[dantheman349]

I did basically this same exact thread a couple weeks ago. Great job, Lance, you dug deeper and presented it better. But the bottom line remains, nearly 90% of the teams we played while Manny was the DC had a BELOW AVERAGE game on offense. Sometimes well below average, but almost every single week, the offense we played put up numbers below their average. It just makes no sense to me how people could have really watched 2016 - 2018 and say Manny is a ****** DC. If you think he's not elite, that's fine. But he was ABSOLUTELY a high end DC here in those 3 seasons. Thankfully data exists and it can poo-poo on nonsense opinions.

Now imagine those defenses with a offense that can actually move the ball.

 

[407cane]

The "analyst" title doesn't sound too great but don't be mistaken. Shoop will be doing a TON on the defense.

That mixed with trob who was just a DC at an SEC program and Simpson who was an NFL DL coach I mean manny has plenty of valid options to lean on for delegation

 

[dantheman349]

bottom line our defense will be a lot better. Also now we have legit position coaches at LB/CB and that will change things around for us with the better offenses.

 

[The Franchise]

Awesome work as always, Lance. As with all statistical analysis, there’s a ton of leeway in the definition of what is a similarly talented opponent.

I’d surmise Miami has a pretty big talent edge in most of the sample you pulled from even though you removed FCS teams. I’d love to see what Manure did at other stops where he might not have had that advantage or what he did here against competent offenses.

Another key factor here is that Manure hasn’t recruited as well on D as Folden did. All of Manure’s best players were Folden recruits. When they left, our results started to decline noticeably.

We’re also going to see if he can do two major jobs at the same time when he’s struggled so badly at his main job. I’m hopeful but skeptical.

 

[Lance Roffers]

Any chance you could run the numbers on Shoop?

I did. They weren’t as good as you’d think. Though his last year with Simmons was elite.

Felt like adding him would distract from my overall point and possibly derail.

 

[The Franchise]

Ed Orgeron.
Mario Cristobal.
Mike Leach.
Dave Aranda.

off top.

Don’t think any of those guys were fired from coordinator positions.

 

[The Franchise]

I don't think you can account for this because we have no way of knowing what games he called.

He only called the successful games.

 

[Lance Roffers]

Not sure why you'd answer to people calling Manny an "average or even poor" DC. As a DC, he was above average and it's hard to argue anything beneath that level.

Currently, despite the fact he's not our DC, but rather our HC/DC, the real debate seems to be around whether he is above average or elite, and how that relates to the execution of his defensive philosophy. And, for that, I think you'll need more than YPP.

I don't question your conclusion based on the narrow scope you selected. But, the methodology tells me there may be a separate discussion/analysis to have. See, part of the problem we've watched unfold is Diaz's philosophy is designed to be good in that YPP metric (btw, I think it's a good indicator, too, but often needs more context).

He coordinates for negative plays to get an offense off schedule. Those negative plays come at a risk, which as a DC he often did a good job of mitigating. The result is a nice YPP metric. However, the philosophy is vulnerable, as we saw in multiple instances or when he (or Baker) is missing a piece. That said, his most impressive year was when he combined a strong YPP with elite 3rd down %. I thought he was on his way to really evolving his defense at that point, and then we saw a disjointed philosophy the last two years. Sure, it was "Baker," but it was Manny's defense and he said himself he was in on everything.

I think the more important question is if his philosophy, when executed as it has been in the past, is solid enough for us to win bigger games consistently. I would be curious to see an analysis of how he performed in "big games" (however way you want to define that) vs how other coordinators/philosophies performed in "big games." Say, Top 25 teams. If there are common Top 25 opponents, all the better.

All of the above is moot if he tinkers with his philosophy and allows Jess Simpson, Shoop, and T-Rob to offer different perspectives to come up with a more controlled version of the negative play madness, and one that has the appropriate coverage behind it. For instance, we played a ton more zone behind the madness these past two years under Baker, resulting in obvious vulnerabilities.

I’ll look into answering your questions.

 

[Shogungts]

This is a great breakdown. I think the concern most of us have isn't Manny as DC, but Manny as head coach AND DC. I'd love to see a breakdown of head coaches who also call plays (preferably on defense) and their relative success rates.

This is my concern as well - particularly with it being the D. If you are the HC and OC you get to control the speed of the game when you are calling plays, but as the DC Manny will have to call the D and handle HC matters at the speed dictated to him by the opposing OC.

There are some well known coaches to have called the plays as a HC as pointed out by Manny Navarro in a recent article:

"TCU’s Gary Patterson has been calling defensive plays since he took over as head coach in 2000. He’s been named the Coach of the Year twice (2009 and 2014) and led TCU to six AP Top 10 finishes. Others such as Virginia’s Bronco Mendenhall and former Tennessee coach Jeremy Pruitt and Vanderbilt coach Derek Mason have handled both responsibilities, too, and relented or handed it back to others before."

That is obviously not a breakdown like Lance did, but it does show a bag of mixed results (although not a huge sample size).

 

[mirvin88]

@Lance Roffers can you add something related to third down success?

 

[Rellyrell]

Yeah, a big dawg in the middle as NT or DT changes a defense one step up (i.e., bad to average, above average to elite, etc.) that's what Gerald Willis did that year. Weird that he's been a no show in the NFL.

One of the great mysteries in recent memories. Just knew he would be a stud at the next level. He was playing out of his mind in ‘18

 

[Lance Roffers]

@Lance Roffers can you add something related to third down success?

Has been shown to have little stickiness.

 

[Lance Roffers]

I judge Manny by the eye test. How does Miami look against good offenses? Not very good. On the other side Miami usually looks good defensively against the weaker teams (Duke last year) padding statistics making the defense look statistically better. The attacking style works against less talented teams but gets burnt against high powered offenses.

This did not bear out in the data and is false.

 

[Empirical Cane]

With the news coming out that Manny will serve as the playcaller for the defense, returning to the de facto DC position, I wanted to dive deeper to establish whether Coach Diaz was a top-shelf DC or not.

Methodology
It started with a Twitter debate, where fans online perceived Coach Diaz as either an average DC, or even a poor one. From there, I asked for fans to identify the top-5 DC’s currently in college and received a list of various names. I settled on these names:

• Brett Venables- Clemson
• Marcus Freeman- Cincinnati (now Notre Dame)
• Dave Aranda- LSU (now HC at Baylor)
• Jim Leonhard- Wisconsin
• Jon Heacock- Iowa State
• Blake Baker- Miami

There are other worthy candidates, but this list seemed like a representative start of identifying how standout DC’s perform in their roles.

I am using a sample size of three years for each of them, except for Baker, as I am using him for a control to show how Miami fared under Baker as opposed to Manny Diaz.

The metric I am using is Yards Per Play. I chose Yards Per Play (YPP) because it is simple to calculate, readily available, and is a good catch-all for how a defense fares against an opposing offense overall.

To gauge how a team fared against peers, I am removing teams that are not of like-quality. I.E. a Power-5 team, I am removing all non-Power-5 teams from their results. For Cincinnati, I am removing lower tier FBS teams and all FCS teams. This helps to stabilize the talent of the teams and removes a defense beating up on the Missouri State’s of the world.

From here, I calculated the “Win Percentage” that a DC against opposing offenses. A win is holding the offense under their standard average against other peer-like teams, a loss is allowing them to outperform their YPP against that DC’s defense.

Additionally, I wanted to calculate the “Difference” between what each DC allowed on YPP and what those opponents averaged against other peer opponents.

I also calculated the standard deviations of the opponents’ offensive performance and tabulated the number of times each DC held their opponents one standard deviation under their average and two standard deviations. I calculated how often the DC allowed their opponent to outperform their average by one and two standard deviations as well.

Results
By looking at performances against peer institutions and then then weighing it against their offensive performance against other peer opponents I believe you are getting at the actual contributions of each DC and can accurately gauge the quality of coach. The results pass the “smell test” as well, with Venables being far and away the best and Blake Baker being far and away the worst.

Here are the results for % of games holding opponents under their norms (Win Percentage):

View attachment 143007

Coach Diaz fares very well here, finishing second among the group of DC’s with a win percentage of 77.4%. Venables leads the pack, as he does in every category, by holding an opponent under their norms an astonishing 86.5% of the time. Keep in mind, this is against peer institutions, so no Bethune Cookman’s propping up that number for Venables (or anyone).

Surprisingly, Blake Baker did not come in last here, as Jim Leonhard and Jon Heacock both finished lower than his 68.4% showing.

Here are the results for Yards Per Play (YPP):

View attachment 143008

Again, Venables leads the way, with Coach Diaz coming in second. Marcus Freeman is the third DC to hold opponents to under 5 YPP, but keep in mind he is doing that at Cincinnati, with a decided talent advantage against even some conference foes. Clemson may have a talent advantage over their opponents, but the revenue from the ACC to each team at least allows them to be competitive in resources. Some teams in the AAC have a wide budget gap between themselves and Cincinnati.

Blake Baker comes in last in this metric, not surprisingly. That is a gap of .58 yards per play allowed between Baker and Diaz, which is cavernous.

Finally, here are the results for how each DC fared at holding their opponents below their norms (Difference):

View attachment 143009

Venables comes in at a ludicrous -1.17 YPP against peer institutions and their offensive averages. Marcus Freeman is second, but Manny comes in third at -0.84 YPP.

Blake Baker does not finish in last place in this metric, surprisingly, Jim Leonhard does. Leonhard appears to be a bit overrated by fans currently.

Overall:
For those of you who just skip to the end, suffice to say, Coach Diaz comes in second amongst this group of DC’s over a three-year sample. Over this time, keep in mind that Coach Diaz only had Gerald Willis for one season and had three true freshmen LB’s to scheme around. Then he introduced the Striker into his defense in this sampling as well. Venables, Aranda, Leonhard, Heacock are considered among the best names in the game, so Coach Diaz is definitely an excellent DC, who will be a massive upgrade over Blake Baker.

In looking at the performances of Coach Diaz, a few things stand out:

• Coach Diaz will consistently hold an offense under their norms, but will generally not dominate an offense (over one standard deviation under their norms), as he did this in only 29% of games. Venables did this an absurd 57% of the time. Coach Diaz has only held an offense two standard deviations under their norm once (Against Virginia in 2018), while Venables has done that six times over the last three years.
• Coach Diaz’s defenses are remarkably consistent in their ability to scheme negative plays against an offense and hold their overall performance down.

I went into this exercise expecting to find Coach Diaz was outside of that elite tier of DC’s and would be solidly in the “good” tier, but the data confirms that Coach Diaz profiles in the 90th percentile of DC’s in the country and will put up a performance far better than we have seen over the past two seasons.

Lance... excellent start.

One thing that stands out however: Offense (of same team) is now a critical component to defensive performance.

In your "under norms" data analysis, how could we pair it with that same team's offensive performance? My guess is either ToP or PPG.

If I'm right, it should show Venables' holding opponents under their norm was at least partially (maybe significantly) enabled by Clemson's O putting pressure on the opposing team's playcalling and making it a touch easier on Venables.

WDYT?

 

[mirvin88]

Not sure why you'd answer to people calling Manny an "average or even poor" DC. As a DC, he was above average and it's hard to argue anything beneath that level.

Currently, despite the fact he's not our DC, but rather our HC/DC, the real debate seems to be around whether he is above average or elite, and how that relates to the execution of his defensive philosophy. And, for that, I think you'll need more than YPP.

I don't question your conclusion based on the narrow scope you selected. But, the methodology tells me there may be a separate discussion/analysis to have. See, part of the problem we've watched unfold is Diaz's philosophy is designed to be good in that YPP metric (btw, I think it's a good indicator, too, but often needs more context).

He coordinates for negative plays to get an offense off schedule. Those negative plays come at a risk, which as a DC he often did a good job of mitigating. The result is a nice YPP metric. However, the philosophy is vulnerable, as we saw in multiple instances or when he (or Baker) is missing a piece. That said, his most impressive year was when he combined a strong YPP with elite 3rd down %. I thought he was on his way to really evolving his defense at that point, and then we saw a disjointed philosophy the last two years. Sure, it was "Baker," but it was Manny's defense and he said himself he was in on everything.

I think the more important question is if his philosophy, when executed as it has been in the past, is solid enough for us to win bigger games consistently. I would be curious to see an analysis of how he performed in "big games" (however way you want to define that) vs how other coordinators/philosophies performed in "big games." Say, Top 25 teams. If there are common Top 25 opponents, all the better.

All of the above is moot if he tinkers with his philosophy and allows Jess Simpson, Shoop, and T-Rob to offer different perspectives to come up with a more controlled version of the negative play madness, and one that has the appropriate coverage behind it. For instance, we played a ton more zone behind the madness these past two years under Baker, resulting in obvious vulnerabilities.

Your last paragraph is most important. I don't necessarily HATE his philosophy of negative plays because there is a lot of data that says if x team has a negative play they're much less likely to score points that drive.

But it needs work and most glaringly as you mentioned, the coverage concepts.

 

[Lance Roffers]

@Lance Roffers, would comparing the success rate of opposing teams help to eliminate the outliers that skew the YPP?

Success rate is honestly more prone to outliers than YPP.

Example: A 3-yard run on 3rd & 2 and a 75-yard TD run on 3rd & 2.

Both are 100% success rate and obviously vastly different impacts to the game. Manny’s scheme looks to impact explosive plays from the offense and create their own.

There is no perfect metric, and Success Rate has its place, but not here in my view. The best would be to use Expected Points Added, but would be far more work than I want to do because it’s difficult to code to exclude certain games, using different years for different coaches, and take out the performance against that defense (Miami’s opponents for instance) in their sample of offensive performance.

It can be done, but for the marginal improvement and far more complex outcomes that would be more difficult for readers to understand, I didn’t see the value.

 

[Lance Roffers]

Lance... excellent start.

One thing that stands out however: Offense (of same team) is now a critical component to defensive performance.

In your "under norms" data analysis, how could we pair it with that same team's offensive performance? My guess is either ToP or PPG.

If I'm right, it should show Venables' holding opponents under their norm was at least partially (maybe significantly) enabled by Clemson's O putting pressure on the opposing team's playcalling and making it a touch easier on Venables.

WDYT?

I did it tie to the opponents offensive performance against other peers.

TOP has shown to be completely useless. R^2 under .10

if Clemson’s offense is performing well, that would impact Venables negatively because we know that teams pass more when they’re trailing and most every team has higher YPP when passing.

YPP is best for this review in my mind because it scales it on a per play basis and accounts for tempo and situations better.

It doesn’t account for fact that Clemson has more talent than anyone they play most every game. You do see a degradation of performance from Venables against SEC. Typically big games or playoff games.

 

[88]

I’m not sure any defenses stop great offenses anymore. Clemson, Bama, Georgia, OSU couldnt. It’s going to come down to crushing the average offenses and keeping them out of the game and being able to go blow for blow with top offenses.