Here are my numbers and analysis for the players who played since the 1967 expansion. They rest on the idea that there are three separate phases of the game of hockey - even-strength play, power play, and penalty killing - and these should be evaluated separately.
These don't include playoffs or intangibles and can't be taken as the final word, but can hopefully provide some information in certain areas.
Stat Glossary:
$ESGF/G - even-strength goals for per game, adjusted for scoring level. Higher is better.
$ESGA/G - even-strength goals against per game, adjusted for scoring level. Lower is better.
R-ON - Player's even-strength on-ice goal ratio (ESGF/ESGA). Should be higher than R-OFF.
R-OFF - Player's even-strength off-ice goal ratio (ESGF/ESGA).
XEV+/- - Players expected EV+/-, based on off-ice results.
EV+/- - Even-strength plus-minus, adjusted for scoring level.
AEV+/- - Adjusted even-strength plus-minus. =(EV+/-) - (XEV+/-). If you look at just one number, make it this one.
/82 - Adjusted even-strength plus-minus per season.
SH% - . Percentage of team's PPGA the player was on the ice for. Measures a players role in killing penalties, but not effectiveness.
PP% - Percentage of team's PPGF the player was on the ice for. Measures a players role on the power play, but not effectiveness.
$PPP - Scoring-adjusted power play points per game.
$ESP - Scoring-adjusted even-strength points per game.
Borje Salming
Player | Year | Seasons | $ESGF/G | $ESGA/G | R-ON | R-OFF | XEV+/- | EV+/- |
AEV+/-
|
/82
|
SH%
| PP% |
$PPP/G
| $ESP/G
Borje Salming | 74-75 | 1.72 | 1.15 | 0.85 | 1.35 | 0.97 | -3 | 42 |
45
|
26
|
57%
| 46% |
0.18
| 0.38
Borje Salming | 76-82 | 6.59 | 1.31 | 1.03 | 1.27 | 0.83 | -81 | 152 |
233
|
35
|
58%
| 81% |
0.38
| 0.50
Borje Salming | 83-86 | 3.14 | 0.93 | 1.18 | 0.79 | 0.74 | -57 | -65 |
-7
|
-2
|
57%
| 58% |
0.23
| 0.33
Borje Salming | 87-90 | 2.93 | 1.10 | 0.93 | 1.19 | 0.79 | -41 | 42 |
83
|
28
|
39%
| 25% |
0.06
| 0.23
Borje Salming | 74-90 | 14.37 | 1.17 | 1.02 | 1.14 | 0.82 | -181 | 172 |
353
|
25
|
54%
| 62% |
0.26
| 0.40
Salming was an even strength difference maker from the time he came to the NHL. He went on to have a really excellent seven-year peak from 1975-76 to 1981-82. During this time the Leafs went from being very bad to very good when he was on the ice at even strength, and he was also a productive power play quarterback and their best penalty killer.
His offensive contributions and overall impact dropped off later in his career, but he was an important player for Toronto right to the end.
Dave Keon
Player | Year | Seasons | $ESGF/G | $ESGA/G | R-ON | R-OFF | XEV+/- | EV+/- |
AEV+/-
|
/82
|
SH%
| PP% |
$PPP/G
| $ESP/G
Dave Keon | 68-71 | 3.81 | 0.85 | 0.73 | 1.17 | 1.05 | 9 | 38 |
29
|
8
|
37%
| 50% |
0.25
| 0.64
Dave Keon | 72-75 | 3.82 | 0.70 | 0.61 | 1.15 | 0.99 | -2 | 29 |
31
|
8
|
38%
| 57% |
0.28
| 0.52
Dave Keon | 80-82 | 2.93 | 0.53 | 0.82 | 0.64 | 0.95 | -5 | -70 |
-65
|
-22
|
31%
| 23% |
0.07
| 0.41
Dave Keon | 68-82 | 10.56 | 0.71 | 0.71 | 0.99 | 1.00 | 1 | -3 |
-5
|
0
|
36%
| 47% |
0.21
| 0.53
Keon's numbers for this time period are good for the first few years post-expansion, but not terribly impressive. He was a contributor on the power play, but not a major one. He was a positive force at even strength, but not to the degree that his reputation suggests. Sergei Fedorov, to name another defensively strong centre who isn't up for voting yet, had much better even strength results. He did contribute to a very good Toronto penalty kill while scoring a ton of shorthanded goals, and that's probably his best selling point according to the numbers.
He was overmatched in his final NHL stint with Hartford, so his career averages dropped misleadingly.
That said, these numbers don't cover most of his prime years, so that limits the conclusions we can draw from them.
Scott Stevens
Player | Year | Seasons | $ESGF/G | $ESGA/G | R-ON | R-OFF | XEV+/- | EV+/- |
AEV+/-
|
/82
|
SH%
| PP% |
$PPP/G
| $ESP/G
Scott Stevens | 83-84 | 1.94 | 1.01 | 0.77 | 1.31 | 1.09 | 9 | 38 |
30
|
15
|
25%
| 22% |
0.10
| 0.29
Scott Stevens | 85-90 | 5.58 | 1.11 | 0.96 | 1.15 | 1.15 | 45 | 66 |
21
|
4
|
47%
| 64% |
0.31
| 0.41
Scott Stevens | 91-95 | 4.78 | 1.22 | 0.88 | 1.40 | 1.24 | 61 | 136 |
75
|
16
|
54%
| 54% |
0.23
| 0.48
Scott Stevens | 96-01 | 5.79 | 1.11 | 0.77 | 1.45 | 1.20 | 56 | 163 |
107
|
18
|
71%
| 20% |
0.07
| 0.34
Scott Stevens | 02-04 | 2.45 | 1.09 | 0.89 | 1.22 | 1.25 | 31 | 39 |
8
|
3
|
71%
| 18% |
0.02
| 0.28
Scott Stevens | 83-04 | 20.53 | 1.13 | 0.86 | 1.31 | 1.19 | 202 | 443 |
241
|
12
|
56%
| 40% |
0.17
| 0.38
Scott Stevens famously made the transition from being an offensive defenceman to a defensive defenceman, and this shows up in the stats. Earlier in his career he was not really an even strength difference maker, although he was productive on the power play and played a major role on the penalty kill. In the 90s he became a better even strength defencemen, cutting down on his goals against.
Later in his career his power play role declined, but he took on an even larger role on the penalty kill. In fact, there isn't a single player for whom we have data (post-1967) who was on the ice for as large a share of his team's penalty kill as Scott Stevens was. Some of that may be because the Devils were famously disciplined and spent less time on the PK, but he still deserves a ton of credit for the Devils' fine penalty kill.
The Devil Made Me said that Scott Stevens was more of an even strength difference maker than Al MacInnis. I can't agree with that. During their primes, MacInnis had a better on-ice goal ratio than Stevens, and less support from his teammates (particularly in goal). Stevens likely played a slightly more defensive role than MacInnis, making it more difficult to put up a good plus-minus, but they were both big minute #1 defencemen, so the difference there can't be too big. Add in MacInnis's massive advantage on the power play, and I don't think Stevens' edge as a penalty killer can make up the difference.
Al MacInnis
Player | Year | Seasons | $ESGF/G | $ESGA/G | R-ON | R-OFF | XEV+/- | EV+/- |
AEV+/-
|
/82
|
SH%
| PP% |
$PPP/G
| $ESP/G
Al MacInnis | 82-85 | 1.68 | 0.80 | 0.61 | 1.31 | 1.15 | 10 | 26 |
17
|
10
|
3%
| 90% |
0.50
| 0.28
Al MacInnis | 86-90 | 4.93 | 1.08 | 0.74 | 1.46 | 1.30 | 66 | 137 |
71
|
14
|
39%
| 82% |
0.47
| 0.39
Al MacInnis | 91-96 | 5.03 | 1.12 | 0.72 | 1.54 | 1.02 | 6 | 162 |
155
|
31
|
42%
| 88% |
0.56
| 0.44
Al MacInnis | 97-03 | 6.05 | 1.19 | 0.90 | 1.32 | 1.07 | 23 | 142 |
119
|
20
|
46%
| 87% |
0.47
| 0.40
Al MacInnis | 82-03 | 17.68 | 1.10 | 0.78 | 1.41 | 1.12 | 106 | 468 |
362
|
20
|
39%
| 86% |
0.50
| 0.40
I'm surprised MacInnis wasn't voted in last time. He was far more than just his shot. He was a very good even-strength defenceman and an all-time great power play quarterback for a long time. He didn't get as many all-star spots as he could have in the 90s because he missed 10-20 games in a few years, but he was still providing a ton of value in those years. Few defencemen were elite for as long as he was.
Brett Hull
Player | Year | Seasons | $ESGF/G | $ESGA/G | R-ON | R-OFF | XEV+/- | EV+/- |
AEV+/-
|
/82
|
SH%
| PP% |
$PPP/G
| $ESP/G
Brett Hull | 87-89 | 1.85 | 0.78 | 0.72 | 1.08 | 1.14 | 11 | 9 |
-1
|
-1
|
0%
| 61% |
0.32
| 0.60
Brett Hull | 90-92 | 2.89 | 1.22 | 1.01 | 1.21 | 1.05 | 8 | 51 |
42
|
15
|
3%
| 86% |
0.51
| 0.94
Brett Hull | 93-98 | 5.51 | 0.95 | 0.94 | 1.00 | 1.07 | 19 | 2 |
-17
|
-3
|
29%
| 78% |
0.46
| 0.73
Brett Hull | 99-06 | 5.71 | 0.90 | 0.77 | 1.16 | 1.26 | 64 | 58 |
-5
|
-1
|
14%
| 59% |
0.40
| 0.69
Brett Hull | 87-06 | 15.96 | 0.96 | 0.87 | 1.11 | 1.13 | 102 | 120 |
18
|
1
|
16%
| 72% |
0.43
| 0.76
Hull had a great three year peak. Outside of that, he was basically not a difference-maker at even strength - his goals just made up for his shortcomings in other areas. He was a very good power play scorer for a very long time, that's probably the best thing he has going for him.
Peter Stastny
Player | Year | Seasons | $ESGF/G | $ESGA/G | R-ON | R-OFF | XEV+/- | EV+/- |
AEV+/-
|
/82
|
SH%
| PP% |
$PPP/G
| $ESP/G
Peter Stastny | 81-88 | 7.54 | 1.07 | 0.91 | 1.18 | 1.02 | 8 | 101 |
93
|
12
|
9%
| 73% |
0.47
| 0.83
Peter Stastny | 89-92 | 3.61 | 0.77 | 0.89 | 0.87 | 0.84 | -31 | -34 |
-3
|
-1
|
7%
| 69% |
0.37
| 0.53
Peter Stastny | 93-95 | 1.07 | 0.56 | 0.51 | 1.11 | 1.10 | 3 | 5 |
2
|
2
|
5%
| 55% |
0.25
| 0.39
Peter Stastny | 81-95 | 12.22 | 0.94 | 0.87 | 1.08 | 0.97 | -20 | 71 |
91
|
7
|
8%
| 71% |
0.42
| 0.70
Stastny was an excellent offensive player in North America for about 8 years - and he didn't get started until he was 24. On the other hand, his goals against are poor, and the Nordiques were an average team with him off the ice, so this isn't a situation where a terrible defense was dragging him down. He didn't kill penalties either, making him a fairly one-dimensional player.
Overall, he was a positive force at even strength, if not dominant, while also a very good power play scorer.
Others
Nighbor is near the top of my list. I hope seventieslord can post those early 70s save percentages soon, as I'm interested to see Bernie Parent's numbers. I believe his save percentages were actually very good for his first stint in Philadelphia.