Early NFL "MVPs"

[Bob Gill]

As far as I know, Doc Elliott played with these teams outside the NFL:

1922: Frankford Yellowjackets (two years before the team entered the NFL)
1926: Cleveland and Philadelphia in the AFL
1927: Ironton Tanks and that Millville team
1928: Millville and Ironton
1929: Ironton again

Given that he played in an NFL game in 1931, I'd say it's likely that he played somewhere else in 1930 and '31 too.

[TanksAndSpartans]

That's awesome! Thanks Bob. I didn't remember the Ironton connection - hope they don't revoke my avatar

That's a good bit of football - combined with 3 championships, postseason honors, etc. I may have to nominate him for HOVG - worth a look at least.

[JameisLoseston]

For 1920, I agree that the selection should be someone from Akron. It's really hard to tell who was the most important player on the team, but I wouldn't be surprised if it was Pollard. It was unquestionably Pollard in 1921, but for some reason they weren't quite as good.

All-Pro teams and such are definitely a fortunate resource to have for this era, especially for players like Boynton, where it would otherwise be hard to tell if his outwardly nice numbers were actually indicative of his play, or simply artifacts of inconsistent recording that left him looking good and other guys looking less good who were actually better. But with the corroborating evidence of him being a 1st-team tailback both years, clearly he was an elite talent. The same principle goes for other players like Hamer, although his being 2nd team is a bit puzzling, considering we have plenty of evidence that he was the best player in the league for the year. Perhaps it's merely a reflection of the fact that Frankford wasn't a glamorous team, although they were very good in 24, or that he was a rookie who hadn't been an extremely popular college player.

Another Penn-to-Frankford player I wanted to mention here is Charley Rogers, who has a similar distinction to Friedman's in being the only player to lead the league in both rushing and receiving in a season, and put together the first ever 1000 yards from scrimmage in the process, also as a rookie. Unfortunately, he did it in 1927, which was a Friedman year and most closely contested by McBride, making Rogers a tough pick, especially as Frankford was not good this year and he did not score much. Still a season worth highlighting.

[JeffreyMiller]

https://nflfootballjournal.blogspot.com ... 1.html?m=1

I think this article I wrote a couple of years ago demonstrates some of the problems with relying on early All Pro teams

[TanksAndSpartans]

@JM, I commented on that article . I get it - everything has problems, but an All-Pro is a data point - it represents contemporary opinion - maybe it was biased because the voter is a huge homer or because they didn't see all the games. Some of those problems I would think still exist. Plus add in the fact of how hard it is to get people to change their mind - it may not have mattered how many other players they saw once they decided who the best [fullback or fill in the blank] was.

***

All-Pro teams and such are definitely a fortunate resource to have for this era, especially for players like Boynton, where it would otherwise be hard to tell if his outwardly nice numbers were actually indicative of his play, or simply artifacts of inconsistent recording that left him looking good and other guys looking less good who were actually better. But with the corroborating evidence of him being a 1st-team tailback both years, clearly he was an elite talent.

Well said. I agree.

The same principle goes for other players like Hamer, although his being 2nd team is a bit puzzling, considering we have plenty of evidence that he was the best player in the league for the year. Perhaps it's merely a reflection of the fact that Frankford wasn't a glamorous team, although they were very good in 24, or that he was a rookie who hadn't been an extremely popular college player.

I think Hamer was 1st team Collyer's Eye, 2nd team GBPG, not too shabby. And you make a good point about rookies - I think it can be hard for them to break through at times which is consistent with the idea of the voters already having their "favorites". But Frankford didn't win the title. I can't say without researching it more deeply whether I may go with a player on Cleveland or Chicago - I'd at least check there first before going with Hamer.

Good catch on Rogers. I agree. A nice season.

[Bob Gill]

@JM, I commented on that article . I get it - everything has problems, but an All-Pro is a data point - it represents contemporary opinion - maybe it was biased because the voter is a huge homer or because they didn't see all the games.

This is why I count only all-pro teams chosen by a panel of voters. (With about two exceptions, actually.) In theory, and I think in practice too, using a panel eliminates the problem of unusual bias, because the guy from, say, Washington in 1937 who has a special fondness for Ed Justice can vote for him, but that's the only vote Justice will get, so there's no way he'll make the all-pro team.

Occasionally, you might get somebody who does show up on an all-pro team but doesn't seem to merit the selection based on whatever stats we have available now. I think that's valuable in itself, though, because it shows what people at the time believed, even if that belief seems questionable.

[Bryan]

Having done more research into Elmer Oliphant, I've come to find out that he's actually a very inner-circle College Football HOFer, and one of the best football players of the 1910s. Looking at his stats, I figured he was probably a highly regarded college player who left pro football for all the reasons guys left back then, and that is precisely the case; it definitely wasn't because he wasn't good enough.

FWIW, Elmer Oliphant was chosen as one of the backs on the 1910-1919 College Football all-decade team.

[JameisLoseston]

To add to the discussion, here are some revelations I've gleaned since receiving my 1994 Neft encyclopedia:

All Packer games for the entire decade from 1923-32 are completely recorded except for one 1930 game. This means that the stats for the likes of Dunn, Lewellen, Blood and so on can be considered everything but official, and Curly indeed threw exactly 29 INTs in 1924.

Red Dunn should be replaced as 1925's pick, because he only has 2 complete games with the Cardinals, and 7 of his 8 recorded interceptions came in those two games. That means his 8 incomplete games are probably full of icky turnovers. Give me Jack Ernst of Pottsville instead, who had a definitive 8-13 TD-INT ratio for the "other" champs, whose games are very complete.

Since Pottsville's games are so complete, that also means Barney Wentz will probably be usurped for 1926 by Paddy Driscoll, who had 12 incomplete games to add onto his totals. This works to the opposite effect as it did for Ernst, because Wentz is a running back who could only benefit from more counting numbers, whereas quarterbacks incur less risk of hidden turnovers with more complete stats.

It's fairly likely that Hamer had over 1000 yards in 1924. He only has 2 complete games, although by my estimation, the majority of his yardage from the incomplete games is recorded, so he probably isn't far over 1000. I have him with about 1100 yards on just over 200 carries.

Friedman's games align roughly with my original estimate of about 90% of his passing stats presently recorded, but maybe only around 50% of his rushing. The only immediate threat I've found to any of his four consecutive MVP claims is Charley Rogers in 1927; two-thirds of his games are complete, and the other third, strangely, have nothing recorded. He may well have had almost 1500 scrimmage yards, but also played in like 18 games; you be the judge of whether that makes one more or less MVP-worthy.

[Bob Gill]

Friedman's games align roughly with my original estimate of about 90% of his passing stats presently recorded, but maybe only around 50% of his rushing. The only immediate threat I've found to any of his four consecutive MVP claims is Charley Rogers in 1927; two-thirds of his games are complete, and the other third, strangely, have nothing recorded. He may well have had almost 1500 scrimmage yards, but also played in like 18 games; you be the judge of whether that makes one more or less MVP-worthy.

Shifting the focus from scrimmage yards to all-purpose yards, did you notice Rogers' punt return stats in 1927 and '28? He was over 500 yards each year. And I think in one of those years he's the leader in interceptions too, at least as far as we can tell.

[JameisLoseston]

Yes, it's all in 1927:

120 carries, 508 yards, 2 TD
38 receptions, 541 yards, 1 TD
44 returns, 620 yards
7 defensive INTs

All his recorded games are complete, but 5-6 are unrecorded. A very fascinating season, I'd bet the over on 2000 total yards. Why the lack of touchdowns, though? 3 TDs in 2000 yards is hilariously low even by today's standards (although he could've had more, since his missing games are completely missing). I'm assuming he's another college star who either got hurt or left.