Another QB ranking from football perspective

[TanksAndSpartans]

This one adjusts QB rating by era. I'm far from an expert on QB rating, but Wikipedia shows that it uses data from 1960-1970 to find averages (I guess this is where the 66.7 number derives from?), so my guess would be they adjusted the average for different time periods and maybe as a result some of the earlier players stood out as being above average:

http://www.footballperspective.com/adju ... thru-2018/

There is also a link at the top of the page to the single season ranking where Luckman is first for his '43 season.

We have Frankie Albert and Ken Anderson in the HoVG, looks like the others are HoFers or active players. Brady is 15th.

If someone can follow what they did to adjust passer rating better than me, definitely chime in - obviously I don't completely follow it, but I never let it stop me from looking at what they come up with.

[ChrisBabcock]

What I think they’re doing without digging too deep into the article is comparing passer rating against what the average rating for that season (or era) was. For example, if a QB’s rating is 10% better than whatever average is, then his actual rating gets multiplied by 1.1 to get his era adjusted rating.

[Rupert Patrick]

I came across this problem when trying to rate QB's against their contemporaries in my manuscript, in fact, I refer to it as the Luckman effect. When you have a small league of ten teams or less, and you have one guy who is really good, and the rest of the league who is mediocre or less, and you compare him against the performance of his contemporaries, it tends to overrate his performance to a degree as he is not being compared to average players but actually below-average players. I ran into this when trying to rate Baugh, Luckman, Graham in his AAFC years, and I think it affected Len Dawson during his AFL years, and it could be argued that it might have also affected kickers also, particularly Lou Groza. The work-around I came up with was by splitting passers into two lists, 1933-69 and 1970-present.

[ChrisBabcock]

I came across this problem when trying to rate QB's against their contemporaries in my manuscript, in fact, I refer to it as the Luckman effect. When you have a small league of ten teams or less, and you have one guy who is really good, and the rest of the league who is mediocre or less, and you compare him against the performance of his contemporaries, it tends to overrate his performance to a degree as he is not being compared to average players but actually below-average players. I ran into this when trying to rate Baugh, Luckman, Graham in his AAFC years, and I think it affected Len Dawson during his AFL years, and it could be argued that it might have also affected kickers also, particularly Lou Groza. The work-around I came up with was by splitting passers into two lists, 1933-69 and 1970-present.

Are you including in in the average or comparing him against the average of everyone except him?

[Rupert Patrick]

I came across this problem when trying to rate QB's against their contemporaries in my manuscript, in fact, I refer to it as the Luckman effect. When you have a small league of ten teams or less, and you have one guy who is really good, and the rest of the league who is mediocre or less, and you compare him against the performance of his contemporaries, it tends to overrate his performance to a degree as he is not being compared to average players but actually below-average players. I ran into this when trying to rate Baugh, Luckman, Graham in his AAFC years, and I think it affected Len Dawson during his AFL years, and it could be argued that it might have also affected kickers also, particularly Lou Groza. The work-around I came up with was by splitting passers into two lists, 1933-69 and 1970-present.

Are you including in in the average or comparing him against the average of everyone except him?

His contemporaries are everybody except for him. When I evaluate Sid Luckman in the context of the 1943 NFL, I evaluate his performance against the aggregate performance of the entire league with his statistics removed.

Due to the law of averages, sometimes in a small group, the collective average performance of the group is not only below average but well below average; this is a sample size issue. The larger the group or sample size, the less the possibility of this occurring. In the late 30's and early 1940's NFL, you had two truly great QB's, Baugh and Luckman, and Herber was quite good also. The rest were fair to poor, and a number of them played at or near what the baseball stat people refer to as replacement level. And things only got worse when WWII came along and the quality of NFL players in general started to peter out for a couple seasons. The gap between the stars who stayed behind and continued to play in the NFL, like Baugh and Luckman, and the rest of the league, widened considerably.

I'm sorry, but it is difficult to defend any kind of rating system that concludes that Sid Luckman's 1943 season was the greatest of all time when he faced teams whose rosters were decimated by the war effort. In 1943, the Eagles and Steelers had to merge for the season in order to help keep the league afloat after the Rams chose not to play the 1943 NFL season

[TanksAndSpartans]

Good points everyone. I noticed as well that the analysis seemed to be benefitting the early passers who were much better than the pack. I think the pack caught up as the game evolved. 9/10 of the best single seasons were prior to '54. Go back a little further and think how much better Friedman was than the the pack in '27. With better records, that probably winds up first. There might just not be a good way to compare across eras. Marino's '84 season came in at 35th.

[Rupert Patrick]

Good points everyone. I noticed as well that the analysis seemed to be benefitting the early passers who were much better than the pack. I think the pack caught up as the game evolved. 9/10 of the best single seasons were prior to '54. Go back a little further and think how much better Friedman was than the the pack in '27. With better records, that probably winds up first. There might just not be a good way to compare across eras. Marino's '84 season came in at 35th.

There really isn't a good way to adjust for eras when it comes to passing stats. The best way to do it is to compare QB's to his contemporaries, and even when you do so, Baugh's 1940 season (when he had a 12-10 TD-INT ratio) rates a little higher than Aaron Rodgers's 2011 season when he had a 45-6 TD-INT ratio. The best way to do it is to create a list for the guys from the early years and a list for the modern guys.

[Jay Z]

Any QB rate is just that, a rate stat. Bob Griese had a great QB rate in Super Bowl VIII. He attempted 7 passes.

In eras with less passing, the position should be discounted, since it doesn't have the same level of impact.

Also, Luckman 1943 was competing against tailbacks in a single wing. 3rd highest in passing attempts what Tony Canadeo. What does that tell you? Canadeo ran the ball, Luckman did not.

[sluggermatt15]

Good points everyone. I noticed as well that the analysis seemed to be benefitting the early passers who were much better than the pack. I think the pack caught up as the game evolved. 9/10 of the best single seasons were prior to '54. Go back a little further and think how much better Friedman was than the the pack in '27. With better records, that probably winds up first. There might just not be a good way to compare across eras. Marino's '84 season came in at 35th.

There really isn't a good way to adjust for eras when it comes to passing stats. The best way to do it is to compare QB's to his contemporaries, and even when you do so, Baugh's 1940 season (when he had a 12-10 TD-INT ratio) rates a little higher than Aaron Rodgers's 2011 season when he had a 45-6 TD-INT ratio. The best way to do it is to create a list for the guys from the early years and a list for the modern guys.

Rupert, this is excellent insight into comparing QB passing over time. How much do you adjust for evolution of the game? For instance, we always hear the Tom Brady is GOAT argument, but he is playing in an era that heavily favors offense. Take a QB like Joe Montana or Johnny Unitas, who played in more defense-favored times, and is there a way to compensate? My view is that these two would thrive in today's game, but a QB from today would not do as well in the 1960s or 1980s. What do you think?

[JuggernautJ]

....In eras with less passing, the position should be discounted, since it doesn't have the same level of impact....

Not to argue but in a way they did more than modern QBs.
Many of the quarterbacks of whom we would be speaking called their own plays.

So I guess I agree with Rupert in that we should think of the eras differently.
Before the helmet-radio (or messenger guards, etc) Quarterbacks called plays and directed the offense on the field.
Nowadays (post whatever criteria you choose to use) Passers throw the football where they are told to.

The criteria for the position has changed as have the rules (especially regarding the passing game).
Not only are there difference between eras but one might almost argue the two positions are the same only in name (and the fact that both throw the ball).

I believe we've discussed this before but where would one draw the line in eras?

I would perhaps consider three distinct "eras."
Before 1940 or so (when passing was a rarity), 1940-1978 (??) when the rules changed and post-1978 for the "modern" passing game.