Professional Football Researchers Association Forum

James wrote:

JWL wrote:The real Mount Rushmore will never change. I think once you decide on your personal Mount Rushmore for something, then, boom, that is was it is. You have to pretend it like the real Mount Rushmore.

You are absolutely right, as originally Baugh, Graham, Unitas, Montana, but I cannot overlook Brady and what he's done accomplished season after season that's why reluctantly I decided to drop Baugh and have it Graham, Unitas, Montana, Brady which this version, with me, will be the one I have till he day I die.

Yeah, that's good. I wasn't being 100% serious earlier. I do think, though, that the idea of a Mount Rushmore is to have four great and important and influential people on it. They must be great but not necessarily the greatest. We should each decide on who those guys are and then set the group in our own figurative stone.

Graham, Unitas, Montana and Brady front and center with Baugh and Manning just slightly farther back on the sides of the mountain.

I took a different approach. I said that I was compiling a Mt. Rushmore of NFL quarterbacks and replaced Graham with Brady. This is not to denigrate Graham's time in the AAFC; I simply omitted it. Graham's 6 years in the NFL were not at all shabby, but Brady's 18 years trump that.

I couldn't take off Baugh. He spent his career throwing off his back foot as linemen (albeit 240 lb linemen) could take more than 1 step to hit him.

conace21 wrote:I couldn't take off Baugh. He spent his career throwing off his back foot as linemen (albeit 240 lb linemen) could take more than 1 step to hit him.

Another perspective: If you're taking impact on the game into account, as somebody else said, I don't think you can omit Baugh. In fact, he might be No. 1 in that regard.

Bob Gill wrote:

conace21 wrote:I couldn't take off Baugh. He spent his career throwing off his back foot as linemen (albeit 240 lb linemen) could take more than 1 step to hit him.

Another perspective: If you're taking impact on the game into account, as somebody else said, I don't think you can omit Baugh. In fact, he might be No. 1 in that regard.

That's why my preferred Mt. Rushmore has six heads on it. I don't see how you can leave Sammy and Peyton off of it.

James wrote:
Damn, what a fantastic post, Rupert. Math and numbers were NEVER my strong suit, still aren't, but damn, I am totally fascinated by your QB Time Machine. Curious as to how Sid Luckman is on your QB Tme Machine some day.

The QB Time Machine is an excel program I designed where it translates a passer's career, season by season, to another era. It is built from my NPR (Normalized Passer Rating) methodology, which is similar to NFL Passer Rating, but instead of comparing the QB to fixed standards, it compares him to the league average (with his stats removed) in that given season. If you understand PR, there are four categories in which you receive points, Completion %, Yards per Attempt, TD % and Interception %, and NPR is exactly the same. The difference to NPR is that each season, with a slight algebraic adjustment, the average NPR for the league comes at exactly 66.7. In this way, an NPR score for a passer fro 1988 compares pretty well to a passer from 2018. Because the NPR scores translate pretty well from one era to the next, you can translate a guy's stats from one era to the next.

However, there are limits to doing this. I have a rule which I refer to ironically as the Luckman effect, and it goes something like this - when you are dealing with small leagues (10 teams or less), and rating players compared to their peers and removing said player's stats from the league stats to do so, and said player is markedly better than the other players in the league, it over-inflates his value due to the small sample size. It is difficult to put into words, but if you look at the top 25 individual seasons for NPR from 1933 to 2017, NFL, AFL and AAFC, nine of them occurred in the 1940's. Four occurred in the 60's, two each in the 50's, 70's, 90's, 00's and 10's, and one each from the 30's and 80's. This is a little abnormal. Here is the top 25 passing seasons from 1933-2017, NFL, AFL and AAFC, split by the 1970 merger:



We'll look at Baugh's 1940 season, which was fifth on the 1933-69 list. The main reason he finished so high was his 12-10 TD-INT ratio, compared to 88-213 for the rest of the NFL. I cannot defend his NPR being higher than that of Aaron Rodgers from 2011, who had a TD-INT ratio of 45-6. The 1940 season was one of the three or four best seasons of Baugh's career, not a season of historic magnitude, but because nobody else in the league had a really decent season, it made his stats look much better when compared to the rest of the league. For this reason, I really cannot tranlate Sid Luckman's career to the modern era.

The system needs a lot of work from people who understand career projections a lot better than I do, but here is what Sonny Jurgensen's career looks like when translated from 2001 to 2018:



The time machine sees Jurgensen as throwing a lot more passes, which of course is due to the 16 game schedule and that more passes are being thrown now than back then. Also, he career completion percentage goes up to 67.8 percent, but his TD percentage drops a little because fewer TD passes are thrown on a percentage basis these days. There are far fewer interceptions nowadays than in the 60's and 70's, so his interception totals dropped. However, his NPR for each projected season is virtually the same as for his actual season. It doesn't yield any actual knowledge, but is fun to play with.

Very interesting work, Rupert. One question: I'm guessing you didn't include Frank Filchock's 1939 season because he didn't throw 100 passes, because otherwise I think he'd show up pretty near the top for that year. Is that right? He did, however, qualify for the NFL's leadership, though he somehow failed to lead the league by whatever system they were using at the time.

Being that we soon will close the books on the 99th season, I imagine there will be all sorts of lists this offseason and then during next season. Either this offseason or next when 100 seasons are in the books (what books these are and where one can found them I am not actually sure), I will decide on my Mount Rushmores for all the key positions. I am sure we will have loads of threads here. I don't care what I did in the past. These Mount Rushmores will be my official ones that I will stick to for the rest of my life.

Bob Gill wrote:Very interesting work, Rupert. One question: I'm guessing you didn't include Frank Filchock's 1939 season because he didn't throw 100 passes, because otherwise I think he'd show up pretty near the top for that year. Is that right? He did, however, qualify for the NFL's leadership, though he somehow failed to lead the league by whatever system they were using at the time.

Flichock in 1939 threw 89 passes and had an NPR of 135.42. For seasons from 1933-1969, I made a rule of eligibility for the leader tables of 10 attempts per scheduled team game and from 1970-2017 observe the NFL standard of 14 attempts per scheduled team game. By the way, the 1939 Redskins played 11 games.

I wrote an article on Normalized Passer Rating for the Coffin Corner 7-8 years ago, which is available wherever fine PFRA products are sold.

Rupert Patrick wrote:

James wrote:
Damn, what a fantastic post, Rupert. Math and numbers were NEVER my strong suit, still aren't, but damn, I am totally fascinated by your QB Time Machine. Curious as to how Sid Luckman is on your QB Tme Machine some day.

The QB Time Machine is an excel program I designed where it translates a passer's career, season by season, to another era. It is built from my NPR (Normalized Passer Rating) methodology, which is similar to NFL Passer Rating, but instead of comparing the QB to fixed standards, it compares him to the league average (with his stats removed) in that given season. If you understand PR, there are four categories in which you receive points, Completion %, Yards per Attempt, TD % and Interception %, and NPR is exactly the same. The difference to NPR is that each season, with a slight algebraic adjustment, the average NPR for the league comes at exactly 66.7. In this way, an NPR score for a passer fro 1988 compares pretty well to a passer from 2018. Because the NPR scores translate pretty well from one era to the next, you can translate a guy's stats from one era to the next.

However, there are limits to doing this. I have a rule which I refer to ironically as the Luckman effect, and it goes something like this - when you are dealing with small leagues (10 teams or less), and rating players compared to their peers and removing said player's stats from the league stats to do so, and said player is markedly better than the other players in the league, it over-inflates his value due to the small sample size. It is difficult to put into words, but if you look at the top 25 individual seasons for NPR from 1933 to 2017, NFL, AFL and AAFC, nine of them occurred in the 1940's. Four occurred in the 60's, two each in the 50's, 70's, 90's, 00's and 10's, and one each from the 30's and 80's. This is a little abnormal. Here is the top 25 passing seasons from 1933-2017, NFL, AFL and AAFC, split by the 1970 merger:



We'll look at Baugh's 1940 season, which was fifth on the 1933-69 list. The main reason he finished so high was his 12-10 TD-INT ratio, compared to 88-213 for the rest of the NFL. I cannot defend his NPR being higher than that of Aaron Rodgers from 2011, who had a TD-INT ratio of 45-6. The 1940 season was one of the three or four best seasons of Baugh's career, not a season of historic magnitude, but because nobody else in the league had a really decent season, it made his stats look much better when compared to the rest of the league. For this reason, I really cannot tranlate Sid Luckman's career to the modern era.

The system needs a lot of work from people who understand career projections a lot better than I do, but here is what Sonny Jurgensen's career looks like when translated from 2001 to 2018:



The time machine sees Jurgensen as throwing a lot more passes, which of course is due to the 16 game schedule and that more passes are being thrown now than back then. Also, he career completion percentage goes up to 67.8 percent, but his TD percentage drops a little because fewer TD passes are thrown on a percentage basis these days. There are far fewer interceptions nowadays than in the 60's and 70's, so his interception totals dropped. However, his NPR for each projected season is virtually the same as for his actual season. It doesn't yield any actual knowledge, but is fun to play with.

I hate to dig up an old thread from months ago, but as I am working on my manuscript for my football history book, I decided not to include the QB Time Machine in my manuscript in order to keep the manuscript under what feels like 3,000 pages. What I have decided to do instead is publicly release the Excel spreadsheet for the QB Time Machine version 1.0 at this time now, along with a Word document primer explaining how it works and how to use it, and just reference it in my manuscript and say that it can be found in a google search as it will be "on the web". It works for about 90 percent of quarterbacks but there are some (Sid Luckman and Sammy Baugh) that it doesn't work well for. In the primer, I issued a challenge for those who are experts with stats or career projection that they are more than welcome to create version 2.0, to play around with it and take a crack at trying to fix the inherent problems in the spreadsheet that I couldn't work around.

Therefore, if you want to play with this, plug in different QB's to see what Bob Griese's stats might look like if he had started his career in 1985, or if you want to see if you can take a crack at fixing the problems with this spreadsheet, PM me with your email address and I will send you the spreadsheet and primer. You will first need to read the primer to understand how to use the spreadsheet before you start playing with the spreadsheet, and if you want to understand how I came up with the Normalized Passer Rating system which led me to the four equations that i use in the spreadsheet, I recommend you read the Normalized Passer Rating article I wrote that is in the Coffin Corner archives.