Thursday, February 21, 2008

Defining bad defence

In general, I'm pretty happy with the Rod Langway look at the best defensive defenders. It's a quick and dirty method for finding which players have played big minutes on defence in a shutdown role and kept pucks out of their own net.

What it's not useful for is finding which defenders have been on the wrong end of the equation.

If finding "good" defence is complicated, then the question of who's played badly is even more so. We all already know Nick Lidstrom and Chris Pronger make it difficult for the opposition to score, but if I asked you who made it the easiest, my guess is there'd be any number of different answers.

What follows are a few different looks at "bad defence" in the NHL. (With some serious help, as per usual, from Behind The Net.)

On the ice for the most 5-on-5 goals against

NAME POS TM GP QUAL COMP GFON GAON
1 PAUL RANGER D T.B 58 0.12 61 70
2 ILYA KOVALCHUK RW ATL 58 0.05 44 60
3 LUBOMIR VISNOVSKY D L.A 62 -0.05 43 60
4 MARTIN ST. LOUIS RW T.B 59 0.11 63 60
5 BRAD RICHARDS C T.B 59 0.06 35 58
6 VINCENT LECAVALIER C T.B 59 0.14 62 58
7 FILIP KUBA D T.B 58 0.13 51 58
8 VACLAV PROSPAL LW T.B 59 0.13 52 54
9 BRIAN CAMPBELL D BUF 59 0.00 53 54
10 STEVE STAIOS D EDM 61 0.07 36 54
11 WADE REDDEN D OTT 60 0.01 64 53
12 TODD WHITE C ATL 61 0.06 38 51
13 ANZE KOPITAR C L.A 62 0.03 35 51
14 ERIC STAAL C CAR 62 0.05 49 51
15 MIKE GREEN D WSH 60 0.04 49 51
16 JACK JOHNSON D L.A 62 0.05 39 51
17 TOMAS KABERLE D TOR 61 0.05 44 51
18 NICLAS HAVELID D ATL 60 0.10 50 50
19 BRAD STUART D L.A 61 -0.08 34 49
20 ALEXANDER OVECHKIN LW WSH 60 0.06 59 49

That there is as basic as it gets. Bad defence, in essence, as being out there for a lot of goals against.

This, of course, doesn't take into account ice time, so we see some big-time minute munchers on bad teams way up the list. Tampa's Big Three are taking up prime real estate, and young Mr. Ranger is running away with the lead.

No trophy for that one.

Let's bring ice time in here:

5-on-5 goals against per 60 minutes

NAME POS TM GP TOI/60 QUAL COMP GF/60 GA/60
1 MICHAEL NYLANDER C WSH 40 13.38 0.02 2.47 4.37
2 ILYA KOVALCHUK RW ATL 58 14.67 0.05 3.10 4.23
3 RYAN SMYTH LW COL 39 15.16 0.03 2.94 3.86
4 TODD WHITE C ATL 61 13.16 0.06 2.84 3.81
5 ALEXANDER SEMIN LW WSH 41 11.93 -0.02 2.94 3.80
6 JOE SAKIC C COL 24 13.63 0.06 2.57 3.67
7 LUBOMIR VISNOVSKY D L.A 62 15.95 -0.05 2.61 3.64
8 BRAD RICHARDS C T.B 59 16.24 0.06 2.19 3.63
9 PAUL RANGER D T.B 58 19.99 0.12 3.16 3.62
10 SIMON GAGNE LW PHI 25 11.92 -0.03 2.42 3.62
11 PATRICK O'SULLIVAN RW L.A 62 13.01 0.03 2.60 3.57
12 VINCENT LECAVALIER C T.B 59 16.61 0.14 3.80 3.55
13 MICHAL HANDZUS C L.A 62 11.88 0.08 1.87 3.50
14 DANIEL BRIERE C PHI 60 13.37 -0.02 2.02 3.44
15 ANZE KOPITAR C L.A 62 14.40 0.03 2.35 3.43
16 ROBERT BLAKE D L.A 53 14.53 0.03 2.34 3.43
17 MARTIN ST. LOUIS RW T.B 59 17.86 0.11 3.59 3.42
18 FILIP KUBA D T.B 58 17.70 0.13 2.98 3.39
19 VACLAV PROSPAL LW T.B 59 16.38 0.13 3.23 3.35
20 JIRI TLUSTY LW TOR 37 10.19 0.01 1.59 3.34

More big, ugly numbers, and more Lightning in the parade.

There are some surprising members here, including Burnaby Joe, Todd White, Vinny Lecavalier and Michal Handzus, but mostly what we're still seeing here is players from teams that have had either poor goaltending or poor team defence.

There's also the matter of strength of competition, which is why the Tampa fellows are all getting hammered.

But I'll add that in next. First up, let's compare players to their own teammates' performance:

Difference between 5-on-5 goals against on and off ice

NAME POS TM GP QUAL COMP GA/60 GAOFF/60 GAA dif
1 MICHAEL NYLANDER C WSH 40 0.02 4.37 2.05 2.32
2 SIMON GAGNE LW PHI 25 -0.03 3.62 1.48 2.14
3 ILYA KOVALCHUK RW ATL 58 0.05 4.23 2.27 1.96
4 RYAN SMYTH LW COL 39 0.03 3.86 2.22 1.64
5 DANIEL BRIERE C PHI 60 -0.02 3.44 1.98 1.46
6 TODD WHITE C ATL 61 0.06 3.81 2.38 1.43
7 SEAN AVERY C NYR 36 0.06 2.66 1.26 1.40
8 KEITH YANDLE D PHX 26 -0.10 3.22 1.86 1.36
9 JAROSLAV SPACEK D BUF 47 0.01 3.20 1.85 1.35
10 SCOTT NIEDERMAYER D ANA 29 0.11 2.42 1.12 1.30
11 ALEXANDER SEMIN LW WSH 41 -0.02 3.80 2.53 1.27
12 JOE SAKIC C COL 24 0.06 3.67 2.41 1.26
13 SANDIS OZOLINSH D S.J 38 -0.10 2.97 1.73 1.24
14 BRIAN CAMPBELL D BUF 59 0.00 3.19 1.95 1.24
15 MIKE COMRIE C NYI 59 0.04 3.28 2.06 1.22
16 LUKAS KRAJICEK D VAN 39 -0.12 2.86 1.69 1.17
17 WADE REDDEN D OTT 60 0.01 3.27 2.12 1.15
18 NICLAS WALLIN D CAR 46 -0.03 3.22 2.08 1.14
19 PHIL KESSEL LW BOS 59 -0.00 2.89 1.75 1.14
20 PAUL RANGER D T.B 58 0.12 3.62 2.55 1.07

Some more new names at the party, and a few guys — like Yandle, Ozolinsh and Krajicek — who are getting beat badly during easy minutes.

Then again, we see someone like Scott Niedermayer show up, mainly due to the fact he's carrying the mail when it comes to tough minutes (he may also be a tad rusty from, you know, sitting on his tuchis for a while).

Sakic? A 38-year-old offensive, smallish centre who was playing hurt and filling tough minutes on a team with some goaltending woes? And his linemates were Andrew Brunette and Ryan Smyth/Wojtek Wolski?

Okay, let's factor the quality of competition numbers in here, weight it against the GAA dif numbers, and see what's left:

NAME POS TM QUAL COMP GAA dif GAA dif rank QUAL COMP rank
1 LUKAS KRAJICEK D VAN -0.12 1.17 16 12
2 KEITH YANDLE D PHX -0.10 1.36 8 23
3 SANDIS OZOLINSH D S.J -0.10 1.24 13 23
4 JOHN-MICHAEL LILES D COL -0.12 0.75 51 12
5 RHETT WARRENER D CGY -0.08 0.82 42 35
6 SEAN HILL D MIN -0.08 0.81 44 35
7 AARON MILLER D VAN -0.08 0.70 66 35
8 LUBOMIR VISNOVSKY D L.A -0.05 0.94 30 79
9 BRENDAN MORRISON C VAN -0.05 0.92 32 79
10 ANDREI ZYUZIN D CHI -0.08 0.62 80 35
11 DUVIE WESTCOTT D CBJ -0.16 0.48 118 3
12 SIMON GAGNE LW PHI -0.03 2.14 2 122
13 MILAN LUCIC LW BOS -0.07 0.64 75 50
14 NICLAS WALLIN D CAR -0.03 1.14 18 122
15 MARKUS NASLUND LW VAN -0.05 0.72 61 79
16 FRED MEYER IV D NYI -0.07 0.58 90 50
17 BRYAN BERARD D NYI -0.07 0.57 94 50
18 ANDERS ERIKSSON D CGY -0.03 1.04 23 122
19 CAM BARKER D CHI -0.04 0.81 44 102
20 ANDREJ SEKERA D BUF -0.06 0.58 90 66
21 DANIEL BRIERE C PHI -0.02 1.46 5 152
22 VITALY VISHNEVSKI D N.J -0.04 0.72 58 102
23 CHRIS KUNITZ LW ANA -0.03 0.85 39 122
24 MATT JONES D PHX -0.05 0.60 83 79
25 ALEXANDER SEMIN LW WSH -0.02 1.27 11 152
26 SCOTT WALKER RW CAR -0.04 0.70 62 102
27 MAXIM AFINOGENOV RW BUF -0.04 0.70 62 102
28 STEVE WAGNER D STL -0.04 0.68 70 102
29 MIKE CAMMALLERI C L.A -0.06 0.47 121 66
30 JEREMY ROENICK C S.J -0.03 0.67 71 122
31 PETER SCHAEFER LW BOS -0.07 0.38 143 50
32 BRAD STUART D L.A -0.08 0.31 159 35
33 MARC-ANDRE BERGERON D NYI -0.06 0.42 129 66
34 PATRICK MARLEAU C S.J -0.01 1.06 22 178
35 STEVE DOWNIE RW PHI -0.06 0.39 139 66
36 SHEA WEBER D NSH -0.02 0.73 57 152
37 TODD BERTUZZI LW ANA -0.03 0.59 87 122
38 MARTIN SKOULA D MIN -0.02 0.72 58 152
39 TOM PREISSING D L.A -0.05 0.41 132 79
40 BRIAN CAMPBELL D BUF 0.00 1.24 14 200

Essentially, this is a list of players whose coaches keep them away from tough minutes, but who still manage to somehow have more goals scored against than when they're on the bench.

It relies heavily (probably too heavily) on the quality of competition numbers, giving a free pass to anyone who plays remotely difficult minutes, but maybe that's the point. One would assume that a coach knows his team's weaknesses as well as anyone, and anyone linematching would therefore hide those players.

I'm surprised to see a few players on here (Weber, for one), but there are a lot of offence-first type defenders who one would expect to see on there. And quality of teammates could still be an issue.

Still, I'd take this over plus-minus.

And I'd be curious to see what others can come up with.
.

Labels:

11 Comments:

At 4:38 AM, February 21, 2008, Anonymous David Johnson said...

I don't have them sorted by worst defensive players (though I probably should do that) but I have player rankings at stats.hockeyanalysis.com. Click on the division links under Player Rankings for offense, defense and overall rating for every player in the NHL that has play a significant number of minutes. Most of the players on your list do poorly on my rating scheme but a few exceptions exist. For mine a rating over 1.0 is better than average and a rating under 1.0 is below average. Kunitz is on your list but received a 1.01 rating on my list.

The formula I use is based on the quality of the players linemates and the individual players he plays against while he is on the ice. If a teammate performs better when the player is on the ice that is a 'plus' for the player. Conversely if teammates perform worse when they are on the ice with the player then that is a 'minus' for the player. A similar process is done for the individual opponents he plays against.

The results end up being fairly close to yours but factors in one more level of detail.

 
At 5:06 AM, February 21, 2008, Anonymous Anonymous said...

I just gotta ask: What are "easy minutes"? Are they 60 seconds long like other minutes?
Isn't the point of hockey to score more than you allow?
Then couldn't we just stop at the ratio of GF/60 v GA/60?
That takes into account GF v GA and factors in icetime. All easily countable.
The Quality of Competition thing has become a bit of a meme.

 
At 8:40 AM, February 21, 2008, Blogger Adam C said...

That's a lot of work to determine that Sandis Ozolinsh is poor defensively...

Kidding aside, nice analysis, James. It's great to see it broken down like this, instead of just a final list, so we can get an idea of how each factor affects players' results.

 
At 12:17 PM, February 21, 2008, Blogger MetroGnome said...

I just gotta ask: What are "easy minutes"? Are they 60 seconds long like other minutes?

There's a difference between playing against Jarome Iginla or Sidney Crosby all night rather than, say, David Moss and Maxime Tablot. Lidstrom does the former and Christoph Shubert does the latter. Shubert, therefore, plays "easier" minutes.

 
At 12:31 PM, February 21, 2008, Anonymous Anonymous said...

Christoph Shubert is no worse than Antonio Volsenkov.

 
At 1:12 PM, February 21, 2008, Anonymous Dan said...

Who is Antonio Volsenkov? Maybe the Leafs should try to pick him up.

 
At 2:09 PM, February 21, 2008, Blogger Bruce said...

Nice work, James, although I think you're unnecessarily limiting it at the end.

It relies heavily (probably too heavily) on the quality of competition numbers, giving a free pass to anyone who plays remotely difficult minutes, but maybe that's the point.

Maybe that's your point, but I don't see the logic in simply letting all players who are above the water-line in terms of QUALCOMP completely off the hook. Some of those guys are getting eaten alive.

One would assume that a coach knows his team's weaknesses as well as anyone, and anyone linematching would therefore hide those players.

Well, he would hide them unless they're Tampa's guys and he rolls them out there against anybody. Brad Richards for example faces a slightly softer QUALCOMP (+0.06) than either Lecavalier (+0.14) or St. Louis (+0.11) yet has the worst defensive record of the three.

More importantly, those latter two make up for their poor defence by trading goals when they're out there. Both are + players (VL +62/-58; MSL +63/-60) whereas Richards is +35/-58 for an horrendous differential of -1.44/60. With his exorbitant ice time it follows that he has the worst minus in the entire NHL.

Just looking at the guys on your second list, the worst such diffentials on a per-60 basis are Nylander (-1.90/60), Tlusty (-1.75), Handzus (-1.63), B.Richards (-1.44), Briere (-1.42).

While most guys on the list are in the neighbourhood of -1/60, an exception can be found in the group of other Tampa players, of whom VL and MSL are the only two net plusses and Prospal and Ranger are at least close to break-even. No wonder it's Richards who has been put on the trading block. Caveat emptor.

 
At 3:20 PM, February 21, 2008, Blogger The Forechecker said...

Don't be too surprised to see Shea Weber in there. When he was hurt earlier this season and came back in November, he was terrible for about 15 games until a second injury stint and recovery right around the New Year. Since then, he's been more of the player people expected.

During that first run, he was even getting outscored in his power play time, having been on the ice for 4 shorties against, and only three PP goals scored.

 
At 2:06 AM, February 22, 2008, Blogger Mogen_david said...

Have you thought about using a ratio for on ice vs off ice ga/60 instead of a difference? This would penalize a skater even less for being on a team with a poor team defence. It's a little less intuitive but scales a player performance to his team performance.

 
At 9:27 AM, February 22, 2008, Blogger Julian said...

Shouldn't something about team SV% be included? A guy who plays infront of a .890 goalie is gonna have worse numbers than a guy who plays infront of a .925 goalie.

I know, I know, players affect the shot quality, but I'm not sure that makes enough of a difference that one should ignore the impact of the goalie behind the players.

 
At 9:34 AM, February 22, 2008, Blogger James Mirtle said...

The goals against on/off metric will take that into account; I'm not looking at just goals against, but how the team's goals against changes with certain players on the ice.

Look at how many Canucks are on the final list; how can goaltending be a factor for them?

 

Post a Comment

<< Home

Links to this post:

Create a Link


.

Free Page Rank Checker
eXTReMe Tracker