1932

[JameisLoseston]

So I've come across a problem in some of the research I'm doing that I couldn't have foreseen. PFR considers 1932 to be the first officially statted NFL season, whereas David Neft considers it the last unofficial season, and I have no idea who's right. Normally, I would just assume Neft accidentally calculated PBPs for a season he didn't need to, but the major problem with that is that some of his numbers actually exceed the "official" stats. For example, season "MVP" Arnie Herber is listed at 639 passing yards on PFR, but Neft gives him 774, and he considers all of GB's games complete. Despite that, he also gives Herber 3 fewer INTs than PFR does. Benny Friedman has 319 yards on PFR and 360 in Neft, who marks 4 very incomplete games among that. All of this leaves me having no idea where PFR even got its numbers from, and why it considers them official; I can't imagine it just took them from an earlier edition of Neft, considering it doesn't do that for any preceding season. Neft seems to acknowledge the stats from 1933 on as being the same as what's listed on PFR, leaving 1932 as the only anomalous season. I'm sure this is something that has confounded members here in the past doing similar work, so which stats should I use?

[TodMaher]

So I've come across a problem in some of the research I'm doing that I couldn't have foreseen. PFR considers 1932 to be the first officially statted NFL season, whereas David Neft considers it the last unofficial season, and I have no idea who's right. Normally, I would just assume Neft accidentally calculated PBPs for a season he didn't need to, but the major problem with that is that some of his numbers actually exceed the "official" stats. For example, season "MVP" Arnie Herber is listed at 639 passing yards on PFR, but Neft gives him 774, and he considers all of GB's games complete. Despite that, he also gives Herber 3 fewer INTs than PFR does. Benny Friedman has 319 yards on PFR and 360 in Neft, who marks 4 very incomplete games among that. All of this leaves me having no idea where PFR even got its numbers from, and why it considers them official; I can't imagine it just took them from an earlier edition of Neft, considering it doesn't do that for any preceding season. Neft seems to acknowledge the stats from 1933 on as being the same as what's listed on PFR, leaving 1932 as the only anomalous season. I'm sure this is something that has confounded members here in the past doing similar work, so which stats should I use?

The 1932 stats on PFR (and my site) are the official stats.
Neft includes - I assume you are referring to the The Football Encyclopedia - 1932 with 1920-31 because that's the era before the playoffs and making the forward pass legal from anywhere behind the line of scrimmage. It has nothing to do with when the stats were official. He compiled his own stats (like he did for 1920-31) because before Total Football all the officials stats for the 1930s and 1940s were not available to the general public. He didn't do any years after that because of the different format of the 1933+ sections of the book.

[RyanChristiansen]

Unfortunately, I've run across errors in Sports Reference statistics, and I've submitted corrections, and they took them. I've never submitted corrections for the pro football site but for the college football site. Each time I pointed out an error, I submitted box scores from newspaper articles as my sources. I didn't submit any earth-shattering changes, just a few forgotten, miscellaneous carries or catches by reserve players. The site isn't perfect, that's for sure.

[JameisLoseston]

So I guess I'll go with what PFR has. Any idea why Neft calculated the stats so differently, even in cases where all games are marked complete (e.g. Herber)?

[TodMaher]

So I guess I'll go with what PFR has. Any idea why Neft calculated the stats so differently, even in cases where all games are marked complete (e.g. Herber)?

As I said, Neft went by what was available at the time (1970s and 1980s) which was next to nothing so he had to recreate it from the newspapers which outside of Green Bay and Portsmouth in 1932 was skeletal at best. Only when Total Football was published in 1997 did the public know what the official totals were.
The stats published in the papers were not always the official stats. Even the official stats of the time were revised by Elias in the 1960s when they took over as the official statisticians.

[JameisLoseston]

Great info, Tod. My edition of Neft is from 1994, so it precedes that release date by just a bit.

I suppose this means that, even for the complete games, Neft's stats for 1920-31 may contain the same margin of error from what "really" happened as found between the official stats in 1932 and Neft's?

[TodMaher]

Great info, Tod. My edition of Neft is from 1994, so it precedes that release date by just a bit.

I suppose this means that, even for the complete games, Neft's stats for 1920-31 may contain the same margin of error from what "really" happened as found between the official stats in 1932 and Neft's?

Yes, but it's the only data that is available.

You also may notice that the 1932 official individual receiving stats don't always add up to the official team totals.

Brooklyn: Individuals 37 for 560, Team 42 for 586. The reception and yards totals for John "Stumpy" Thomason are unofficial. In Total Football he has only 1 TD catch listed as his receiving stats. The 25 yards is what his TD catch was reported as in the newspapers. Same thing goes for Stu Wilson of Staten Island's stats (1 for 11).
Boston: Individuals 14 for 216, Team 17 for 216.
Chicago Cardinals: Individuals 40 for 612, Team 40 for 628.
New York: Individuals 71 for 949, Team 82 for 949.
Portsmouth: Individuals 43 for 563, Team 43 for 623.
Staten Island: Individuals 24 for 378, Team 33 for 518.

Only the Bears and Packers add up.

Of course, we will never know where the errors/discrepancies are because the NFL will probably never release the official game-by-game totals.

[RichardBak]

Of course, we will never know where the errors/discrepancies are because the NFL will probably never release the official game-by-game totals.

I'm curious--why is that?

[TodMaher]

Of course, we will never know where the errors/discrepancies are because the NFL will probably never release the official game-by-game totals.

I'm curious--why is that?

Because the NFL just doesn't care about its history.

[RichardBak]

If that's the case, I don't see why they'd oppose some established researcher/s (like yourself or Chris Willis or others) doing a deep dive into those early game records. I don't know any of the politics involved, but now I'm becoming slightly more intrigued about the 1920-32 records. If nothing else, might be gist for someone (not me) to do a feature (a blog, CC, SI?) on the whole issue. Why doesn't the NFL cooperate and "open up the books" (as Tony Soprano would say)? Can't see any particular reason why they wouldn't. Of course, I'm trying to think about it logically, which is not exactly the league's strong suit.