The Hockey News 1998 list of Top-100 NHL Players voting results

[seventieslord]

Brian Costello, current senior editor at THN dug this up for me, and gave me permission to share.

THN had their panel send in top-50 lists, from which they created a top-100 list. No one at THN was privy to the individual lists - they were sent to an accounting firm that tabulated the results and returned a final top-100 list (plus extras).

Knowing how many voters there were, and looking at the point totals, we may be able to reverse engineer what kind of point system they used. The negative point totals for players near the bottom likely means the accounting firm placed value on appearing on the list at all - which I would also do.

Well, take a look and see what you can make of it.

Gretzky	2726
Orr	2713
Howe	2681
Lemieux	2308
Richard	2142
Harvey	2123
Beliveau	2054
Hull	1993
Sawchuk	1899
Shore	1846
Lafleur	1646
Messier	1574
Plante	1543
Bourque	1410
Morenz	1400
Hall	1338
Mikita	1321
Esposito	1299
Potvin	1261
Bossy	1220
Lindsay	1106
Kelly	1056
Clarke	943
Robinson	928
Dryden	898
Mahovlich	889
Schmidt	767
Coffey	767
H.Richard	721
Trottier	684
Moore	618
Lalonde	599
Apps	569
Durnan	560
Roy	501
Conacher	490
Jagr	469
Dionne	461
Malone	451
Chelios	434
Clapper	432
Geoffrion	408
Horton	399
Cook	387
Bucyk	383
Hainsworth	366
Perreault	351
Bentley	344
Park	336
Kurri	332
N. Stewart	299
Clancy	280
Cowley	275
Lindros	246
Jackson	232
Stastny	206
Kennedy	196
Bathgate	177
Pilote	177
Broda	168
Boucher	165
Denneny	163
Parent	153
Hull	150
Joliat	147
Blake	146
Brimsek	136
Lach	134
Keon	116
Fuhr	115
Leetch	97
Seibert	78
Bentley	65
Salming	59
Vezina	44
Gardiner	43
Benedict	29
Yzerman	24
T.Esposito	23
Smith	0
S.Savard	-5
Delvecchio	-7
Dye	-12
Chabot	-20
Abel	-21
Gainey	-31
Bower	-33
Cleghorn	-40
Gartner	-47
Ullman	-55
Schriner	-56
Primeau	-59
Sittler	-59
Sakic	-63
Hasek	-64
Pratt	-66
J.Stewart	-68
Cournoyer	-74
Gadsby	-77
Nighbor	-77
Hawerchuk	-83
McDonald	-85
Middleton	-87
Goodfellow	-89
Lemaire	-90
Ratelle	-90
D.Savard	-90
Hedberg	-92
Pronovost	-92
Cheevers	-94
Coulter	-94
Goulet	-94
Neely	-95
Howell	-96
C.Johnson	-96
Quackenbush	-96
Lapointe	-97
Weiland	-97
Ciccarelli	-98

[TBODY] [/TBODY]

Click to read more...

 

[Iceman]

Thanks for posting Seventies!

Double Shore in the beginning.

On a side note, Lindros is ahead of surprisingly many good players. Roy and Hasek are surprisingly low.

 

[seventieslord]

For starters, there were 50 voters, and the lowest scores approach -100, so it's safe to say that players were given a score of -2 every time they didn't make the list.

This means Ciccarelli probably made one list at 50th, earning zero points, and was (rightly) left off the other 49, earning -98 for those.

It's quite possible Nighbor made #100 on this list from just one astute backer placing him ~25th while every other panelist ignored him.

Gretzky earned 54.5 points per list, meaning there had to be a slight uptick in the value of votes towards the top. (Orr earned 54.3 and Howe 53.6). Knowing that these three were probably top-3 on more ballots than they weren't, that tells us that there were about 165 points given out to the top-3. Perhaps 60, 55, 50? They would have had a total of 8250 points if they were ranked 1-3 100% of the time, and they had 8120, meaning they missed out on 130 possible points, or about 2-3 per ballot. It's possible that on about every 2nd ballot, one of the big 3 was ranked 4th, or on about every 3rd ballot, one was ranked 5th. This would give us results like the above.

There are about 1149 points per ballot. I imagine the number of points per ballot was variable, because as soon as an outlier player is named, in addition to the points he receives from that ballot, he immediately becomes a part of other ballots that don't contain his name, earning -2 each time. (so in essence, every NHL player in history has a score of -100 by this method; they just aren't all displayed here).

 

[seventieslord]

Thanks for posting Seventies!

Double Shore in the beginning.

On a side note, Lindros is ahead of surprisingly many good players. Roy and Hasek are surprisingly low.

Thanks!

After removing the duplicate shore, the number of total points actually becomes a nice round number - 55600, which is 1112 per ballot. Surely there's something we can do with that to reverse engineer the results and scoring system.

 

[Noldo]

Thanks!

After removing the duplicate shore, the number of total points actually becomes a nice round number - 55600, which is 1112 per ballot. Surely there's something we can do with that to reverse engineer the results and scoring system.

But a straight sum of numbers 1 to 50 gives a grand total of 1275, meaning that either low positions were of equal value or the point total is somehow incorrect. Did you include negative scores? If we assume that each player was given a score of -2 for being outside top-50 in a list we should be able to calculate the total amount of negative scores given (assuming the list includes all players who received votes)

 

[seventieslord]

But a straight sum of numbers 1 to 50 gives a grand total of 1275, meaning that either low positions were of equal value or the point total is somehow incorrect. Did you include negative scores? If we assume that each player was given a score of -2 for being outside top-50 in a list we should be able to calculate the total amount of negative scores given (assuming the list includes all players who received votes)

Yes that includes the negatives.

Good point. 119 players were named. 50 lists, 50 names per list. That's 2500 instances of a player being named. If 119 players were on every list there would be 5950 instances, 3450 more than the minimum. So there should be 6900 negative points given.

Did I do that right? If so, that leaves 48700 positive points, 974 per list. But that's only 19 per player.

I can't imagine they'd count a 46th and a 50th the same.

 

[Canadiens1958]

Interesting effort.

Hull seems to be listed twice - 1993 and 150.

Henri Richard top 30. Bucyk surprising. 1990s hockey seems to be under the radar.

 

[seventieslord]

Interesting effort.

Hull seems to be listed twice - 1993 and 150.

Henri Richard top 30. Bucyk surprising. 1990s hockey seems to be under the radar.

Bobby and Brett.

 

[Nick Hansen]

Messier at 12th. Some have disputed his high placing in stuff like this, but everywhere I look, he's always between 10 and 15 (usually 12 or 13) in top players. That has to account for something in and of it self.

Anders Hedberg seems like a weird one to have on this list, though...

 

[JackSlater]

Interesting effort.

Hull seems to be listed twice - 1993 and 150.

Henri Richard top 30. Bucyk surprising. 1990s hockey seems to be under the radar.

... those Hulls would be Bobby and Brett. It's a pretty well known list.

Very good effort to get the scores. Interesting to see little tiers that exist between the players.

 

[seventieslord]

Anders Hedberg seems like a weird one to have on this list, though...

Yes, definitely. Although based on what we can figure out, he likely was ranked about 40th on one list and not at all on any others.

 

[Nick Hansen]

Yes, definitely. Although based on what we can figure out, he likely was ranked about 40th on one list and not at all on any others.

I mean, he was a good player and it might have been sort of a nod to another great hockey country, but even then you have players like Håkan Loob and Kent Nilsson who were better.

 

[JackSlater]

I took a stab at some of the numbers. If the data is complete, then I agree with a lot of seventieslord's conclusions. It seems reasonable that each player received a -2 for not appearing on a given list given the results of the players near the bottom. Using the total amount of points given, each voter's ballot totaled 1112, as already noted. If we remove the 138 negative points that each voter gave (there are 119 players listed, so on each ballot 69 of them received -2) then each voter gave out 1250 positive points. At the very least that is a nice number.

Now we must consider how the points were distributed. If the voters gave out 1 point for 50th place, 2 points for 49th place etc. then the total would be 1275. This number is a bit too high. This system also makes Ciccarelli's score impossible, as the worst possible score (that could still make the list) would be -97. So, I'm going to assume a system where 50th place gets 0 points, 49th place gets 1 point and so on (first place receives a base of 49 points). This system gives a total of 1225 points. This is close to the total of 1250 points but is 25 points short. It makes sense that the top players would get a bit of a bonus, and that bonus (spread among all the top players) would have to equal 25 points per ballot. I'm guessing that the bonus was limited to either the top five or the top three as this allows for the sizable gap found among the top players (Gretzky, Orr, Howe).

Since Gretzky received 2726 points, if we assume that he was included on every ballot then he received 54.52 positive points on average. Since Gretzky very likely split first place votes with other players, the top player on each ballot must receive more than 55 points. Anyway, this is my guess at a scoring system: #1 received 49 points for being first, plus a bonus of 9 points for 58 total points. #2 received 48 points plus a bonus of 7 points for a total of 55 points, #3 received 47 points plus a bonus of 5 points for a total of 52 points, #4 received 46 points plus a bonus of 3 points for a total of 49 points and #5 received 45 points plus a bonus of 1 point for a total of 46 points. The bonus points are now gone (9 bonus points + 7 bonus points + 5 bonus points + 3 bonus points + 1 bonus point = 25 bonus points), so player #6 received 44 points, #7 received 43 points and so on. This system is just a guess, but the numbers add up. The maximum score (a player receiving 50 first place votes) would be 2900 points, which is comfortably above Gretzky's total. The maximum points available for the top three would be 8250, which is higher than the actual total received by the top three (Gretzky, Orr and Howe) which was 8120.

Just messing around with numbers here. It's difficult to suss out what they did, but I am guessing that the system is at least similar to the above.

 

[Canadiens1958]

Bobby and Brett.

Okay, thanks but this also hilites the issue with a lack of 1990s players. No Fedorov,Lidstrom, Oates - not that huge a gap with Brett Hull, plus goalies.

 

[Michael Farkas]

Heh, I liked going through the OP, believe me...but man, I liked reading JackSlater's post ^ even more...bravo, whether it's right or wrong, who cares...that was cool...

 

[JackSlater]

Okay, thanks but this also hilites the issue with a lack of 1990s players. No Fedorov,Lidstrom, Oates - not that huge a gap with Brett Hull, plus goalies.

The list was made in 1997. There would be no reason to include Lidstrom in 1997 and players like Fedorov, Oates and Hull had time to add to their careers (to varying degrees) since that point.

 

[Michael Farkas]

Okay, thanks but this also hilites the issue with a lack of 1990s players. No Fedorov,Lidstrom, Oates - not that huge a gap with Brett Hull, plus goalies.

Not the be-all, end-all...but in 1998, Lidstrom hadn't even won a Norris yet...he was just getting his name solidified on the map after the 1997 Final. I'm not sure when exactly this was all voted on, but Lidstrom had about as much of a resume then as like Oliver Ekman-Larsson has now...he wouldn't make any historical ranking, ya know?

 

[Nick Hansen]

Who is this Primeau fella being ranked ahead of Sakic, Hawerchuk and Hasek?

It can't be Keith Primeau can it...

 

[Michael Farkas]

Wayne...super 4th liner!




(Joe Primeau, I presume: Joe Primeau Stats | Hockey-Reference.com )

 

[Canadiens1958]

Not the be-all, end-all...but in 1998, Lidstrom hadn't even won a Norris yet...he was just getting his name solidified on the map after the 1997 Final. I'm not sure when exactly this was all voted on, but Lidstrom had about as much of a resume then as like Oliver Ekman-Larsson has now...he wouldn't make any historical ranking, ya know?

True but what had Lindros done to deserve a 246 total ahead of past multiple Ross, Hart and Norris winners? Fedorov by that time had a Hart, two Selke's and a Lindsay.

Fedorov and Lidstrotm are virtually ignored.

 

[Hockey Outsider]

Who is this Primeau fella being ranked ahead of Sakic, Hawerchuk and Hasek?

It can't be Keith Primeau can it...

It was "Gentleman" Joe Primeau. He was one of the premier playmakers of his generation (1930s). He centred Charlie Conacher on the Leafs. He led the league in assists three times in four years. When he retired in 1936, he ranked 4th all-time in assists, and held the single-season record.

Primeau was a perennial Lady Byng contender back when the award was taken seriously. He finished runner-up in scoring twice. When the Leafs won the Cup in 1932 he recorded six assists in seven games, which was great for that era.

A decent modern comparable might be Adam Oates, but with way less longevity. Still, I wouldn't have room for him in my top 100.

 

[Nick Hansen]

Wayne...super 4th liner!




(Joe Primeau, I presume: Joe Primeau Stats | Hockey-Reference.com )

Phew. That makes sense.

Amongst other players that are missing, Ron Francis surprises me. Doug Gilmour and Pat Lafontaine can also be mentioned.

 

[Michael Farkas]

I think we're identifying how difficult it is to digest careers without chewing. Lindros was probably the best player in the game or thereabouts at the time this vote was taking place. Slightly more recently than Fedorov's incredible 1994. Before Lidstrom ramped it up. So Lindros was a beanie baby that everyone wanted for a few years but soon after realized...how easily the...heads...pop...off...? I'm not sure where I was going with that. But the moral of the story is he got bonus points for being in his peak, peak at the time of voting...

 

[Canadiens1958]

I think we're identifying how difficult it is to digest careers without chewing. Lindros was probably the best player in the game or thereabouts at the time this vote was taking place. Slightly more recently than Fedorov's incredible 1994. Before Lidstrom ramped it up. So Lindros was a beanie baby that everyone wanted for a few years but soon after realized...how easily the...heads...pop...off...? I'm not sure where I was going with that. But the moral of the story is he got bonus points for being in his peak, peak at the time of voting...

Love the analogy. Why such lists should feature retired players only.

 

[JackSlater]

True but what had Lindros done to deserve a 246 total ahead of past multiple Ross, Hart and Norris winners? Fedorov by that time had a Hart, two Selke's and a Lindsay.

Fedorov and Lidstrotm are virtually ignored.

You can certainly argue that Fedorov is underrated by the list (and Lindros overrated) but Lidstrom would have no place on the list. The list is from 1997. Lidstrom first became an all star level player in the 1995-1996 season.