Page 2 of 2

Re: Reggie Roby HOVG? HOF?

Posted: Wed Oct 31, 2018 2:28 pm
by Ronfitch
JWL wrote:The best metric hasn't been made into a statistic yet. There should be something regarding the best possible outcome for each punt. For example, Joe Mama punts from the 50-yard line. The best outcome is the punt pins the opponent at the 1-yard line. A touchback, in this particular scenario, would not even really be good. A touchback could be scored as a -19. If the punt is returned and the returner takes it to his team's 30 yard line, then you could score it a -29.

I'm just throwing something quick out there. I think the punting stats as they exist now need a ton of context.
Not a jersey-owning kind of guy, but I would consider one if it was a Joe Mama replica jersey.

Back on topic - I grew up next to Waterloo, Iowa, where a generation/half-a-generation before, where Roby attended Waterloo East and was firmly entrenched in local lore for his thundering punts as a high school player.

He was also a pitcher in high school, drafted by the Reds and intended to play both football and baseball at Iowa when he accepted the scholarship (he earned all state honors as a punter and tight end in high school, also played linebacker and basketball . https://hawkeyesports.com/news/2005/2/2 ... _Dies.aspx).

When former Iowa assistant Dan McCarney scouted Roby as a high school senior in the first game of that season, he saw Roby drive the opening kickoff through the opponent's goal post (http://data.desmoinesregister.com/hall- ... php?id=364).

Re: Reggie Roby HOVG? HOF?

Posted: Mon Jan 14, 2019 5:45 pm
by RRMarshall
Until "Hang time" is kept as an official statistic getting a proper metric may be extremely arduous. Ray Guy was the King of Hang Time, and it took him forever to get in. I actually had an argument with Peter King about Guy's viability as a candidate. His argument was looking at Guy's numbers his net punting average wasn't significantly above that of other punters in his era. I countered with Guy's hang time and the difficulty opponents had returning his punts because of it. The following year King relented and voted for Guy, the year he was elected. I still haven't gotten my Thank You card in the mail from Ray!!

Re: Reggie Roby HOVG? HOF?

Posted: Mon Jan 14, 2019 6:29 pm
by JohnTurney
RRMarshall wrote:Until "Hang time" is kept as an official statistic getting a proper metric may be extremely arduous. Ray Guy was the King of Hang Time, and it took him forever to get in. I actually had an argument with Peter King about Guy's viability as a candidate. His argument was looking at Guy's numbers his net punting average wasn't significantly above that of other punters in his era. I countered with Guy's hang time and the difficulty opponents had returning his punts because of it. The following year King relented and voted for Guy, the year he was elected. I still haven't gotten my Thank You card in the mail from Ray!!

What King and others missed is that the difference in net punting is razor thin...meaning even though Guy was not that much ahead of the average over his career, he WAS the best in that era in terms of highest net over the average.

Re: Reggie Roby HOVG? HOF?

Posted: Mon Jan 14, 2019 9:44 pm
by Rupert Patrick
JWL wrote:The best metric hasn't been made into a statistic yet. There should be something regarding the best possible outcome for each punt. For example, Joe Mama punts from the 50-yard line. The best outcome is the punt pins the opponent at the 1-yard line. A touchback, in this particular scenario, would not even really be good. A touchback could be scored as a -19. If the punt is returned and the returner takes it to his team's 30 yard line, then you could score it a -29.

I'm just throwing something quick out there. I think the punting stats as they exist now need a ton of context.
I'm responding to a four-month old post here, but it caught my eye. I've always felt along the same lines for a stat that covers something similar to best possible outcome, and in my opinion it is by measuring punts as a percentage based on the maximum length of a punt before it becomes a touchback. Let's say a punter's line of scrimmage is at his 40-yard line, and booms a 58 yard punt to the opponents 2-yard line. Of the 60 yards the punt could possibly travel before it becomes a touchback, it actually covered 58 of those 60 yards, and then you make a percentage of that 58 divided by 60 result, and keep track of punts this way. Over time, you would make this into an average of sorts. However, if his punt had gone for a touchback, his punt percentage, if you will, would be 35/60 as the punt is returned to the 25-yard line, as the punt got a 25-yard deduction as it went for a touchback. The averages of punt percentage for every punt for each punter would be expressed as an average punt percentage.

Of course, this doesn't work that well if you're at your own 20, and the punter is standing at his five-yard line and punts it 45 yards (from the line of scrimmage) to the opponents 35-yard line. In that case, the punt percentage would be 55/80, which would not be that good. Such a stat would hurt punters on bad teams who punt often from deep in their territory. One way around this would be to choose a point, perhaps when the line of scrimmage is 60 yards from the end zone you are punting toward, where punters would be rated from. In other words, if the line of scrimmage is between a team's one and 39-yard lines, this punt would not count toward his average punt percentage.

I think 60 yards from the end zone you are punting toward is the right place to start this punt percentage, as i have studied large numbers of punts, noting distance from the opponents end zone and the distance the ball was punted. I did an analysis of punting in Denver and Heinz Field where I tracked distance from LOS to opponents end zone, and distance the punt traveled. From 99 to 61 yards, the number is pretty straight, but when you hit the 60-yard point, punting average drops off like a ski slope. This is the point where punters start to sacrifice distance to keep it from becoming a touchback, or angling the punt toward the sidelines. I found the same exact graph when I studied the Pittsburgh and Denver data separately. It looks like this:

Image

Image

The "Away" graphs are for all away games, for Broncos games during the Invesco Field era and for Steelers games during the Heinz Field era, which cover roughly 2000 or 2001 to 2017. From further than 60 yards out, the punter is generally trying to punt is as far and as long as he can, but once that number hits 60, is the place that separates the good punters from the bad punters. I haven't studied the data, but I would guess with few exceptions, all punters would have a pretty similar punting average on punts between 99 and 61 yards from the end zone they are punting toward, the variation or standard deviation of punting averages would be much lower from 99 to 61 as opposed to 60 and lower.

To make a long story short, I think average punt percentage from 60 yards and closer may be the best way to rate punters. I don't have complete punt data for all punters, but that information can be extracted from game books pretty easy. The formula would look something like this, I think:

Punt Average = Punt Yards / 100 - New Line of Scrimmage + Punt Return Yards + 25*Touchbacks