Hidden Points- First Down Percentage

Hidden Points- First Down Percentage

Last season Miami converted at least one first down on 64% of their offensive possessions. On the surface, it sounds like a good result, but this percentage actually ranked 63rd out of 66 P5 teams. Only Rutgers, Florida State, and Louisville converted a first down on a lower percentage of their possessions than Miami, as Miami was more than 1.50 standard deviations below the mean in this area. The top teams are converting at least one first down on much higher percentages: Alabama (88%), Oklahoma (86%), Georgia (83%) of the time, so Miami has a long way to go to get to the upper tier in this area.

When I first ran the numbers, I expected to be writing about the impact that this has on the defense, but multiple studies have shown that rest time for the defense, number of plays for defense etc. have zero impact of the ability to score on that defense. The first study from 2011 has shown the same results as the latest study in 2018 on the subject.

What the lack of first downs on offense DOES impact is field position. Surprisingly, Miami had the 10th best starting position among P5 teams last year (34-yard line). What this means is we have a baseline that we can use to start to quantify the number of expected points that moving into higher tiers of 1st down percentage would add. The average team converts a 1st down on 73.3% of their offensive possessions. One standard deviation is 79% and two standard deviations is 85%.

To put this in chart form:
View attachment 87176

Excluding kneel-down possessions to end the half or the game, there were 164 Miami possessions, of which 59 possessions did not result in at least one Miami first down. Adding an average of 12 yards for a first down (“expected” yardage on a first down) for the 15 additional possessions that would need a first down to get to the average of 73% would add 180 yards to Miami’s season total. As you would expect due to Miami’s abysmal punting game last year, Miami opponents actually had better starting field position than Miami had on their drives, despite an elite defense (Miami’s defense was third in the entire country by only giving up 1.12 first downs per opponent possession behind Clemson and Michigan State).

By moving Miami’s percentage of at least one first down to even an average level, Miami would gain roughly three points on the scoreboard. This isn’t a major impact on the scoreboard, but when you factor in that Miami’s defense was so dominant at forcing three-and-outs by the opposing team, this impact would actually be even greater to the expected gain Miami would have on offense from the additional field position gained. A conservative estimate for the impact that would come towards Miami’s offense from being an average offensive team at first down percentage (FDP) would add seven points net gained. As you would expect due to the sliding scale nature of expected points, adding more than 12 yards average per first down (a very conservative estimate) only increases this net gain for Miami.

To get Miami to one standard deviation above the mean, Miami would have gained at least one first down on 25 more possessions than they did last year (to get percentage to 79% for one SD above the mean). These additional first downs would mean Miami would be 12 points net gained over last year.

For an elite offense, Miami would need to gain a first down on 35 additional possessions, leaving them with only 24 possessions on the season where they did not gain at least one first down. These additional possessions and yards would add an additional 20 points net gained over last year.
When you factor in that the data suggests teams see an incremental expectations on possessions after they create the initial first down than possessions that would normally be “expected” to score from that yard-line initially, and you can see how the impact overall could be even greater for Miami.

Here is the game-by-game chart for Miami last year on FDP:
View attachment 87177

Not surprising that Duke and Wisconsin were our worst games from this standpoint, but it does surprise me that Miami was so consistently bad in this area against most teams last year. It really helps to illustrate just how broken the offense was from a consistency standpoint last year.

Overall:
Miami finished near the bottom in the percentage of possessions where they created at least one first down with a percentage of 64%. This was 62nd out of 66 Power-5 teams.

As a net gain, moving to the average percentage of 73%, Miami could expect to add seven additional points. By moving to one standard deviation above the mean (79%), Miami could expect to add 12 points net gain. For two standard deviations, Miami could expect to add 20 points net gain. Since Miami was essentially the same against P5 teams (62.3%) as they were overall without the Savannah State game (65.1%), I decided to include all games, rather than just P5 games.

Getting to average in FDP only increases Miami’s expected win total by a half game. Getting to one standard deviation above the mean gets them close to one full win, and two standard deviations gets them an additional 1.5 wins. Moving Miami from a true talent 7-win team last year, to a true talent 8.5 win team. It’s important to note that this is a 21% increase in expected win percentage, simply by becoming very good at creating at least one first down on each offensive possession.

FDP is another area that Miami can improve their team and gain Hidden Points during the 2019 season.

If you enjoy my work, please follow me on twitter @HurricaneVision. Thank you

TruCane

Sophomore
Joined
Dec 29, 2012
Messages
813
Last season Miami converted at least one first down on 64% of their offensive possessions. On the surface, it sounds like a good result, but this percentage actually ranked 63rd out of 66 P5 teams. Only Rutgers, Florida State, and Louisville converted a first down on a lower percentage of their possessions than Miami, as Miami was more than 1.50 standard deviations below the mean in this area. The top teams are converting at least one first down on much higher percentages: Alabama (88%), Oklahoma (86%), Georgia (83%) of the time, so Miami has a long way to go to get to the upper tier in this area.

When I first ran the numbers, I expected to be writing about the impact that this has on the defense, but multiple studies have shown that rest time for the defense, number of plays for defense etc. have zero impact of the ability to score on that defense. The first study from 2011 has shown the same results as the latest study in 2018 on the subject.

What the lack of first downs on offense DOES impact is field position. Surprisingly, Miami had the 10th best starting position among P5 teams last year (34-yard line). What this means is we have a baseline that we can use to start to quantify the number of expected points that moving into higher tiers of 1st down percentage would add. The average team converts a 1st down on 73.3% of their offensive possessions. One standard deviation is 79% and two standard deviations is 85%.

To put this in chart form:
View attachment 87176

Excluding kneel-down possessions to end the half or the game, there were 164 Miami possessions, of which 59 possessions did not result in at least one Miami first down. Adding an average of 12 yards for a first down (“expected” yardage on a first down) for the 15 additional possessions that would need a first down to get to the average of 73% would add 180 yards to Miami’s season total. As you would expect due to Miami’s abysmal punting game last year, Miami opponents actually had better starting field position than Miami had on their drives, despite an elite defense (Miami’s defense was third in the entire country by only giving up 1.12 first downs per opponent possession behind Clemson and Michigan State).

By moving Miami’s percentage of at least one first down to even an average level, Miami would gain roughly three points on the scoreboard. This isn’t a major impact on the scoreboard, but when you factor in that Miami’s defense was so dominant at forcing three-and-outs by the opposing team, this impact would actually be even greater to the expected gain Miami would have on offense from the additional field position gained. A conservative estimate for the impact that would come towards Miami’s offense from being an average offensive team at first down percentage (FDP) would add seven points net gained. As you would expect due to the sliding scale nature of expected points, adding more than 12 yards average per first down (a very conservative estimate) only increases this net gain for Miami.

To get Miami to one standard deviation above the mean, Miami would have gained at least one first down on 25 more possessions than they did last year (to get percentage to 79% for one SD above the mean). These additional first downs would mean Miami would be 12 points net gained over last year.

For an elite offense, Miami would need to gain a first down on 35 additional possessions, leaving them with only 24 possessions on the season where they did not gain at least one first down. These additional possessions and yards would add an additional 20 points net gained over last year.
When you factor in that the data suggests teams see an incremental expectations on possessions after they create the initial first down than possessions that would normally be “expected” to score from that yard-line initially, and you can see how the impact overall could be even greater for Miami.

Here is the game-by-game chart for Miami last year on FDP:
View attachment 87177

Not surprising that Duke and Wisconsin were our worst games from this standpoint, but it does surprise me that Miami was so consistently bad in this area against most teams last year. It really helps to illustrate just how broken the offense was from a consistency standpoint last year.

Overall:
Miami finished near the bottom in the percentage of possessions where they created at least one first down with a percentage of 64%. This was 62nd out of 66 Power-5 teams.

As a net gain, moving to the average percentage of 73%, Miami could expect to add seven additional points. By moving to one standard deviation above the mean (79%), Miami could expect to add 12 points net gain. For two standard deviations, Miami could expect to add 20 points net gain. Since Miami was essentially the same against P5 teams (62.3%) as they were overall without the Savannah State game (65.1%), I decided to include all games, rather than just P5 games.

Getting to average in FDP only increases Miami’s expected win total by a half game. Getting to one standard deviation above the mean gets them close to one full win, and two standard deviations gets them an additional 1.5 wins. Moving Miami from a true talent 7-win team last year, to a true talent 8.5 win team. It’s important to note that this is a 21% increase in expected win percentage, simply by becoming very good at creating at least one first down on each offensive possession.

FDP is another area that Miami can improve their team and gain Hidden Points during the 2019 season.

If you enjoy my work, please follow me on twitter @HurricaneVision. Thank you
Great stuff. Thank you
 

Cane1992

Recruit
Joined
Jun 26, 2012
Messages
4
And these pitiful stats...

#92 Total Offense
#46 3rd Down conversion offense
#93 Turnovers Lost
#68 in TOP
#118 Net Punting
#100 Kickoff Coverage D

With any sort of offense this team could have been really good in 2018. The D was put in a terrible position in most every possible stat.
 

DTP

Section 102
Joined
Dec 30, 2015
Messages
5,402
The statistical analysis is my third favorite thing here. After new stadium and uniform threads.
Honestly, I find this stuff very interesting. Anybody can tell you the offense struggled, not many people can tell you why backed with numbers. Last year's offense was spectacularly bad. Not only did they not move the ball, they turned it over a lot. When they didn't turn it over, we got shackled with a 12 yard punt on 4th down. Given the talent available, I don't think they could have possibly been worse.
 

g8rh8rMD

Junior
Joined
May 21, 2016
Messages
3,758
It’s remarkable that two units on the same team could have such drastically different average starting field positions. The offense was top ten in the country in average starting position while the defense was among the worst.

This really goes to show how drastic the performance difference was and how much better that defense could have looked if paired with even an average offense.

Great stuff as always, @Lance Roffers!
 

grover

Senior
Joined
Jan 30, 2012
Messages
3,953
What I like is how Lance shows the impact an improvement will have on points per game, based on mathematically sound principles. First back to the mean, then one and two standard deviation above the mean. Isolating a few key stats, over the past three or so posts, he’s found us another 2-3 wins without the need for a Devin Hester miracle.

We of course know this intuitively. Field position and conversions matter. But it’s nice to see a fact-based impact analysis and proof. And it shows that even a moderate level of improvement by Enos (along with special teams) should change our record significantly.
 

fllcane

Recruit
Joined
Oct 5, 2017
Messages
35
Lance,

On your quote:

When I first ran the numbers, I expected to be writing about the impact that this has on the defense, but multiple studies have shown that rest time for the defense, number of plays for defense etc. have zero impact of the ability to score on that defense. The first study from 2011 has shown the same results as the latest study in 2018 on the subject.
Although it might not directly impact defense rest time, it does present the issue of the opposing teams (assuming they do better at first downs than us) getting more offensive plays - aka: more chance to score. Also there is a psychological factor that cannot be discounted, football is a mental toughness game. If the offense keeps f'ing up and the kick team keeps give the opposition great field position, it starts to drag on a defenses mental focus - especially if that defense is not truly, really mentally tough (Duke/Virginia anyone). I love our defense, but I did wonder at times last year about their mental toughness. Now let me say when I say mental toughness its not just the individual player, but the unit as a whole. The leadership in the huddle, the position leadership. Its not as easy as saying Shaq Quarterman is tough (which he is in spades) or xyz player is a tough guy. That is not what I talking about. Its the unit as a cohesive one, playing with mental toughness. Unless the head coach is instilling leadership, pride, and toughness then as the game/season goes on and the offense keeps sputtering, the defensive units toughness start to erode. Now with that said, I think the defensive unit was much tougher than the offensive - which I attribute to Manny. But a coordinator only goes so far. It really comes from the top down. As it does in any organization. Apple under Jobs == Butch Davis 2000/JJ, under Tim Cook == Larry Coker.
 

JD08

Freshman
Joined
Dec 19, 2014
Messages
2,130
Understand my brand of writing is not for everyone. If you have feedback on what makes it boring I’m happy to listen.
Data doesn't lie.

I, for one, absolutely love how analytics is being used now and the fact that you provide it here is appreciated.
 

imyrhuckleberry

Sophomore
Joined
Oct 30, 2018
Messages
739
Rutgers, Florida State, Louisville and Miami leading the bottom of converted first-down percentage . . . just . . . Wow . . .
I would expect we could beat that just running the wildcat . . .

Thanks, Lance . . . great, as usual . . . and sobering, as usual . . .
 

Empirical Cane

We are what we repeatedly do.
Joined
Sep 3, 2018
Messages
2,965
For years, Miami's two biggest strategic production problems have been 1) inability to sustain offensive drives via 1st downs (several reasons why) and 2) inability to make timely 3rd down defensive stops (also several reasons why).

Solve those two strategic issues, Miami wins a Natty--probably several.
 

2020 Commits

OT
6'6"
330
Orange Park, FL
WR
6'2"
190
Bradenton, FL
QB
6'4"
215
Glastonbury, CT
WR
5'10"
160
Miami, FL
DT
6'3"
290
Hollywood, FL
RB
5'11"
205
Miami, FL
DE
6'4"
240
Orange Park, FL
DE
6'3"
265
Miami, FL
S
6'0"
185
Homestead, FL
C
6'3"
270
Ocala, FL

Latest Predictions

by Matthew_Suero
High
by Swampcity
Medium
by Swampcity
Low
by Swampcity
Medium
by Swampcity
Certain

2019 Schedule

08/24
Florida
Orlando, FL
L 24 - 20
09/07
North Carolina
Chapel Hill, NC
09/14
Bethune-Cookman
Miami Gardens, FL
09/21
Central Michigan
Miami Gardens, FL
10/05
Virginia Tech
Miami Gardens, FL
10/11
Virginia
Miami Gardens, FL
10/19
Georgia Tech
Miami Gardens, FL
10/26
Pittsburgh
Pittsburgh, PA
11/02
Florida State
Tallahassee, FL
11/09
Louisville
Miami Gardens, FL
11/23
FIU
Miami, FL
11/30
Duke
Durham, NC
Top