Navigation: Jump to content areas:


Pro Quality. Fan Perspective.
Login-facebook
Around SBN: Devils Beat Rangers, Head To Stanley Cup Finals

Graphical goodies from Behindthenet.ca

There's nothing like an unexpected present to make a girl happy. Although I have to admit that I'm pretty weirded out by how excited I am to get a statistics present. I have no idea at what point this number thing became less of a chore and more of an obsession.

I had been doing some research on other things when I got distracted by the Lightning's shot data. I spent quite a bit of time trying to see what sort of trends I could find in the data from last season and this one, and then over Christmas, Behindthenet.ca came out with some fascinating new graph features that essentially made the hours and hours I put into this project redundant. ... Yay.

In all honesty, it really is a "Yay!" sort of thing, especially for folks like me who aren't quite sure what to do with all these numbers that are floating around. What these new graphs help to show are (a) puck possession over the course of a season and (b) sustainability. [Note that all of this data is cumulative and not game by game. It's also 5v5 data. No power plays or penalty kills.] They've got graphs for every season back to 2007-08.

Star-divide

T

via behindthenet.ca (click to enlarge)

On the left, you have the team's shot ratio in blue (this includes missed shots but not blocked shots) and goal ratio in green. Both are expressed in percentages of total shots or goals in the games. The idea is that teams that possess the puck better will win more games. Since it's logistically impossible to determine who technically "has" the puck in a hockey game and for how long, shot data stands in as a metric for puck possession. Possession leads to shooting, shooting leads to scoring, and scoring leads to winning, theoretically.

Ideally the blue line in that graph will be above 50% (or .5). [Well, ideally both lines will be up there.] That would mean that the Lightning were taking more shots than their opponents. It certainly was that way all of last season. This season, not so much. It appears to be getting a bit better. But then, in general, over the course of a season shot ratios tend to regress towards the 50% mark, although there's still a lot of variation from team to team.

Essentially, this chart shows in graphic form what I've been arguing numerically over the past month and a half: The Lightning have a shot problem.

5-on-5 Shot Data

Total Shots

Total TOI (seconds)

TOI/60

Shots/60

2011 SF to game 37

817

104519

29.03

28.14

2011 SA to game 37

847

104519

29.03

29.17

2010-11 SF full yr

1993

227945

63.32

31.48

2010-11 SA full yr

1730

227945

63.32

27.32

2010 SF to game 37

907

101340

28.15

32.22

2010 SA to game 37

722

101340

28.15

25.65


Not only are Shots For down significantly over last season, but Shots Against are up significantly. You can see how those two things are related. The Bolts simply haven't had the same control over the puck as they did in 2010-11 and unless that gets corrected, there will continue to be a Goals Against problem. [This doesn't address the power play issues in any way, but there is no doubt that power play success helped the Lightning out in 2010-11.]

The graph on the right is PDO, which as I've mentioned in an earlier post, is a measure of sustainability. Over the season, that line will get closer and closer to 1. When it's below it will go up; when it's above it will go down. From what I've seen of the last few years' data, anything between .99 and 1.01 is pretty sustainable. The Bolts ended last season at .991, by the way.

For most of this season, the Lightning's PDO has been not only below 1.000, but it was declining between late November and late December (after game 19: Nov. 19 vs. the New Jersey Devils). That decline has been a result largely of a sinking shooting percentage (as represented by that big gap between the blue shot line and the green goal line in the middle of graph 1). Over that period, team shooting percentage dropped from around 9.5% to about 8.4%. It has come back up to 9.4% over the past three games.

Both shooting percentage and save percentage vary wildly from game to game. They both need to be taken as trends rather than individually or even over a short series of games. Cumulative SV% has been hovering between .897 and .907 since game 11 (Oct 29 vs Winnipeg.) That's not unheard of, and although it's lower than fans might want it to be, it's higher than it was until about game 60 of last season. The goalies have turned in ESSV% above .920 in 6 of the last 10 games, and it needs to stay that way to give the team the chance to win games.

As of right now, the Lightning are sitting very close to 1 in PDO, which means that "luck" (really, randomness) isn't affecting their play all that much at the moment. They aren't getting any particularly good or bad bounces that would explain their scoring or save rates. That doesn't mean that the team will necessarily continue to win games. It simply means that they aren't being especially hindered or helped by elements outside of their control.

Anyone who wants to check out these lovely charts can do so at behindthenet.ca/T.B_2011.html. On that page you can find the graphs, as well as the data behind them in a table at the bottom. If you do go look, remember that the data is cumulative and not game-by-game, that the TOI is measured in seconds (while all /60 stats are in hours), and that there are drop-down boxes at the top of the page that give you a quick way to move from season to season or team to team.

Comment 22 comments  |  1 recs  | 

Do you like this story?

Comments

Display:

So it looks like bad luck for the last month or so

being counteracted by good luck the last two weeks or so, yielding composite results of about what we should expect.

Good stuff. where’d you find the partial season data for last year? I was hunting for that yesterday and missed it

If I cared more about my UNC side, I'd call myself "Tar Volon," and that'd be awesome.
Bolts, Canes, Preds (now in different conferences!). Canes mini-STH. Southern hockey solidarity
Rocky Top Talk

by Incipient_Senescence on Jan 2, 2012 6:07 PM EST reply actions  

I still think that this team is below where it should be expected, given the talent on the roster. But the improvement and bad luck that Boucher has been talking about over the past weeks seems to show up in this data.

As for the partial season data, I had to take their cumulative data and “unstack” it. I still can’t find power play shots broken down by date, which I really, really want. Their data, btw, is given by game number, and you have to then go find out what the associated dates were.

One feature I didn’t get to was that if you click on the game number in the table at BTN it will give you a Fenwick graph for that game.

R.I.P. Belak, Rypien, Boogaard, Lokomotiv.
Rakastan suuria maalivahteja Skandinaviasta.
Love is but a song we sing And fear's the way we die.

by CAustin on Jan 2, 2012 6:20 PM EST up reply actions  

unstack?

I don’t follow

If I cared more about my UNC side, I'd call myself "Tar Volon," and that'd be awesome.
Bolts, Canes, Preds (now in different conferences!). Canes mini-STH. Southern hockey solidarity
Rocky Top Talk

by Incipient_Senescence on Jan 2, 2012 6:33 PM EST up reply actions  

They add each game’s stats to all the games before it and give only the totals.
Game 1: TOI: 2597 GF: 3 SF: 26 GA: 1 SA: 27
Game 2: TOI: 4853 GF: 4 SF: 41 GA: 5 SA: 62

I had to create a spreadsheet that subtracted it all out to get
Game 2: TOI: (4853-2597=) 2256 ….and so on.

R.I.P. Belak, Rypien, Boogaard, Lokomotiv.
Rakastan suuria maalivahteja Skandinaviasta.
Love is but a song we sing And fear's the way we die.

by CAustin on Jan 2, 2012 6:41 PM EST up reply actions  

would you mind just linking to the part that has that stuff

I’m having a lot of trouble navigating the site. I can’t even find the fenwick numbers without following the Arctic Ice Hockey direct link

If I cared more about my UNC side, I'd call myself "Tar Volon," and that'd be awesome.
Bolts, Canes, Preds (now in different conferences!). Canes mini-STH. Southern hockey solidarity
Rocky Top Talk

by Incipient_Senescence on Jan 3, 2012 3:14 PM EST up reply actions  

Not sure what in particular you're looking for

behindthenet.ca/T.B_2011.html

The data is at the bottom below the graphs, sorted by game number. Click on the game number for that game’s Fenwick chart. You have to connect game number to date on your own.

R.I.P. Belak, Rypien, Boogaard, Lokomotiv.
Rakastan suuria maalivahteja Skandinaviasta.
Love is but a song we sing And fear's the way we die.

by CAustin on Jan 3, 2012 4:23 PM EST up reply actions  

ah, I gotcha

didn’t realize you were talking about estimating. but that’s better than nothing. thanks

If I cared more about my UNC side, I'd call myself "Tar Volon," and that'd be awesome.
Bolts, Canes, Preds (now in different conferences!). Canes mini-STH. Southern hockey solidarity
Rocky Top Talk

by Incipient_Senescence on Jan 3, 2012 5:49 PM EST up reply actions  

I didn’t estimate on this post. I used their data. But their data was for ALL shots and saves made at 5v5 and I had to separate it out to find out the individual game save percentages, for instance. If you want to look at the Fenwick data by situation (tied, up one, down one, etc.) and by date you can’t get it from this particular feature. You can try to figure out the difference between 18 and 19 shots on the chart, but you’d have to estimate. And it’s still 5v5 data.

One of the really, really big issues with the hockey stats movement is the sheer difficulty of getting to the data. I have yet to find Fenwick data (or special teams data) sorted by date so you simply cannot see change over time or do a before and after comparison unless you are able to pull the data out of the RTSS reports. It’s a horrible way to convince people that statistical analysis is worthwhile.

R.I.P. Belak, Rypien, Boogaard, Lokomotiv.
Rakastan suuria maalivahteja Skandinaviasta.
Love is but a song we sing And fear's the way we die.

by CAustin on Jan 3, 2012 6:20 PM EST up reply actions  

Note that the Fenwick charts don’t have the associated data listed out. You’d have to estimate, add and subtract to get actual Fenwick data to play around with unless you’re looking for season totals.

If anyone wants BTN’s actual Fenwick data it’s at http://behindthenet.ca/fenwick_2011.php?sort=6&section=close but it’s not broken down by date.

R.I.P. Belak, Rypien, Boogaard, Lokomotiv.
Rakastan suuria maalivahteja Skandinaviasta.
Love is but a song we sing And fear's the way we die.

by CAustin on Jan 3, 2012 4:28 PM EST up reply actions  

although if you just want totals through X game you don’t have to do that. Only if you want to look at game-by-game results—which I did for other projects. Just background on this post.

R.I.P. Belak, Rypien, Boogaard, Lokomotiv.
Rakastan suuria maalivahteja Skandinaviasta.
Love is but a song we sing And fear's the way we die.

by CAustin on Jan 2, 2012 6:47 PM EST up reply actions  

yeah, I was just looking through X games

I didn’t see that they had that

If I cared more about my UNC side, I'd call myself "Tar Volon," and that'd be awesome.
Bolts, Canes, Preds (now in different conferences!). Canes mini-STH. Southern hockey solidarity
Rocky Top Talk

by Incipient_Senescence on Jan 2, 2012 9:11 PM EST up reply actions  

Excuse my ignorance real quick.

I’m not sure I understand the Goal% line on the graph….so that’s like Fenwick, but calculated using goals instead of shots?

Also, great find with all these charts…this is really cool stuff. And it does paint the picture well….the recent bounce in their PDO jives with what you’d expect. They really have been playing considerably better of late.

I love Casey Fossum. Now try and take me seriously.

by Steve Slowinski on Jan 3, 2012 12:19 AM EST reply actions  

It’s GF/(GF+GA). Oh, and I found out why Fenwick doesn’t use blocked shots: because shot blocking is a skill and not random. For whatever that’s worth. I’m not sure I buy the logic there, because it doesn’t mean a guy didn’t have the puck when he shot it. And because I consider goaltending just as much a skill, too, and that affects the outcome of a shot. Anyhoo…..

I like being able to see it all at a glance. I’d been making up charts for that very reason when these came out.

R.I.P. Belak, Rypien, Boogaard, Lokomotiv.
Rakastan suuria maalivahteja Skandinaviasta.
Love is but a song we sing And fear's the way we die.

by CAustin on Jan 3, 2012 10:42 AM EST up reply actions  

Makes sense...is it the same formula for the Shot% presented above?

Because I was assuming it was Fenwick, but I’m just now realizing that it has the Lightning with a higher % than I would expect.

…Although that may only be because I’m looking primarily at tied Fenwich (46%), while the Bolts are spending lots of time behind in the game (in which cases they’re well over 50%).

Interesting about Fenwick and the blocked shots….huh. Yeah, if the idea is to simulate puck control, I’m not sure why you wouldn’t include blocked shots. Corsi% would almost seem to make more sense.

I love Casey Fossum. Now try and take me seriously.

by Steve Slowinski on Jan 3, 2012 11:23 AM EST up reply actions  

Yes, the shot ratio is the same, not broken down to get at score effects. All 5v5 shots or goals.

R.I.P. Belak, Rypien, Boogaard, Lokomotiv.
Rakastan suuria maalivahteja Skandinaviasta.
Love is but a song we sing And fear's the way we die.

by CAustin on Jan 3, 2012 11:51 AM EST up reply actions  

FWIW

almost everyone is well over 50% when they’re behind, but the Bolts are still in the middle of the league. Almost everyone is under 50% when ahead, but the Bolts are still in the middle of the league. When tied, the Bolts are awful

If I cared more about my UNC side, I'd call myself "Tar Volon," and that'd be awesome.
Bolts, Canes, Preds (now in different conferences!). Canes mini-STH. Southern hockey solidarity
Rocky Top Talk

by Incipient_Senescence on Jan 3, 2012 12:14 PM EST up reply actions  

Haha!

I’m not sure how predictive this kind of information is. I tend to think of it more as a diagnostic tool than a predictive tool, except in very broad strokes. Basically it confirms what we’ve been seeing with other ways of looking at the data—that the season has been really disappointing so far and that that disappointment has to do with puck possession. IF the team keeps playing the way they have very recently, and IF they don’t run into a bunch of teams having extraordinary luck, and especially IF they are able to improve their control of the play, there is hope for a playoff run. But that’s a lot of IFs.

As always, your mileage may vary.

R.I.P. Belak, Rypien, Boogaard, Lokomotiv.
Rakastan suuria maalivahteja Skandinaviasta.
Love is but a song we sing And fear's the way we die.

by CAustin on Jan 3, 2012 11:23 AM EST up reply actions  

Others possibly disagree with me.

But I’m in the Hold camp. I don’t think there’s any one piece they can bring in that’ll turn the Lightning into contenders overnight, and there’s no sense mortgaging the future on a long shot of a playoff run.

The Bolts will bring their core back again next season (and the season after that, and the season after that), so I don’t see there being a rush to make any big moves. We can hopefully fix the goaltending problem after the season (and in the meantime, see how Garon does with the primary duties), and shore up the defense some more. And who knows, the Bolts have a talented roster and could still make things interesting with the roster as is (and some good luck down the stretch).

I suppose if there’s an trade to be made for a defender that could help the Bolts this season and in the future, I’d be in favor of an in-season move. It depends on what they’re asking, though (obvious statements are obvious).

I love Casey Fossum. Now try and take me seriously.

by Steve Slowinski on Jan 3, 2012 11:29 AM EST up reply actions  

if we can net either a youngish defenseman or a youngish goalie mid-season without mortgaging in the future

we’ll obviously do it. and that’ll help a lot. but if there’s none available at the right price, I imagine we’ll try to work free agency

If I cared more about my UNC side, I'd call myself "Tar Volon," and that'd be awesome.
Bolts, Canes, Preds (now in different conferences!). Canes mini-STH. Southern hockey solidarity
Rocky Top Talk

by Incipient_Senescence on Jan 3, 2012 12:16 PM EST up reply actions  

I have to agree that summer movement's the most likely scenario.

As frustrating as it can be to watch the team have a “bad” season, I’d expect Yzerman to be extremely cost-conscious when it comes to the trade market. Short-term gains aren’t going to be enough. At some point depth has to be added and getting a Nabokov for the season, for instance, at the cost of either picks or prospects (i.e., the future) seems to be a bad deal. I’d expect him to be looking for help both defensively and in goal—or anywhere else, really—but I’d also expect him to have very high standards for what’s an acceptable trade.

R.I.P. Belak, Rypien, Boogaard, Lokomotiv.
Rakastan suuria maalivahteja Skandinaviasta.
Love is but a song we sing And fear's the way we die.

by CAustin on Jan 3, 2012 12:31 PM EST up reply actions  

Comments For This Post Are Closed


User Tools

"The static cling that brings Tampa Bay Lightning fans together", SB Nation's Bolts community is your place for news, commentary and camaraderie.

Pages

Posting Guidelines on Raw Charge

Lined-Up Update (2010 off-season)

Raw Charge apparel

Featured Poll

Poll
20th Anniversary Season poll: Favorite Tampa Bay Lightning logo?

  51 votes | Results

Raw Charge on the Web


Raw Charge on Tumblr

eXTReMe Tracker

Managing Editors

2011-03-06_12_small John Fontana

Imag0026_small Cassie McClellan

Writing Staff

Small Dani Toth

Ads_small Tina Robinson

Beardweekone_small Matt Amos

Photo_small Clark J Brooks

Snowman_in_net_small CAustin

Small shawndubstep