Hidden Points- Punting

Hidden Points- Punting

Lance Roffers
Lance Roffers
Welcome to the first installment of a new series that tries to identify areas Miami can improve over last year where I quantify "hidden points."

While the focus of the 2019 offseason has been on coaching changes, what will the offense look like under Coach Enos, and who will be the QB, I wanted to focus on some areas where Miami can impact games outside of offensive philosophy.

Those who read my work know that I tend to dip into the numbers to better understand what is happening or how to evaluate different outcomes within a given game/season/recruit. I decided to try something a little different in this study and try to quantify several areas that Miami struggled with last season outside of overall offensive rank and place a number on where Miami can make drastic improvements with merely average play during the 2019 football season.

Methodology
It is no secret that many of my readers just want to see the results so, if that is you, feel free to jump down to the results. For the rest of you, here is how I went about my study:

I looked into the concept of “expected points” that was made famous in the NFL by Brian Burke, who joined ESPN on the basis of his excellent work with statistics. Taking each drive starting position for CFB you get a cluster of outcomes, to get a smooth line you can run a LOESS regression and have a line like this:

85908


I plotted each yard line with a corresponding point value and it turns out to be an expectation of 5.50 points at the 1-yard line (no extra point included) and, conversely, you are expected to lose 0.50 points from your own 1-yard line.

From here, I looked at several different areas of the teams’ performance and plotted their expected points and worked to analyze how different levels of outcome could contribute to Miami scoring or allowing fewer points over the upcoming season.

Punting
I’ve long wanted to try and quantify the impact of Miami’s punting last year and what sort of impact average punting could have on the team. Out of 66 teams in Power-5 conferences (I’m including Notre Dame), Miami finished #62 in terms of net punting. That drops to #64 if you exclude the Savannah State game.

Some housekeeping on how I conducted my study: I split results into “Average,” “Good”, “Great” net punts. Average is the average of all teams, good is one standard deviation above the mean, great is two standard deviations above the mean. For directional punting, I applied an average punt as the 15-yard line, a good punt as to the 10-yard line, and a great punt as the 5-yard line. This means that my results are quite conservative compared to what a great directional punter would add to Miami’s expected points.

Since the entire play is driven by the punter, I hold returns against the punter (net punting). The punter can legally kick the ball out-of-bounds, kick the ball high into the air and force a fair catch, or directionally kick as well, so even if there was poor coverage, the return yards are held against the punter.

In 2018, Miami punted the ball 59 times (excluding Savannah State game), with only 16 of the punts even reaching “average” status (27%) and only 10 punts reaching “great” status on the season (17%). This means that 73% of the punts Miami had on the season actually cost the team expected points.

The totals come out this way:
85912


With just an average punting game (38 net yards a punt), Miami would save eight points on the upcoming season. With a good punting game (40 net yards a punt), Miami would save fourteen points. With a great punting game (42.4 net yards a punt), Miami would save twenty-four points! Because you cannot score fractional points, the figures are rounded up and include extra points.

Keep in mind this was with a conservative designation on directional punting. Essentially, if Louis Hedley can come in and provide us with a good punting game, he will represent an addition larger than any other position has to gain outside of the QB position.

Results
If you looked simply at the Pythagorean expectation of Miami's win-loss record last year, you'd see a winning percentage of .716. This number is heavily skewed by the fact that Miami beat up on Savannah State, FIU, and Toledo. If you take only the results against P5 opponents last year, you see that Miami was expected to have a .500 winning percentage (actual record was 4-6). Giving Miami just an "average" punting game last year would have added a half win to their total expectation. Giving Miami a "good" punting game last year would have added an entire win to their total expectation. Giving Miami a "great" punting game last year would have added an entire win and a half to their total expectation without changing any other item within the program.

Adding 1.5 wins to a 10-game win expectation (P5 opponents) is a massive change and helps to illustrate how much Miami can improve simply by finding hidden points on the football field. Here's hoping Louis Hedley can provide us with a boost this year in the punting game.

Next week I will bring part two of "Hidden Points" and will explore another area where Miami can make huge gains with simply average performance. If you like my work, please give me a follow on Twitter @HurricaneVision.
 

Comments (43)

Out of 66 teams in Power-5 conferences (I’m including Notre Dame), Miami finished #62 in terms of net punting. That drops to #64 if you exclude the Savannah State game.

Who could possibly have been worse?
 
You apply a whole new meaning to the expression “all other things being equal”. Super appreciated. Love your work.
 
Advertisement
This is an important and under discussed topic. Our punting was atrocious last year, and our field position on O and D was impacted as a result. We were effectively playing with a hand tied behind our back.

Would be interested to see what our field position by opponent was on a relative basis. What was our average starting field position on O and our opponents? Exclude opening kick-offs of each half.
 
"Adding 1.5 wins to a 10-game win expectation" - I thought it was a 5 win expectation? You said 4-6 but .500 based on pyth math
 
"Adding 1.5 wins to a 10-game win expectation" - I thought it was a 5 win expectation? You said 4-6 but .500 based on pyth math

Probably poorly worded. My intention was to express that you're adding the 1.5 wins over only a 10-game sample (because I am talking only about the P5 games there).

The "win expectation" was stating a "win expectation over 10 games" essentially.
 
Lance, great work per usual mate. We all knew that our atrocious punt game last season cost us games but surely didn’t know if we were in the great category that it would’ve translated to 1.5 more wins per 10 game sample. Here’s praying Hedley gets us at minimum into the good range, if not the very good. Adding that to the improvements on the offensive side of the ball should translate to a comfortable margin of victory as it relates to the Coastal title. It will be in the ACCCG that we’ll need all the extra hidden yardage we can come by to take on the likes of Clemson.
 
Advertisement
Imo punting on this team over the last two seasons cost us more than 1.5 wins per season.

Starting with the LSU game it's felt like we've started 90% of our drives on our own 5 while other teams started theirs at the 50,
 
Last edited:
Bill Connelly (who I feel is the best in the business with the analytics driven off-season preview pieces) hasn’t published his Miami preview article yet, but he has his raw data public already. He has Miami at 130th in the country in Punt Efficiency last year.

There are 130 FBS teams in the country.

Yep, dead ******* last. Not that those of us who watched the games didn’t know it was abysmal. But wow. Dead last. By the way, he defines this metric as such:

Punt Efficiency: Based on where the punt takes place, we create an expected net punting value. Punt efficiency is simply the percentage of your punts that resulted in a net higher than the expected value.
 
Welcome to the first installment of a new series that tries to identify areas Miami can improve over last year where I quantify "hidden points."

While the focus of the 2019 offseason has been on coaching changes, what will the offense look like under Coach Enos, and who will be the QB, I wanted to focus on some areas where Miami can impact games outside of offensive philosophy.

Those who read my work know that I tend to dip into the numbers to better understand what is happening or how to evaluate different outcomes within a given game/season/recruit. I decided to try something a little different in this study and try to quantify several areas that Miami struggled with last season outside of overall offensive rank and place a number on where Miami can make drastic improvements with merely average play during the 2019 football season.

Methodology
It is no secret that many of my readers just want to see the results so, if that is you, feel free to jump down to the results. For the rest of you, here is how I went about my study:

I looked into the concept of “expected points” that was made famous in the NFL by Brian Burke, who joined ESPN on the basis of his excellent work with statistics. Taking each drive starting position for CFB you get a cluster of outcomes, to get a smooth line you can run a LOESS regression and have a line like this:

View attachment 85908

I plotted each yard line with a corresponding point value and it turns out to be an expectation of 5.50 points at the 1-yard line (no extra point included) and, conversely, you are expected to lose 0.50 points from your own 1-yard line.

From here, I looked at several different areas of the teams’ performance and plotted their expected points and worked to analyze how different levels of outcome could contribute to Miami scoring or allowing fewer points over the upcoming season.

Punting
I’ve long wanted to try and quantify the impact of Miami’s punting last year and what sort of impact average punting could have on the team. Out of 66 teams in Power-5 conferences (I’m including Notre Dame), Miami finished #62 in terms of net punting. That drops to #64 if you exclude the Savannah State game.

Some housekeeping on how I conducted my study: I split results into “Average,” “Good”, “Great” net punts. Average is the average of all teams, good is one standard deviation above the mean, great is two standard deviations above the mean. For directional punting, I applied an average punt as the 15-yard line, a good punt as to the 10-yard line, and a great punt as the 5-yard line. This means that my results are quite conservative compared to what a great directional punter would add to Miami’s expected points.

Since the entire play is driven by the punter, I hold returns against the punter (net punting). The punter can legally kick the ball out-of-bounds, kick the ball high into the air and force a fair catch, or directionally kick as well, so even if there was poor coverage, the return yards are held against the punter.

In 2018, Miami punted the ball 59 times (excluding Savannah State game), with only 16 of the punts even reaching “average” status (27%) and only 10 punts reaching “great” status on the season (17%). This means that 73% of the punts Miami had on the season actually cost the team expected points.

The totals come out this way:
View attachment 85912

With just an average punting game (38 net yards a punt), Miami would save eight points on the upcoming season. With a good punting game (40 net yards a punt), Miami would save fourteen points. With a great punting game (42.4 net yards a punt), Miami would save twenty-four points! Because you cannot score fractional points, the figures are rounded up and include extra points.

Keep in mind this was with a conservative designation on directional punting. Essentially, if Louis Hedley can come in and provide us with a good punting game, he will represent an addition larger than any other position has to gain outside of the QB position.

Results
If you looked simply at the Pythagorean expectation of Miami's win-loss record last year, you'd see a winning percentage of .716. This number is heavily skewed by the fact that Miami beat up on Savannah State, FIU, and Toledo. If you take only the results against P5 opponents last year, you see that Miami was expected to have a .500 winning percentage (actual record was 4-6). Giving Miami just an "average" punting game last year would have added a half win to their total expectation. Giving Miami a "good" punting game last year would have added an entire win to their total expectation. Giving Miami a "great" punting game last year would have added an entire win and a half to their total expectation without changing any other item within the program.

Adding 1.5 wins to a 10-game win expectation (P5 opponents) is a massive change and helps to illustrate how much Miami can improve simply by finding hidden points on the football field. Here's hoping Louis Hedley can provide us with a boost this year in the punting game.

Next week I will bring part two of "Hidden Points" and will explore another area where Miami can make huge gains with simply average performance. If you like my work, please give me a follow on Twitter @HurricaneVision.

giphy (2).gif
 
Advertisement
Excellent write-up Lance. I knew watching the LSU game that our punting was costing us a ton of points and momentum. LSU's avg starting field position was at their 45 yard line. That's just giving away points. The game would have been a lot closer and perhaps a different outcome with good-great punting.
 
Punt Efficiency: Based on where the punt takes place, we create an expected net punting value. Punt efficiency is simply the percentage of your punts that resulted in a net higher than the expected value.

Thanks for sharing. 27% of our punts added value last year. There were two punts in the San State game. One was the best of the year for Miami from an EP standpoint. One was a negative EP, but to a lesser extent.

Interesting how he defines it, as I included that info in my write up as well, but took it a different direction overall. Lots of smarter people than me out there.

Hopefully what I’ve done here adds some actual context to the number of points lost and how it can translate to wins.
 
Outside of QB, Punter was our worst position on the field last year. I'm not asking for much but please put us in the 38-40 range. All we need to do is help our defense and flip the field.
 
Advertisement
Probably poorly worded. My intention was to express that you're adding the 1.5 wins over only a 10-game sample (because I am talking only about the P5 games there).

The "win expectation" was stating a "win expectation over 10 games" essentially.
Thanks!
 
Outside of QB, Punter was our worst position on the field last year. I'm not asking for much but please put us in the 38-40 range. All we need to do is help our defense and flip the field.
Although its hard to imagine anything could be worse than our qb play last year, i think punting managed to do exactly that. Sure bad qb play had a bigger impact because, well, quarterback. But our punting last season was the worst ive ever seen at any level of football.
 
Back
Top