Great analysis as always. The biggest surprises were:
- Howe played the least amount of time on the penalty kill (in terms of his prime). I would have assumed that Howe played more on the PK than Niedermayer and Leetch (though in the latter's case, he likely earned a lot of PK ice time in order to generated shorthanded offense, and also because the Rangers generally lacked depth).
Howe's special teams usage is a bit lower than one would expect, based on his reputation. I guess he was just more of a skating/transition/even-strength player.
Howe was listed at 5'11", 185 lbs. Every great penalty killing defenceman (post-expansion) has been bigger. Most have been 6'2 or taller. Among the shorter ones were Ray Bourque (5'11, 220 lbs) and Chris Chelios (6'0, 191 lbs).
Howe was also a much better offensive player at even strength than on the power play. His skating would have been more of an advantage at even strength. He used his great wrist shot more than a slap shot, but maybe this wasn't as good a fit for the power play?
Here are the year-by-year special teams numbers for Howe, as well as his rank among defenceman on his team and who he was behind, if anyone. Keeping in mind that single season numbers are imprecise...
Year | Player | PP% | Rk | Behind
1980 | Mark Howe | 82% | 1 |
1981 | Mark Howe | 80% | 1 |
1982 | Mark Howe | 63% | 1 |
1983 | Mark Howe | 70% | 1 |
1984 | Mark Howe | 49% | 1 |
1985 | Mark Howe | 47% | 3 | Eriksson
1986 | Mark Howe | 53% | 2 | Crossman
1987 | Mark Howe | 72% | 1 |
1988 | Mark Howe | 60% | 1 |
1989 | Mark Howe | 69% | 1 |
1990 | Mark Howe | 49% | 3 | Murphy, Huffman
1991 | Mark Howe | 56% | 2 | Murphy
1992 | Mark Howe | 56% | 2 | Duchesne
1993 | Mark Howe | 36% | 4 | Coffey, Chiasson, etc
1994 | Mark Howe | 20% | 4 | Coffey, Lidstrom, etc
1995 | Mark Howe | 31% | 4 | Coffey, Lidstrom, etc
Year | Player | SH% | Rk | Behind
1980 | Mark Howe | 49% | 3 | Sims, Ley
1981 | Mark Howe | 46% | 3 | Sims, Barnes
1982 | Mark Howe | 34% | 3 | Wesley, Kotsopoulous
1983 | Mark Howe | 29% | 4 | Marsh, Dvorak, McCrimmon*
1984 | Mark Howe | 35% | 4 | McCrimmon, Marsh, Dvorak
1985 | Mark Howe | 53% | 1 |
1986 | Mark Howe | 53% | 1 |
1987 | Mark Howe | 47% | 2 | Crossman
1988 | Mark Howe | 49% | 2 | Crossman
1989 | Mark Howe | 52% | 2 | Samuelsson
1990 | Mark Howe | 47% | 3 | Samuelsson, Carkner
1991 | Mark Howe | 70% | 3 | Samuelsson, Carkner
1992 | Mark Howe | 33% | 5 | Samuelsson, Carkner, etc
1993 | Mark Howe | 27% | 5 | Chiasson, Konstantinov, etc
1994 | Mark Howe | 18% | 6 | Chiasson, Konstantinov, etc
1995 | Mark Howe | 19% | 6 | Rouse, Coffey, etc
*Howe played forward on the PK this season.
Edit: Another factor, at least during the Mike Keenan years of 84/85 to 87/88, might have been the fact that Keenan preferred to go with four defencemen as much as possible in Philadelphia. So he wouldn't want to play one defenceman on special teams too much, to avoid messing up the rotation.
Here's an indication of what Keenan thought about Mark Howe. I ran across it while reading up on Pronger.
Iain MacIntyre, Vancouver Sun, Dec 11, 1997:
Keenan has always believed a dominant, Norris Trophy-calibre defenceman is essential. Keenan traded for Chris Chelios when he was in Chicago and for Chris Pronger when he coached and managed in St. Louis. He had Brian Leetch when New York won the Stanley Cup in 1994.
Keenan said the only exception to his rule may have been in Philadelphia, where he surrounded superior, but not outstanding, defenceman Mark Howe with a strong supporting cast.
So, as long as the reporter isn't putting words in Keenan's mouth, this suggests that Keenan thought Chelios, Leetch, and Pronger were on another level as compared to Howe.
- I was surprised to see that Lapointe was used more on the penalty kill than Savard (who generally has the reputation of being better defensively). What do others think?
That's in part because of the years I chose to show.
Lapointe played on Montreal's first PK unit from 1973-74 to 1978/79. During those years, Savard was also on the top penalty kill unit. Savard's SH% was 70%, as compared to 72% for Lapointe. I don't think that's a significant difference.
Savard played a larger PK role outside of those years, which is a tiebreaker in his favour. After JC Tremblay left, Savard was the one who replaced him on the first PK unit. Lapointe joined him a year later, replacing Laperriere. And when Robinson was promoted to the first unit, he bumped Lapointe to the second unit. Savard remained on the first unit with Robinson.
Seventieslord probably will be annoyed now, but I just want to mention again that I think the above mentioned stats should be taken with a grain of salt. For example, you mention how Washington's PK stats changed dramatically when Langway wasn't on the team.
There is a risk that players that are good at preventing GA, ends up with too low SH time estimations (and vice versa).
In cases where a player have missed games during the season in question, I think you make a projection (at least I think you did before) that may also be risky. I think for example Bobby Orr ended up being on ice for more than 100 % of the PP GF, which of course is impossible. That probably happened because when Orr didn't play, his team scored fewer PP goals per game. Better would be to only look at the games Orr actually played, but I understand that may be difficult. You may have spent time looking at newer seasons, where one can compare the estimations with real times/percentages, and come to the conclusion that even for the best players, the estimations are considered good enough.
(Interpret all the above as an interest in the numbers you present.)
Yes, that's true. Especially for players who miss some games.
I think this is less of a factor for the SH numbers than the PP numbers. The best penalty killers usually end up matched up against the opponent's first unit power play, and they defend the 5-on-3 situations. So they play in situation where they have a higher chance of getting scored against. This is why I show the team numbers. A great defensive defenceman may end up with a lot of PPGA because of the situation he plays in, but if his team numbers are good, he's probably doing something right.
Scott Stevens is a good example of this. I looked at his actual TOI numbers, and his team allowed PP goals at a higher rate when he was on the ice than when he was off the ice. But that's almost certainly because he matched up against the opponents' first power play units every time, and defended against the 5-on-3's.
(You gave the example of Bobby Orr. A more relevant example to this round is Brian Leetch in 1992-93. He was on the ice for 40 PPGF in 36 games. In the 48 games he missed, the Rangers only scored 37 PP goals.)