Takes and trash talk from both ALL sides of the NHL's most obscure PATHETIC* rivalry

* Thanks, Kevin Lowe!

Thursday, December 27, 2007

Ducks Gameday—Behindthenet does BoC a solid

Anaheim Ducks (18-15-5, t-7th in west) at Edmonton Oilers (16-18-3, t-12th in west)

I'm not going to write too much about tonight's matchup; you know where to go to read up on that. Instead, I'm going to praise Gabe Desjardins at Behind the Net for his always-solid work aggregating and crunching through NHL data to really help promote the notion of data-driven hockey analysis. In the past, I've talked up his strength-of-opposition and strength-of-teammates metrics; they really help to put a context around a player's 5-on-5 production (for example). This season, he's added lots of other interesting stats; you could spend a solid day looking through the metrics on his site.

One thing that really caught my eye over at BTN's blog, though, was tracking shots and goals from where they are taken on the ice. This is a pretty fascinating new layer of data, and while its accuracy or its usefulness is still being explored, I think it's pretty darn cool that the data is being collected. It may not mean much just yet, but I think it's day is coming.

As an example, I asked Gabe in the comments of his blog whether it would be possible for him to run some shot and goal charts for the BoC teams, specifically on their 4-on-5 penalty kills. Why the PK? Well, just for starters we've observed this year how the Sharks are killing penalties with ease, whereas the Ducks and Kings have struggled in that area. Perhaps by looking at the shot charts side-by-side maybe some aspect might become apparent--some difference in how the different PK units are allowing shots, and which shots are becoming goals.

All this data is pre-Christmas, 4-on-5 only, and just for illustrative purposes.

Anaheim Ducks:
200:28 4-on-5 time, 170 shots against, 28 goals against, 26.87 expected goals against


First off, let's just get a handle on what we're seeing here. I've indicated on the chart where the goalie sits and where the red line is; you'll have to imagine faceoff circles and the rest from there. Every shot that didn't score is indicated on the chart with a blue X, and every goal that was scored is highlighted with a red O, which I re-colored myself.

A lot of what's emerged is fairly intuitive; there seems to be a lot of activity and a lot of goals that cluster right around the goalie. These may be one-timers or rebound shots, but it seems that power play units are working to generate shots from that "danger zone", and the better they seem to generate those shots, the likelier it will be that the PP unit will find success. Also somewhat interesting are shots from distance--they seem to be taken with a lot of frequency, but rarely do they seem to beat the goaltender. Still, you have to say that the point shot does have an effective use; overall, teams still score against Duck penalty-killers with good effectiveness.

Los Angeles Kings:
177:16 4-on-5 time, 189 shots against, 30 goals against, 32.39 expected goals against


Here's the same chart for the L.A. Kings this season, another squad that's struggled to keep the puck out of their net when shorthanded. A lot of the same observations apply, although note how much more vulnerable the Kings' "danger zone" leakage appears to be. That seems like a thing for that unit to maybe focus on defending.

Also a bit strange seems to be bleeding from one side of the ice; it looks as if a lot more goals are being scored on the Kings from the goalie's right side than from the goalie's left, even though the shot counts seem kind of equivalent from both sides. I think this is a question maybe for observant Kings fans. Is this a function of having dangerous PP options playing on one side, or maybe is it a flag that there is a weak defender on one side? Perhaps this is a goalie issue that could be explored, who knows? I'm not really sure what the proper response for a team that notices these trends, but it is worth noting where trouble patterns emerge; addressing them might be the key to improving PK performance.

San Jose Sharks:
170:26 4-on-5 time, 107 shots against, 13 goals against, 16.56 expected goals against


And for contrast, here are the results for a successful penalty kill, courtesy of the Sharks. A large difference seems to come from shot prevention; this chart isn't nearly as dense as the Ducks' or Kings' were. Also, it seems that the Sharks are much more adept at defending the so-called "danger zone"; there's a cluster of activity, but it seems much more controlled and limited. There seems to be a lot more ability to prevent point shots from getting through; this probably contributes a great deal to SJ's success.

At any rate, I don't want to put out any super-strong statements about what the data means or indicates; we're still in a bit of an exploratory phase (certainly feel free to leave your own observations in the comments). Mainly, the thing to take away here is that the data is being collected and even at this stage can be examined in different manpower situations. The consistency of the data may be a bit suspect at this point, but for sure that will improve, as will analysts' ability to incorporate these coordinates into aggregated findings. For today, these charts are interesting and a bit revealing, but probably not very "actionable". Looking to the future, though, I can see where these sorts of charts will become fundamental for coaching analysis. We are probably on the brink of a new age in hockey understanding, and I'm glad there's guys out there like Gabe that can put it in a form that idiots like me can toy with.

Thanks again, Gabe.

Prediction: Crap. No Pahlsson tonight, or for any of this road trip. I won't be able to watch this game tonight; I'll be at my mom's birthday dinner. Still, somehow the Ducks claw their first victory from the Oil tonight in their last meeting of the season, 4-2. Goals by Bertuzzi, Weight, Pahlsson, Getzlaf, and Beauchemin.

Go Ducks.

9 comments:

Anonymous said...

Cool stuff! I'm wondering if it's possible to annotate a "second chance" shot. With the amount of pucks getting past JSG from the slot, it'd be interesting to see how many of those were rebounds. Rebound control ought to be coachable - for the goalie, and for the D.

Earl Sleek said...

I think that sort of analysis is possible for sure--mostly what I've seen about rebounds is that they can be selected with criteria--a shot that happens within 2 or 3 seconds of a previous shot, something like that.

Really, my head is spinning a bit thinking of all the possibilities--I think it would be cool to see these sorts of charts when a specific player is on the ice, too. How does Pronger's 4-on-5 experience differ from Beauchemin's (for example)? How would Getzlaf's PP look compared to Weight's?

Like I say, it's tough at this point to draw too much out of these 'first-blush' charts, but it's cool to think about where they could be incorporated and studied.

Anonymous said...

watching the sharks PK last nite..I realized how more aggressive they are as opposed to how passive our team plays the Pk....the sharks play aggressive but somehow do no get out of position...wonder how that factors in...cool chart..fo sho

p.s. frankie is due for a goal no doubt...I think his lack of scoring ahs to due with getting beat when his shot gets blocked...maybe with nieds back hell be a little more confident..case in point (and Ive suggested this to hnidy as well) pronger needs to have a clinic on the art of not getting your shot blocked...chapter one = look up when you shoot

Patty (in Dallas) said...

Fascinating chart! I'm fascinated with penalty kills and I'd love to see the Stars compared to the BoC teams.

I'm not as sophisticated an observer, so I'm not sure what makes theirs so good. Plus, the Stars are the only ones I watch that closely.

One thing the Stars seem to do, when they're on, is just not let the power play get into the zone. I'd guess that the number of shots is very low, like the Sharks. It would just be a guess.

Maybe a Pacific division comparison would be interesting for you BoCers?

Earl Sleek said...

Well, including the Stars might be interesting, but I'm not sure it would be very encouraging.

I don't have the data, though. This was purely a behindthenet favor, based on a request.

Patty (in Dallas) said...

Yeah, we pretty much got the penalty killing down (knock wood).

Anonymous said...

I wonder how many x ed point shots are first assists, similar to the first comment.

andy grabia said...

Hey Earl, can you send me that goalie image? Gabe ran some stuff like that for me, too, and I'd love to use that image.

Earl Sleek said...

Grabia, I love that little guy. I shot an e-mail to your gmail account; let me know if you need it to go somewhere else.