So how good was Gordon "Duke" Keats anyway?

[wilfpaiement]

Duke Keats. By all accounts, the man was an absolute beast. A stud center in the Mark Messier mould, not as swift a skater but a much better stickhandler and playmaker.

It's always been something of an obsession with me - the question "How would X have performed in the modern NHL?"

Now, of course there's no way to answer this question perfectly. There will always be some conjecture and assumption in any cross-era "translation" exercise. However, with numbers sometimes you can arrive at what appears to be a fairly decent result.

Take the case of Mr. Keats. I took his four core seasons:

So right off the bat we have assumption #1 - that we can simply tally up these four seasons to get a nice, fat sample size of The Iron Duke in his prime. This is a huge assumption and very likely a bad one. After all, the level of talent, competition etc. in the "Big 4" Alberta Hockey League and Western Canadian Hockey League in the early 1920s is (it's safe to assume) lower than that of the NHL in 2022-23, which is the season I ended up using as my base - to "translate" Keats' numbers to.

Is there any way to rectify this situation? I don't know. If anyone has somehow managed to measure league equivalencies going back to these long gone eras in hockey history, well, then I'd love to see it. However, for the sake of this exercise, I have no choice but to gloss over this assumption and proceed in my cross-era translation.

Okay, so I tallied up Keats' prime four seasons. Basically what we're doing here is satisfying the need for a big enough sample size (in this case 77 games total). As these numbers are spread out over 3+ years we're more getting a general, in-his-prime snapshot of Duke Keats rather than a true moment in time.

So ... what we end up with for a tally is a scoring line for Keats of: 77 - 95 - 58 - 153 - 196

From here we get a scoring "temperature" for the four seasons. We then determine how Keats performed in that climate and extrapolate these numbers to the 2022-23 NHL league averages for goals, assists and penalty minutes. Finally we assume - from anecdotal evidence and a full NY Times statistical record (including ice times for every player) for the 1927-28 season - that Keats, as a stud center, played 44 minutes a game. If you think that's off, well, you're entitled to believe that, but the rosters for Keats' Eskimos seem to suggest there were other centers that "subbed" him. So I went with 44 minutes of ice for an estimate.

The final result, if you're into this kind of historical study - putting flesh on the bones of dead men - is not surprising given all we know about Keats from sportswriters of the day as well as interviews with Keats' peers. Let's drop The Iron Duke into the 2022-23 NHL. Here are your "new" Top 5 scorers:

Keats was a beast. Does this prove it? No. But it certainly supports all the gushing his peers did about him back in the day.

Talk about forgotten greatness. Keats is to me one of the greatest hockey players who nobody knows a damn thing about.

Anyway, hope you enjoyed this little exercise.

Click to read more...

 

[jigglysquishy]

I wouldn't go so far as to say Keats is a 140 point player in 2023, but he was definitely a star player.

We voted him the 9th top centre/rover of the pre-merger era. If we limit it to the 1915-1930 time frame, he comes up a bit, but he's not a top 3 centre of his era.

For guys who were in their prime 1915-1930, he's clearly behind Nighbor, Taylor, Lalonde, Malone, and Fredrickson. MacKay and Keats are close and to me the clear 3rd/4th best centres of their era in western Canada.

MacKay and Keats will get compared a lot. MacKay the clear defensive edge. Keats the clear offensive edge.

 

[wilfpaiement]

I wouldn't go so far as to say Keats is a 140 point player in 2023, but he was definitely a star player.

We voted him the 9th top centre/rover of the pre-merger era. If we limit it to the 1915-1930 time frame, he comes up a bit, but he's not a top 3 centre of his era.

For guys who were in their prime 1915-1930, he's clearly behind Nighbor, Taylor, Lalonde, Malone, and Fredrickson. MacKay and Keats are close and to me the clear 3rd/4th best centres of their era in western Canada.

MacKay and Keats will get compared a lot. MacKay the clear defensive edge. Keats the clear offensive edge.

Excellent comments, thank you. Fascinating. You mentioned "We voted." Who's "we"? Just curious.

I just now did a similar cross-era exercise for Russ Bowie, using his six best "seasons" to get a 52-game sample. Anyway, I know these exercises are a very imperfect science, but I'll share anyway. Bowie I assumed 50 minutes of ice per game. His league scoring climate was:

5.94 goals per game
2.37 assists per game
9.10 PiM per game

Bringing Bowie to 2022-23 standards (3.13 / 5.30 / 9.19):

Now, again, I realize this is far from perfect science. I guess I'm just one of those people who likes to get a "portrait" of someone who played long ago in the context of "my" times.

Do I think prime Duke Keats would have been good for 138 points or Bowie would have been good for 143 in the modern NHL? Likely not. There are just too many factors at play (conditioning/fitness, diet, size, how the player would respond to development, etc.) There are a hell of a lot of assumptions in cross-era anything and I completely admit it. However, it does put a little more flesh on the bones of dead men, I think, if only in relation to each other.

Looking at Keats and Bowie, for example, we see that one guy (Keats), while good for more than his share of goals, was a master playmaker. This is supported by all we know about the man from those old sports pages. Bowie, however, we see a true sniper who also set up goals though not at Keats' clip.

Connor McDavid's line for 2022-23 -----> 82-64-89-153
Russ Bowie's line for 2022-23 -----> 82-65-78-143

This is of course not to say Russ Bowie - the Gretzky of his era - would have been breathing down McDavid's neck in the modern game. McDavid is a freak of nature. And yet, you read those old game accounts, you read about how good this guy Russ "Dubbie" Bowie was, and you think maybe, just maybe, if he'd had the same benefits growing up - the nutrient-dense diet, the great equipment, access to gyms, training and development etc. - maybe he WOULD have been another McDavid. This, I think, is where cross-era exercises shine. They give us just a little bit more "meat" in our arguments. However, the limitations remain.

 

[Professor What]

Excellent comments, thank you. Fascinating. You mentioned "We voted." Who's "we"? Just curious.

I think there's a list of the voters on the individual round threads for the project. Even if not, a lot of us who voted were active in the participation for the project. Here's a link to the main thread for the project though: HOH Top-60 Pre-Merger Players Of All-Time Hopefully it'll help.

 

[rmartin65]

There are a lot of issues with the stats from the amateur era, as they were not officially kept- everything we have is from what people have compiled from the papers, and coverage of the games was not always very in depth. There are several seasons where we don't have all the goals scored. There are dozens of goals that different papers awarded to two, three, or even four players- I've been splitting credit for those goals, but that isn't a perfect system. And if there is this much confusion over goals, you can only imagine what the assists look like.

I have been through the scans for the AHAC/CAHL/FAHL/ECAHA through 1909 in two Montreal papers and two Ottawa papers; all of the data can be found in the thread that @Professor What listed above. I won't claim that it is 100% accurate, but everything was compiled in good faith.

I am also making my way through another Montreal papers and a Quebec City paper, with some pretty positive results; I'll be posting updated scoring tables shortly (I've finished with the AHAC- for a spoiler, Bob 'the demon' MacDougall was a big winner in terms of new-found data).

 

[wilfpaiement]

There are a lot of issues with the stats from the amateur era, as they were not officially kept- everything we have is from what people have compiled from the papers, and coverage of the games was not always very in depth. There are several seasons where we don't have all the goals scored. There are dozens of goals that different papers awarded to two, three, or even four players- I've been splitting credit for those goals, but that isn't a perfect system. And if there is this much confusion over goals, you can only imagine what the assists look like.

I have been through the scans for the AHAC/CAHL/FAHL/ECAHA through 1909 in two Montreal papers and two Ottawa papers; all of the data can be found in the thread that @Professor What listed above. I won't claim that it is 100% accurate, but everything was compiled in good faith.

I am also making my way through another Montreal papers and a Quebec City paper, with some pretty positive results; I'll be posting updated scoring tables shortly (I've finished with the AHAC- for a spoiler, Bob 'the demon' MacDougall was a big winner in terms of new-found data).

Wonderful. Just wonderful. Please report when ready to make the big reveal.

As a side question, will your updated stats be provided to SIHR so they can update their player records?

 

[wilfpaiement]

Below are translations of PCHA, NHA and early NHL stars to 2022:23, just for kicks along with a table that shows the ice times estimates I refer to in my calculations. Critiques and suggestions always welcome - always happy for the chance to refine. Cheers ... and check out the big, fat numbers on Howie Morenz.

 

[rmartin65]

Wonderful. Just wonderful. Please report when ready to make the big reveal.

Here. I'm looking for feedback on how people prefer the data presented (season by season? totals at the very end, not after each season? something else?), so if you have any ideas, let me hear 'em. Feedback will shape how I present these stats in the individual AHAC, CAHL, FAHL, ECAHA, WPHL, and IPHL threads that I plan on making over the next several months.

As a side question, will your updated stats be provided to SIHR so they can update their player records?

I am not an SIHR member, so I won't provide the stats to them, but the SIHR is absolutely free to use them if they so choose. They have their own stat guy, though, so I imagine they won't be interested.