Is it time for another top 100 or so project?

[bobholly39]

---Best single season teams of all-time
Would it include European teams, like CSKA Moscow? (CSKA might actually be the only European team qualified? But they might have been the best team ever.)
How do you determine the 1970s MTL? Were they really as good as CSKA? And was MTL actually better than previous great NHL teams if looking only at their results against the other O6 teams?
How to evaluate last years Tampa Bay, that had a great regular season but not a good playoff?
What would we learn and gain?

---Best single seasons of all-time by a player
Doesn't this topic come upp all the time? And wouldn't it mostly just be personal preferences determining how we'd rank players? Which Gretzky season is his best? Which Mario season? Was Lafleur even the best player in the world during his peak? (Yes, I guess most North Americans here would say.)
How to treat great players on teams that didn't go far into the playoffs?
What would we learn and gain from a project like this? Wouldn't it just be very familiar players being discussed yet another time? That could as well be done in an ordinary thread?

---Best coaches of all-time (though this might be too research intensive...)
Seems very interesting, especially since coaching can greatly affect the stats and careers of individual players.
But, unforunately (as you and others say)... how would we be able to determine how good the coach or coaching actually was?

---HOH Hall of Fame or Hall of Excellence (we've talked about this one the most in the past)
As a Swede, I'm still surprised about the North American focus on HOF. Like "What does it really matter which players are 'in' or not?". What would the criteria be? Wouldn't the results be quite similar to the best/greatest player rankings? (Or would guys like Börje Salming and Sven Tumba end up much higher on this list than on the previous rankings? Tumba was like "Mr Hockey" in Sweden, and Salming is a legend here.)

The best single season of all-time:

It would have to be regular season only, not playoffs. If you include playoffs - i think it changes from "let's find the best of the best" to "let's identify the most fortunate timing of all time", which sounds a lot more boring to me. Some players have their best season and playoff happen in same year - but more often it does not, since playoff success is team dependent. Fedorov in 1994 had a regular season people are in love with - but in the playoffs his team got bounced in round 1 by SJ (very embarrassing loss).
If we somehow insisted on wanting to include playoffs - I think it should be "best season + best single playoff run" combo - but without them being in the same year. So sticking to Fedorov - you could combine his 1994 season to his playoff run in 97 or 98. This is probably a bit too complex though - single best regular season is preferable to be.

In many ways - there would be 2 layers to this project. First - you have to identify the player's best season. Second - you have to compare it to others. So deciding which of Gretzky's season is best for example is a debate in itself - and then you have to compare it to the best from Orr and Lemieux too.

As to "this topic coming up all the time" and "what would we learn and gain" - I disagree, I don't think we've ever done anything of the sort actually. It's true that at the top we'd rank 1 of Gretzky, Orr, Lemieux Howe seasons, and that always happens. Whose #5? Seasons like Sakic 2001, or Fedorov 1994, or Yzerman 1989, or many other players who normally are nowhere near discussion for "~top 10" would come into play. Have you ever compared Sakic or Yzerman's best season to Beliveau or Hull's best?

Obviously for this to make sense - rank one season per player only. We don't want half the list to be 25 different seasons of Gretzky/Orr/Lemieux/Howe in some fashion at the top. This also makes it easier. Because I can say in a heartbeat that Gretzky's best season is better than Hasek's. But what about his 6th best? or 8th best? I think having more than one season per player would just bog the process down.

 

[Theokritos]

So for "best players of the season" - are you thinking to rank up to 3-5 players for each season, or just 1 player

Five would be the minium IMO. Perhaps top 10?

and playoffs also or just season?

Either would be fine with me. I'm running an annual Best Player voting on the Main board for both regular season and playoffs and leave it up to the individual voters how much weight they want to give to each segment.

There are some facts that I think could make a "best players of the year" list significantly different from Hart results. For one thing, defensemen get very little consideration for the Hart, and they're certainly often among the best players in the league. I mean, Erik Karlsson once led the league in assists as a defenseman and finished 9th in Hart voting. Secondly, especially in recent years, a team missing the playoffs seems to skew Hart voters away from their players to an extent. How was Connor McDavid 5th in 2018? Third, the Hart isn't supposed to be an award to the best player, but the player most valuable to his team. Therefore, any voters that actually follow the rules might not vote for the "best" players. And, fourth, sometimes, they just plain get it wrong.

Good points.

I think one advantage of the birth year approach (and probably the season approach too?) is that each year is independent of each other. So it might be okay to not participate for a couple of weeks. One will of course miss out on that year, but will be fully ready to participate in the next year.

That's also a very good point.

That being said, just as a devil's advocate... 100 seasons x 2 weeks per season = 200 weeks = we are talking about spending 4 years on these projects.

True. I guess we'd have to break it down into limited segments, like 10 seasons/years at a time (20 weeks = 5 months), then take a break.

 

[Fixxer]

It's never a good idead to make top 100 lists with lots of leftovers who would qualify. Not of a fan of any of these subjecvie lists.
Malkin is a poor man's Jonathan Toews... (wtf)

 

[bobholly39]

Wouldn't the best players by birth years be very easily conducted by a set of polls? Doesn't even have to be anything so formal as a project. Easily identify the top ~20 players of each year - list them all in a poll and have people vote. Can rank top 3-5 players per year quite easily this way.

The other issue is that most years most rankings would be slam dunks with very little discussion to be had. There would be some exceptions of course.

I'm pretty sure I've seen someone do polls like this on the poll section previously. I know @tony d ran a set of polls a couple of years ago comparing rosters made of players born the same year - not necessarily ranking within the year. But I think I remember someone else doing it, but i couldn't find the threads.

Even a year like 1965 which has been cited as a great one - the rankings seem pretty obvious at the top, with not much to discuss:

Lemieux
Roy
Yzerman
Belfour
Lafontaine
Neely
Claude Lemieux

If we really wanted to go down all the way to 20th best each year - you'd likely be comparing players like JJ Daigneault and Sergio Momesso. Doesn't seem super interesting - and that's in a very strong year.

If there's definite interest in doing this project - and even if we want to do it as a series of discussions vs as a set of polls. Someone upthread said "2 weeks per seasons X 100 seasons = 4 years". I feel most seasons would have very little discussion to be had - I'd be more likely to say "4 seasons per week = done in 25 weeks total", and have multiple seasons discussed concurrently.

 

[Theokritos]

Wouldn't the best players by birth years be very easily conducted by a set of polls? Doesn't even have to be anything so formal as a project. Easily identify the top ~20 players of each year - list them all in a poll and have people vote. Can rank top 3-5 players per year quite easily this way.

I think the idea is to go deeper than just top 3-5 in the ranking and to attempt actual comparisons between players from different backgrounds (NHL vs Non-NHL).

 

[VanIslander]

The best single season of all-time:

It would have to be regular season only, not playoffs. If you include playoffs - i think it changes from "let's find the best of the best" to "let's identify the most fortunate timing of all time", which sounds a lot more boring to me. Some players have their best season and playoff happen in same year - but more often it does not, since playoff success is team dependent. Fedorov in 1994 had a regular season people are in love with - but in the playoffs his team got bounced in round 1 by SJ (very embarrassing loss).
If we somehow insisted on wanting to include playoffs - I think it should be "best season + best single playoff run" combo - but without them being in the same year.

1. Just the regular season? Too much reliance on just stats.

2. Mix matching playoffs from OTHER years? Frankenstein.

3. THE BEST SINGLE SEASON - ALL INCLUSIVE (regular season + playoffs + international play: Summit Series, Canada Cups, Olympics, World Cups).
After all, the historical significance of a performance is as important as the dominant skill displayed. Let's not go wild over Cooney Weiland, Bronco Horvath and Roger Crozier.

 

[bobholly39]

1. Just the regular season? Too much reliance on just stats.

2. Mix matching playoffs from OTHER years? Frankenstein.

3. THE BEST SINGLE SEASON - ALL INCLUSIVE (regular season + playoffs + international play: Summit Series, Canada Cups, Olympics, World Cups).
After all, the historical significance of a performance is as important as the dominant skill displayed. Let's not go wild over Cooney Weiland, Bronco Horvath and Roger Crozier.

I agree #2 is frankenstein

#1 - No reason not to rely a lot on stats? Best single season ever is very much about ranking peak seasons, something we've never done before (and quite frankly - that i've not seen done in a good way, ever). You'd probably rely just as much on stats for #3.

#3 - all inclusive. I think that's definitely interesting if the goal is to rank the top 5 or so players per year, as someone suggested. So rank the top 5 all-inclusive seasons in 1999, 1998, etc. But if we're ranking best individual seasons per each player all-time - I think it has to be regular season only.

 

[VanIslander]

... all inclusive. I think that's definitely interesting if the goal is to rank the top 5 or so players per year, as someone suggested. So rank the top 5 all-inclusive seasons...

Top-3 stars of the all-inclusive season! (Regular season & playoffs.)

The NHL loves 3 stars of the game, 3 stars of the week, 3 stars of the month, 3 finalist for trophies...

The magic number is 3...

 

[plusandminus]

The best single season of all-time:

It would have to be regular season only, not playoffs. If you include playoffs - i think it changes from "let's find the best of the best" to "let's identify the most fortunate timing of all time", which sounds a lot more boring to me. Some players have their best season and playoff happen in same year - but more often it does not, since playoff success is team dependent. Fedorov in 1994 had a regular season people are in love with - but in the playoffs his team got bounced in round 1 by SJ (very embarrassing loss).
If we somehow insisted on wanting to include playoffs - I think it should be "best season + best single playoff run" combo - but without them being in the same year. So sticking to Fedorov - you could combine his 1994 season to his playoff run in 97 or 98. This is probably a bit too complex though - single best regular season is preferable to be.

In many ways - there would be 2 layers to this project. First - you have to identify the player's best season. Second - you have to compare it to others. So deciding which of Gretzky's season is best for example is a debate in itself - and then you have to compare it to the best from Orr and Lemieux too.

As to "this topic coming up all the time" and "what would we learn and gain" - I disagree, I don't think we've ever done anything of the sort actually. It's true that at the top we'd rank 1 of Gretzky, Orr, Lemieux Howe seasons, and that always happens. Whose #5? Seasons like Sakic 2001, or Fedorov 1994, or Yzerman 1989, or many other players who normally are nowhere near discussion for "~top 10" would come into play. Have you ever compared Sakic or Yzerman's best season to Beliveau or Hull's best?

Obviously for this to make sense - rank one season per player only. We don't want half the list to be 25 different seasons of Gretzky/Orr/Lemieux/Howe in some fashion at the top. This also makes it easier. Because I can say in a heartbeat that Gretzky's best season is better than Hasek's. But what about his 6th best? or 8th best? I think having more than one season per player would just bog the process down.

Is this NHL only, or would the non-NHL Europeans be included? Especially some Soviets had a habit of winning the domestic championship, the World Championship or Olympics, possibly the European club championship, and a few times the Canada Cup. Usually players like Makarov ended up on the all star teams too, both domestically and internationally.

 

[VanIslander]

ANY player from ANY league should be eligible...

But the quality of competition should matter: the best on best comps are key.

The Summit Series, Canada Cups, Renez'Vous '87, touring Soviet-NHL matchups, 1998+ Olympics (minus Pyeongchang 2018 amateurs), World Cups, ... all should count.

This project should encourage us to research and talk about hockey history that matters now and will be worth a damn in 2050.

 

[TheDevilMadeMe]

ANY player from ANY league should be eligible...

But the quality of competition should matter: the best on best comps are key.

The Summit Series, Canada Cups, Renez'Vous '87, touring Soviet-NHL matchups, 1998+ Olympics (minus Pyeongchang 2018 amateurs), World Cups, ... all should count.

This project should encourage us to research and talk about hockey history that matters now and will be worth a damn in 2050.

Well said.

As for the bolded - While I think this forum should be open for any form of discussion, projects, polls, etc, that members want to do, any project that gets the "stickied" treatment really should be relevant decades from now, even as just a snapshot in time.

 

[bobholly39]

Is this NHL only, or would the non-NHL Europeans be included? Especially some Soviets had a habit of winning the domestic championship, the World Championship or Olympics, possibly the European club championship, and a few times the Canada Cup. Usually players like Makarov ended up on the all star teams too, both domestically and internationally.

If we're ranking best single peak season - it should probably be NHL only. Can probably included non-NHL seasons too - but since they're typically in weaker leagues it'll be hard to assess and compare properly.
I really don't think if we're doing single peak seasons we need to include playoffs. What's the point of such a project? Not everyone had the luxury of having an Olympic tournament during his best regular season, or having his team go far in the playoffs. If we're trying to rank peak seasons, it should be regular season only.

If instead of peak seasons we want to rank the top ~3-5 players of each year - for that project looking at absolutely everything (NHL season, non-NHL, international play + playoffs) over the course of a season makes absolute sense.

 

[TheDevilMadeMe]

If we're ranking best single peak season - it should probably be NHL only. Can probably included non-NHL seasons too - but since they're typically in weaker leagues it'll be hard to assess and compare properly.
I really don't think if we're doing single peak seasons we need to include playoffs. What's the point of such a project? Not everyone had the luxury of having an Olympic tournament during his best regular season, or having his team go far in the playoffs. If we're trying to rank peak seasons, it should be regular season only.

If instead of peak seasons we want to rank the top ~3-5 players of each year - for that project looking at absolutely everything (NHL season, non-NHL, international play + playoffs) over the course of a season makes absolute sense.

Ranking best seasons NHL-only, regular season only, seems so... boring.

 

[Professor What]

If we're ranking best single peak season - it should probably be NHL only. Can probably included non-NHL seasons too - but since they're typically in weaker leagues it'll be hard to assess and compare properly.

I think that even applies to the WHA. We have much more information to use to compare the NHL and WHA, and yet, every attempt I've ever seen or made to accurately compare them always feels like it falls short. Bobby Hull's 1974-75 season is a perfect example. I don't think there's any way he would have scored 77 goals in the NHL that year, but how many should you take off? I strongly suspect he would have been shy of Esposito's 61, but can I say that with no doubt whatsoever? No, I can't. Even if he falls short, how short does he fall? I don't really know where I think Hull's campaign would have fallen into the NHL that season. European leagues would be much harder than that.

 

[plusandminus]

I think that even applies to the WHA. We have much more information to use to compare the NHL and WHA, and yet, every attempt I've ever seen or made to accurately compare them always feels like it falls short. Bobby Hull's 1974-75 season is a perfect example. I don't think there's any way he would have scored 77 goals in the NHL that year, but how many should you take off? I strongly suspect he would have been shy of Esposito's 61, but can I say that with no doubt whatsoever? No, I can't. Even if he falls short, how short does he fall? I don't really know where I think Hull's campaign would have fallen into the NHL that season. European leagues would be much harder than that.

I think WHA accomplishments can be ignored completely. I did a rather deep study some years ago on how to possibly "adjust" WHA points to NHL points, but don't remember much about the results except that WHA was considerably weaker. I think one could easily divide WHA points by 1.5 or so. Anyway, I don't think any WHA player had a season good enough to be top-50 or top-100 (or?).

Soviet league is different. Had much better players.

Håkan Loob of Sweden had one very good Swedish season, scoring 76 pts in 36 games as 22 year old, which was a scoring record that stood for many years. But Loob then likely spent his prime/peak years in the NHL, and we know his NHL peaks are far from making a top-100 list.
Not sure about guys from earlier seasons, but no one probably makes the top-100 or so.

A 40 year old Kent Nilsson scored at a 3.33 pts per game during the regular season, and at a 5.50 pts per game pace during the playoffs. 6 games and 2 games, respectively, in the Spanish league.


By the way, I still would like the birth year approach best.

 

[Professor What]

Soviet league is different. Had much better players.

Right. That's the lone exception, and we know those guys could compete, but it's still a nearly impossible comparison. With the WHA, we had plenty of guys going back and forth, so it's theoretically possible to get some sort of equivalency, but I think we're mostly in agreement on it. Still, the Soviet league had virtually no players moving between it and the NHL, and we generally got to see them against NHL caliber competition once every few years in a Summit Series/Canada Cup. There's no way to quantify that, and I'd never be comfortable with any list that I, or anyone else could make that compared them to NHLers. That's why I'm so loathe to try to give guys like Tretiak and Kharlomov all-time rankings.

 

[Hockey Outsider]

I think WHA accomplishments can be ignored completely. I did a rather deep study some years ago on how to possibly "adjust" WHA points to NHL points, but don't remember much about the results except that WHA was considerably weaker. I think one could easily divide WHA points by 1.5 or so. Anyway, I don't think any WHA player had a season good enough to be top-50 or top-100 (or?).

See this paper from Gabriel Desjardins - League Equivalencies « Hockey Analytics

He estimates that the WHA, in its first season, was only slightly better than the AHL. By the end, it was very nearly NHL calibre (which makes sense - when you look at the top scorers of its final season, and see what they did the NHL the following year, there's only a very small drop-off in per-game production).

The results are somewhat "spiky". The league equivalency rating is nearly identical in 1976 and 1979, but there's a large, unexplained dropoff in 1977 - likely due, at least in part, to small sample sizes.

But to your bigger point - clearly no WHA-only star is worthy of the Hall (or Top 100, etc) strictly based on their performance in that league. But I think it makes sense to consider the performance of established NHL stars (Howe, Hull; to a lesser extent Mark Howe, Gretzky, Keon, etc) when assessing the totality of their careers.

 

[Professor What]

See this paper from Gabriel Desjardins - League Equivalencies « Hockey Analytics

He estimates that the WHA, in its first season, was only slightly better than the AHL. By the end, it was very nearly NHL calibre (which makes sense - when you look at the top scorers of its final season, and see what they did the NHL the following year, there's only a very small drop-off in per-game production).

The results are somewhat "spiky". The league equivalency rating is nearly identical in 1976 and 1979, but there's a large, unexplained dropoff in 1977 - likely due, at least in part, to small sample sizes.

But to your bigger point - clearly no WHA-only star is worthy of the Hall (or Top 100, etc) strictly based on their performance in that league. But I think it makes sense to consider the performance of established NHL stars (Howe, Hull; to a lesser extent Mark Howe, Gretzky, Keon, etc) when assessing the totality of their careers.

I know the one you're talking about. I've used it to analyze data many times. You basically have two years of relative equality, and otherwise, the WHA is meh. Guys like Howe and Hull are the main reason I do deal with it. They had too many years in that league to ignore, so you can't really get the full tale without it, but the apparent inconsistency in the strength of the league is another reason why trying to rank WHA players' seasons against NHL players' seasons is such an uphill battle.

 

[plusandminus]

Another thing I would be interested in is looking at player careers up until a certain age, for example the age of 28 or 29.
That would make comparisons between Orr, Gretzky, Mario, Gordie, Bourque, Coffey and others far more "comparable", since Orr and many other all time greats had their careers shortened by injuries.
I also find that most players have peaked before the age of 28, even though there are exceptions (like Martin St Louis, or possibly/likely Lidström).
Many Europeans would be a bit "punished" for having started their NHL career later than the North Americans, so when looking at total points scored that should be kept in mind.

For me this would be a very interesting study because by the time of 28 or 29, most players have showed their peak and prime. For example, Orr might by age 23-24 already have showed the world that he was the best (and/or most talented) player ever. Same with Gretzky. Roy did it early too. (Hasek later, at least to North Americans.)

Another study could then be done for players after a certain age, like 28 or 29. (Would Gordie be the greatest? How great would Lidström and Bourque seem? Gretzky, leading scorer of the 1990s but an enough good allround game? Jagr. No Orr. Probably no Lafleur, being aged 23-28 during 6 strongest seasons.)


The "project rankings" being done here focuses a lot on the whole career, which differs so much between players. Longevity is great, but so dependent on health, which is partly dependent on luck or unluck (cancer, Suter hit, troubling knees), although playing style (Lidström vs Lindros) might also factor in.


A disadvantage for setting the age limit is that it will affect players differently. One player might have had a serious injury at age 25, or 27, while another had it at 30. A guy like Phil Esposito had his best seasons being 26-31, or 25-32 or so.

A solution might be to be flexible about the age limit. Use say 28 or 29 as a standard. Then give the guys who got injured earlier the benefit of the doubt. Focus on how good the player actually was, rather than if he managed to play until age 28 or not. For cases like Esposito, we might need another solution.

(An alternative may be to choose the best 5-7 years or so from a player's career. But I personally like to split the career in two halves.)

 

[ted2019]

I think that it's best to do projects that we have a chance of completing - a 4 year long project seems like it wouldn't be that...

Ideas that I think could work:

• Best single season teams of all-time
• Best single seasons of all-time by a player
• Best coaches of all-time (though this might be too research intensive if we want results that aren't somewhat arbitrary)
• HOH Hall of Fame or Hall of Excellence (we've talked about this one the most in the past)

I'm digging my 30 all time best rosters. You have to possibly have some rules for the salary cap era. Would we include international teams in this, or just NHL/NHA/PCHA?

 

[ted2019]

I think the idea is to go deeper than just top 3-5 in the ranking and to attempt actual comparisons between players from different backgrounds (NHL vs Non-NHL).

If we do some sort of comparison thread, I would want to do more myself. It would give a true ranking.

 

[ted2019]

Wouldn't the best players by birth years be very easily conducted by a set of polls? Doesn't even have to be anything so formal as a project. Easily identify the top ~20 players of each year - list them all in a poll and have people vote. Can rank top 3-5 players per year quite easily this way.

The other issue is that most years most rankings would be slam dunks with very little discussion to be had. There would be some exceptions of course.

I'm pretty sure I've seen someone do polls like this on the poll section previously. I know @tony d ran a set of polls a couple of years ago comparing rosters made of players born the same year - not necessarily ranking within the year. But I think I remember someone else doing it, but i couldn't find the threads.

Even a year like 1965 which has been cited as a great one - the rankings seem pretty obvious at the top, with not much to discuss:

Lemieux
Roy
Yzerman
Belfour
Lafontaine
Neely
Claude Lemieux

If we really wanted to go down all the way to 20th best each year - you'd likely be comparing players like JJ Daigneault and Sergio Momesso. Doesn't seem super interesting - and that's in a very strong year.

If there's definite interest in doing this project - and even if we want to do it as a series of discussions vs as a set of polls. Someone upthread said "2 weeks per seasons X 100 seasons = 4 years". I feel most seasons would have very little discussion to be had - I'd be more likely to say "4 seasons per week = done in 25 weeks total", and have multiple seasons discussed concurrently.

Bolded part. That is 3/4 of the fun with these sort of things. That's what we do in the HOH section. It's fun to compare players like this as it gives more knowledge to everyone involved and those who might just be reading/following along. It also might make more posters who were possibly scared off from posting here, to stay around here and build this community and isn't that what's it's all about on here, is learning?

 

[ted2019]

ANY player from ANY league should be eligible...

But the quality of competition should matter: the best on best comps are key.

The Summit Series, Canada Cups, Renez'Vous '87, touring Soviet-NHL matchups, 1998+ Olympics (minus Pyeongchang 2018 amateurs), World Cups, ... all should count.

This project should encourage us to research and talk about hockey history that matters now and will be worth a damn in 2050.

Sounds like on how the top 40/60/100 lists should be done.

 

[TheDevilMadeMe]

Another idea - the 101st - 200th best players of all-time, based on the recent top 100 list.

Or if that's too much, do the 100th to 150th best all-time.

I know some posters wanted the list to go beyond 100, but that was deemed too much of a commitment. But now, with a break, maybe we could continue?

 

[ted2019]

Another idea - the 101st - 200th best players of all-time, based on the recent top 100 list.

Or if that's too much, do the 100th to 150th best all-time.

I know some posters wanted the list to go beyond 100, but that was deemed too much of a commitment. But now, with a break, maybe we could continue?

I know that this has been suggested also a few times. I love this idea.