Are shots off the rush on offense pushing Rangers’ SH% higher?

The New York Rangers are in the playoffs. But Ryan Lambert of Puck Daddy raised a question that struck a nerve with a lot of fans: Should the Rangers scare anyone? The article uses raw SAT%/CF% and SPSV%/PDO to justify that the Rangers are a poor possession team that are riding a great season of luck. These were some concerns I raised two months ago. I also revisited it one month ago, and found some interesting trends.

But here’s the issue with using just those numbers: It doesn’t take into account certain variables (like systems and rush shots, more to come), it doesn’t break down trends, and it looks at a full season instead of how a team is playing now. Not all numbers can be taken in a vacuum, some need context. This is one of those times where context is needed.

First things first: Lambert is correct when he says the Rangers aren’t the best puck possession team and that they have an unnaturally high SPSV%/PDO. But that’s about the only thing he’s 100% right with. I’ve tackled New York’s possession in the past, but we can revisit here.

Using score-adjusted FF/USAT (better for cross-team comparisons, as it eliminates system bias for shot blocking) as of December 1, the day the Rangers had both Ryan McDonagh and Dan Boyle in the lineup for the first time, the Rangers are sporting a 51.4% possession rate. That’s mediocre, at 14th in the league and 7th in the Eastern Conference. Their last 11 games have been pretty bad though:

USAT%/FF% as of 12/1, rolling 10-game average
USAT%/FF% as of 12/1, rolling 10-game average

Games against Buffalo (74.5%) and Ottawa (60.7%) are the only two games above 50% possession for the Rangers during that 11-game stretch. But yet, the Rangers kept winning. Based on raw SPSV%/PDO (Rangers are tops in the league at 102.1 since December 1), one can make the conclusion that this is all luck driven. But that may not be the case with the Rangers.

We all know the Rangers are sporting an unnaturally high SH% at 8.7% since December 1 and 8.9% for the season. That part we know. But let’s get into this in a deeper dive.

The Rangers are built on speed. On offense, this means long passes, speed through the neutral zone, and getting that trailer in on the rush, trapping forecheckers and outnumbering the opposition as they gain controlled zone entries. We see this all the time with the “Chris Kreider play,”

rangers senators 2015
The Kreider play, Stepan with the puck outlets to Kreider off the boards for a chance off the rush.

the stretch passes from blue line to blue line,

rangers senators 2015
This is the end result of a Yandle pass (from NYR’s blue line) to Hayes, creating a rush.

and cross-ice passes to the trailer following zone entry.

rangers senators 2015
Trailer (Boyle) joining the rush, created off a turnover in the defensive zone.

All three of these plays happened against the Ottawa Senators last week.

On defense, the Rangers use their speed with one forechecker deep, two on the blue line, and then two on the red line to pressure the opposition into turnovers, something else that was on display against Ottawa.

rangers senators 2015
Speed/Forecheck creates turnovers.

The forecheckers close the gap well, and the puck support in case of a turnover is ready to transition, thus creating more chances off the ensuing transition rush.

The common theme here is the rush. This is an area of shot quality analysis that has trended towards the notion of not all shots are created equal. It’s something that we know to be theoretically true, but has been tough to quantify with the tools available. Hockey Analysis has taken a stab at it though, and has concluded what we believe to be true: Chances generated off the rush have a higher SH% than league average SH%.

Hockey Analysis defines a rush shot as:

  • A shot within 10 seconds of a shot attempt by the other team on the other net.
  • A shot within 10 seconds of a face off at the other end or in the neutral zone.
  • A shot within 10 seconds of a hit, giveaway or takeaway in the other end or the neutral zone.

It’s not perfect, since it omits turnovers in the high-risk area from the blue line to the top of the circle in the offensive zone (this is where the Boyle goal above comes from). That’s a small, though important, subset of attempts. But for the sake of this post, and for the sake of the types of goals we generally see from the Rangers (noted above), it more than suffices.

The article from Hockey Analysis shows that the Rangers jumped from a SH% under 7% on non-rush shots to a 9% shooting team* for rush shots. Hockey Analysis also shows that SV% drops by about 2% when a goalie is facing a rush shot*. The Rangers were also second in the league to Boston over a three year period for percentage of shots that are rush shots (25.72%).

*-This is using last season’s data.

This year, the Rangers have the largest difference of shots taken off the rush this year from last year, up 7.24%. That 7.24% increase in rush shots has led to a 2.47% increase in raw SH%.

Last season, the Rangers took 33.2 shots per game over 82 games, which is roughly 2,722 shots. Of that, 25.72% –700 shots– came on the rush (based off a three year average, best I can do here). The Rangers take 31.4 shots per game this season. Multiply that by the 75 games they’ve played this season, and you get 2,355 shots this season. Their number of rush shots this year is up 7.24% from 25.72% to 32.96%. So the number of rush shots this season, thus far, is 776.

Based on numbers from HA, here’s the best I can do for a breakdown of these numbers from last year to this:

Season Shots Overall SH% Rush Shots Rush SH% Rush Goals Non-Rush SH% Non Rush Goals
2013-2014 2722 6.70% 700 10.09% 71 6.40% 129
2014-2015 2355 8.70% 776 12.50% 97 7.60% 120

These are rough estimates, as this stuff isn’t readily available for me to break down on my own. I have to rely on raw shot data (which apparently varies from site to site), some limited single season data from last year (from Hockey Analysis), and three year averages (also from HA). These numbers will not match up to actuals. Until this is readily available, this is the best I can do here, sorry.

I got to the increased SH% this year by simply taking the three-year SH% averages in rush/non-rush shots and adding the 2.47% increase noted above manually. I’m aware that’s not how it works, but again, guess work here.

Let’s theorize a bit here. First what we know:

  • SH% goes up on rush shots while SV% goes down, for the most part.
  • The Rangers are styled to get more chances off the rush.
  • The Rangers statistically got more shots off the rush than the rest of the league over the previous three seasons.
  • The Rangers statistically generate more shots off the rush this season than last year.
  • The Rangers have seen an increase in SH% from last year to this year.

So we, from the above, we can reasonably and logically theorize that the Rangers will have a higher SH% than the league average. This, unfortunately, is something we cannot actually prove, but we have enough facts that point us in this direction. Maybe, just maybe, the Rangers’ high SH% is due to their score-off-the-rush design.

Show More
  • I think this is right. I hope this is right! I saw the discussion in Hockey Prospectus and it makes sense. PDO is excellent at predicting the type of goal when the defense is worn down after a long shift and the goal is often scored by jamming the crease. I think/hope that is not the only way to win playoff hockey games. A skill speed team lives off the rush and the cross ices pass, as you diagram. Whichever style is used, the team will need to be at top form and have a lot of luck to win four playoff rounds.

    • I wish there was more info available so that I could prove this, but it won’t be available until SportsVu player tracking is available.

  • I think the Rangers rush style of attack also leads to their low(ish) possession numbers as they don’t necessarily have the zone time to put up shifts of 5-8 SAF and with the defense being not so fleet of foot they can’t get back after the usual one and done rush attempt and it leads to more opposing chances. This piece definitely goes a long way in explaining the high sh% and hopefully with the new tracking technology the league is using, we’ll be able to get more accurate rush data.

    I will say that the Rangers should definitely change their O zone schemes a bit, instead of working blue line down they should be working goal line up and using their winger speed to create odd man situations down front.

    Kinda like this

  • this is a really good piece that explains not only the high sh% but the low(ish) possession numbers. Generating chances off the rush lead to a lower Ozone time, this then leads to lower shot attempts and with the defense being not so fleet of foot it opens up into more attempts against. I feel that the Rangers could maintain their rush offense and keep up their Ozone time by changing the angle of attack. Instead of going from the blue line down, work from the goal line up and use the speed of their wingers to create odd man situations down low.

    While they do this from time to time, it does seem to me that the Rangers offense is built solely around those rush chances and once the opposing team shuts down their alleyways they have no idea how to adjust

    • I have to disagree on the last paragraph Kevin. The Rangers do seem to score better off of the rush, but their cycle isn’t anything to laugh at. The trio of Hayes, Miller. and Hags do a nice job of it, as well as the Brass line. Over all the team does a decent job of possessing the puck, but we seem to be snake bitten at this time with the scoring!!!!!!!

      Now, if only the PP started to work????????????? Seems to me that the PP has been an issue ever since the days of Brian Leetch. Can anyone out there tell me how many attempts at fixing the PP, and at what cost to the organization these attempts. Brad Richards, Dan Boyle, Keith Yandle, and the list goes on, and on. We haven’t had a decent PP since the days of Jagr playing on it!!!!!!!!!!! Oh well, some things never change…………

      • Their cycle isn’t bad, but it’s not something they use to regularly generate chances. They are a rush scoring team.

    • The offense is one dimensional, and it’s why the Kings give them fits. The Kings prevent those long passes, they clog the NZ so that the Rangers can’t gain speed, and force dump ins. The Kings win because of their defense. They match up well against the Rangers.

      • Exactly, I think we’re saying the same thing here. The Rangers have the players to run a very very good O zone cycle in Hayes, Stepan, Zuccarello et al but they, for some reason, choose not to.

        • I don’t think it’s them choosing not to. One of the drawbacks of always shooting on the rush is that these don’t generate high offensive zone time. I don’t know if there’s data available for that though.

          • The cycle debate here….. I am starting to think that they would love to generate a cycle if they could but it comes down to ability. They shoot off the rush, don’t score and the puck is in the corner. If the Rangers players were ABLE to, they’d recover more pucks by numbers and establish zone time.

            They haven’t been doing that regularly enough though. I fear that it comes down to ability. It’s not like they aren’t trying to win pucks and establish possession. Rush chances / zone time are certainly negatively correlated, but not mutually exclusive.

    • I think you nailed it, Kevin. I’ve been pretty upset about our mediocre to poor possession this season, but lately have been trying to convince myself it’s ok if we can stay very proficient at counter acting. It’s a very fine line which will be challenged in every 7 game series which we’re in.

      At the end of the day, though, I am uncomfortable with the amount of hockey I’ve been watching where we are scrambling in our own zone. We do not impose our will enough on other teams and I’m nervous about the sustainability of the game plan once the playoffs start.

      We keep hearing about our speed speed speed- a very important element to this team. However, so is their creativity and passing ability. The PROBLEM here is that we do not spend enough time with the puck in the offensive zone to allow the creativity to manifest itself.

      To be honest, I am very happy that we seem to have hit a stretch of bad hockey. We’ve been winning for so long and quite frankly, winning games allows players to develop bad habits. Not until the winning stops will teams (for the most part) clean things up and get back on track.

      Let’s start tonight. I want more than a win. I want the Rangers to prove they are better. Dominate and domesticate the other team. No more of this ‘lets take the 2 points and run bs’

      • The bad hockey over the weekend is bringing the cliff jumpers back unfortunately. All teams hit stretched like this.

        The teams that will give NYR trouble over a seven game series are TBL and NYI because they have the speed to match. In the west, it’s LAK since their entire defensive system is designed to force dump and chase hockey.

  • I don’t know all this stuff, and I’m not smart enough to figure out all of these diagrams and calculate all of these SAT/PDO% numbers. I did pretty well on my SAT’s, but that was a while ago. Dave, we all love ya, and I enjoy reading your columns, and the witty repartee from all of the posters as well. And I try to figure out the graphs and the diagrams and all the stuff that you get from Hockey Analysis and others; I really do. So don’t take this the wrong way; it’s not a knock on you. But….

    Do any of these calculations tell us how many times Alexander Ovechkin will score if no one on the Rangers pays attention to the fact that he’s standing 18 feet in front of Cam Talbot for a half hour waiting for the puck? And which graph can I go to to tell me what will happen if Keith Yandle hands the puck to an opposition forward in the Rangers’ zone? We all know what will happen if the other forward has half a clue what to do with the puck.

    This we know. The Rangers are what we call in soccer a counter-attacking team. They back-check and and take off when the puck changes hands. They’re not a “half-court” possession and fore-checking team. If they can impose their speed game, cross-ice passes and so forth on the other team, they will have 55% or so of the good scoring chances in a game, instead of 45%. If not, the other team will get the 55%. Whether the puck goes in the net depends on a lot of things, not all of which fit onto a graph, right? Sometimes Rick Nash is in a slump. Sometimes the shot hits the pipe. Sometimes the puck hits a little bump in the ice and rolls up Derek Stepan’s stick. Sometimes the goalkeeper makes an awesome save, or a lucky one, or whatever.

    The NHL playoffs can be unpredictable; that’s why we love it so much. Keep up the good work. Regards-


    • You’re looking at singular plays. The definition of statistics is the evaluation of trends over large samples. Your singular plays pointed out above will show up in the larger evaluation process, even if it isn’t called out specifically.

      What part of the charts do you not understand? Maybe I can help. Email me using the mailbag feature or comment here.

    • If LA is like a half court team, then there is no shot clock and they are moving around and around in such a way that the defense gets too tired to defend. Then they clog the paint and look for lay ups and dunks.

      I think the stretch of games we played against Det, Chi, etc on the road a while back is the Rangers team we will actually see in the playoffs. Lots of 2-1 1-0 and overtime games. Yes we will need boatloads of luck to make even the conference finals, but that is true every year.

      • Hockey is largely controlled by luck. You make your own luck with skill, possession, etc, but so many little things control the outcome of the game. It’s why the game is so fun and so maddening.

        Those 2-1, 1-0 games in that stretch were still exciting games. If the Rangers are allowed to gain speed through the NZ, not many can skate with them.

        • Not sure luck is really in play here. You create opportunities with puck possession, and you cannot create opportunities without the puck. To me, it isn’t luck, but superior control that wins games.

          Kind of like keeping Brady or Manning off of the field. Either QB is going to hurt you, but if the ball isn’t in their hands, there isn’t much they can do.

          • Well yes, you can create your own luck, but there are such freak things that can throw off a seven game series. Pucks hopping over sticks (Girardi, Game 1 vs LAK), blind luck on blocked shots (Voynov on Nash, Game 5), the patch of ice behind Hank in Game 4.

  • I don’t pay attention to all the fancy stats that the geek boys come up with. I prefer to just watch the games and let my eyes tell me what I’m seeing. But, don’t you find it just the least bit interesting, that now that the Rangers have lost a few games in a row, and they seem to have hit a little bump in the road; that NOW, everyone is taking notice of them and pointing out why they aren’t legitimate contenders? Their glee at analyzing the losses to Boston and LA is laughable. They’re even talking about a possible goalie controversy with Hank coming back! Its all just a whole lot of nothing. A bunch of guys who don’t think the NYR are any good to begin with; trying to justify their opinions. Just line em’ up and start the playoffs already. And when we’re at the parade, they can talk about how LA’s parade last year was better!

    • Joe I couldn’t agree more. We’ll find out if they’re good enough starting on April 15. Drop the puck already!

      Regards- orange

    • These aren’t “geek numbers made up.” That is an ignorant statement. SH% has been a stat in use for a while. There is nothing “fancy” or “new” about this.

      Side note, enough with the “stats suck” argument. The stats people won. By ignoring the newer stats available, you limit your analysis. Both are needed.

        • I think a lot of it has to do with how the stats are portrayed. They aren’t be all/end all. Need to evaluate using stats and watching games. Too many people think it’s one or the other.

          This post uses both. I theorized that NYR generate a higher SH% because of how they are styled to play, and there are enough facts that point to this conclusion being true.

      • Just because you live and die by the stat numbers, and can’t seem to make any arguments regarding this team’s play, without referring to your precious numbers and %’s; doesn’t make anyone who disagrees with you ignorant. Nobody said the stats were made up.”Come up with” is what I said. I’m aware that some thought is involved in coming up with their results. I never said anything was fabricated. If you’re gonna’ quote me, at least quote what I wrote; not what you read. If we go by statistics only, then why bother even playing the games. This Org. has only won 1 Cup in what, 68 years? If we go by stats alone, why are we wasting our time watching? The stats are for the coaching staff. Let them deal with it. This isn’t “Money Ball”(and by the way, how many Trophies has Billy Beane won with his precious stat #’s?) No one is saying to disregard stat numbers; over time they do predict trends. But there are other factors to consider. Some more important than just the numbers.

  • Nice explanation. I worry that the “home-run-pass” will not be as available in the playoffs, unless we score first or have a lead. Then, maybe you find a defenseman cheating a bit. If we play a team like the Bruins or even Montreal, that seem to be able to cycle and have long stretches of puck possession, I am fearful that we will make that bad coverage decision and allow the opponent to score.

    I believe that a speed team usually isn’t a possession team. What a speed team also needs is strong defense and low turn-overs in the neutral zone. Our speed was not a factor in the last 2 games because we made coverage mistakes, turned the puck over and fell behind.

    The Good – Kreider, Nash and Hags – Need to stretch the opponents coverage with their speed

    The Cycle – Hags, Zucc, Hayes, Brass & Steppan – Need to cycle and possess and find the open slot

    The Defense – Hunwick, Boyle, Yandle – Need to tightly cover the opponent in our zone

    The Goalie – Other teams goalies look great against us if we do not get shots off the rush. We need to play tough, get value out of all 4 lines and Hank to reign again

    • Montreal isn’t a good team. They are riding Price. They don’t really scare me. Boston doesn’t either since they don’t have the depth. The Isles and Lightning scare me.

      Regarding the last two games, they got sloppy in their own end. It is what it is, and people are unnecessarily panicking over it. Kind of humorous actually.

      The Rangers started getting really sloppy when MSL was out.

  • The eye test would confirm your hypothesis. But seriously, this is good stuff and goes a long way toward explaining the Rangers’ success despite mediocre possession numbers. The concern is, as stated above, come playoff time when the ice gets smaller, can the rush style continue to generate this type of success?

    • Especially since SH% can dry up fast in the playoffs (see: Nash, Rick). It’s certainly a concern, and I’d like to see them control the puck more in the OZ. But it’s working.

  • The problem is that we’ve come to believe that these possession numbers are correct. Corsi’s bias is incorrect. Assuming a fundamental understanding of the game, would you trade 2 shots from the point for 1 shot from the slot? I would because a shot from the slot is more that twice as likely to result in a goal than a shot from the point. However, Corsi says that is incorrect because all shots are equal.

    One of the reasons I want Tanner Glass benched is because of the shots he takes. In the 1st period Sunday, he had possession of the puck along the half boards and empty ice between he and the net. Instead of moving toward the net to increase his chance of scoring, Glass took a shot from the boards. Corsi would give this play a thumbs up because it was a shot whereas I thought it was a hope and a prayer.

    • I understand what you’re trying to say, but the example you gave isn’t the best, since Tanner Glass is the worst Corsi player in the league.

      Shot quality hasn’t been proven to be repeatable, which is why the total shot attempts has been relied upon.

      It’s been proven that teams that take more shot attempts control possession and are more likely to make the playoffs.

      • Right. It’s an important distinction. Corsi/Fenwick/SAT% doesn’t attempt to measure shot quality. They are possession metrics. It gives an idea of which team controls the puck more than the other. To try to conclude other analyses from these stats would be incorrect. Generally speaking, teams that possess (control) the puck consistently win more games and make the playoffs almost as a matter of fact.

        • It’s not a generally speaking conclusion, it is fact that teams with higher possession rates make the playoffs and go further in the playoffs.

          The reason why shot quality has generally been ignored is because teams can’t really reproduce shot quality game to game or season to season.

          What that part doesn’t address is how systems drive specific shot types, like NYR and rush shots or LAK and suppressing rush shots.

  • Nice write-up. One thing that may have been missed is the power play. The Rangers don’t get too much power plays therefore they control the puck less. Your thoughts.

    • ^^ I could be wrong, but I think that all possession #s are taken from even strength, so they don’t take into account the PP and PK.

      However, one conclusion we can draw is that we draw fewer penalties than we would if we had higher possession numbers. The more we own the puck / control play, the more the other guys will obstruct, etc.

      • I took that to mean NYR don’t have the puck as much thus draw fewer penalties. I don’t believe there is a correlation there.

    • Ehhh, I don’t know if I buy that. Interesting train of thought, but speed also draws penalties by catching guys flat footed and forcing them to make stick plays.

  • I don’t see a concern. Basically, you have team +/- and possession stats. The advantage of team +/- is that it measures exactly the right thing. The disadvantage of team +/- is that so few goals are scored that small sample size error is a big problem. Possession stats measure not quite the right thing more accurately.

    The study someone here pointed me to said that, for a 20 game sample, possession stats painted a more accurate picture. Even for a 40 game sample. But for a 60 game sample, TEAM +/- gave just as good a picture. So, with as much of the season behind us as we have, the sample size error is smaller and the fact that the Rangers outscore their opposition means something other than luck is going on.

    The way the Rangers attack may be the secret.

    • I don’t know what study you’re talking about, can you link to it please?

      I’m surprised by those conclusions, because possession doesn’t really begin to tell the most accurate story until 70 games into the season, when we have enough of a sample size to see how teams are performing. Team +/-, I’m assuming you mean goal differential, does has some noise to it, since shootout wins get that extra goal (one example).

      There is certainly a correlation between the higher SH% and the type of shots the Rangers prefer to take. I don’t know if it’s a secret, but they may be this year’s litmus test.

  • Back to top button