2017-18 NJ Devils Shootout Breakdown Part 1: The Goaltenders

Earlier this summer, I wrote about how shootouts went fairly well for the New Jersey Devils last season but remain a challenge going into next season. The Devils went 5-4 in shootouts but outside of Drew Stafford, no one scored more than three shootout goals and the only Devil on the roster with a career shooting percentage above 30% is Brian Boyle. I took that as a means to look more closely at shootouts. To do that, I broke down Jesper Bratt’s amazing shootout goal against Mike Condon from last season. It was a great looking play, it was an important goal as it put the Devils up 2-1 in the shootout that they eventually won, and it showed to me how shootouts could be analyzed. I wrote the post as a proof of concept to look at every single shootout attempt for and against the New Jersey Devils last season through NHL.tv. I have completed that task and now I will give you what I found and learned from the experience.

I will reiterate that shootouts matter in the regular season. You may not like them. You do not have to like them. It does not matter. They count. The 2017-18 New Jersey Devils made the playoffs last season on the next-to-last game of the season and ultimately finished one point ahead of Florida. The Devils went 5-4 in shootouts last season. Dropping another shootout would have meant missing the playoffs as Florida had more ROW as a tie-breaker. On the flipside, assuming all other results held, winning one more shootout would have put them past Columbus and given them a matchup with Washington over Tampa Bay and winning two would have put them ahead of Philly. Shootouts mattered in 2017-18. They very well could matter again if the Metropolitan Division and the Wild Card race in the East are tight in the standings again. Therefore, there is value to learning more about what happened and what could be improved for future shootouts.

I am going to start this breakdown with the goaltenders. They cannot score goals to win shootouts, but they can keep the team in them and secure a victory when a lead is obtained. They’re important enough to keep separate from the shooters. Out of the nine shootouts the Devils were in last season, Keith Kinkaid was in net for five of them and Cory Schneider was in net for four. Nine does not seem like a lot, but it was on the upper end of total number of shootouts in the NHL last season. The split of appearances cannot be even and neither was the workload. Two of Kinkaid’s shootouts were short, and one of Schneider’s went extra-long with eleven rounds. As a result, Schneider ended up facing more attempts than Kinkaid. Neither goalie saw the same team twice or saw an opponent the other goalie faced, so there were no common shooters in shootout situations. Yet, we cannot avoid a comparison as the two goalies are set to be the main tandem once again 2018-19. Let’s do it.

Terms

There are a few terms I will be using in this part and future parts that are not so common, but I wanted to clarify them up-front.

First, when I say a shooter makes a move or a goalie takes a move, I am referring to a discernable move intended to try and beat the goalie. Since I used NHL.tv, I did this with the MSG broadcast footage of the Devils games. I cannot see subtle moves like eyes or face placement. I do not mean the common backhand-forehand stickhandling players use when they skate up ice to be a particular move. I am referring to things like sudden dekes, cut-backs, toe-drags, stick-drags, fake shots, shoulder shimmies, and head fakes. Those are moves. A goalie taking the move means they react to it and in that direction. That does not mean they are beaten as goalies are quick and large enough to recover or throw out a limb in desperation to make a save.

Second, I broke down attempts into misses, posts, saves, goals, and lost pucks. A missed shot is, well, missing the net. It is a clean miss, meaning the goalie did not get a piece of it. A post is what happens when the shooter hits the post. The goalie is beaten - the puck got past - but the metal frame denied a goal. Lost pucks are where the shooter just loses the puck either on a move and/or in motion. There was no shot taken - the puck just went away. Only once did a Devils goalie influence that. A save is where the goalie stops a shot and a goal is when the puck goes in the net. I’m stating this here to establish the goaltender’s actual save percentage.

Lastly, a failed attempt is typically where the shooter loses the puck or misses the net so widely or highly that they were nowhere near close to scoring.

In-Game Performances vs. Shootouts

Alex wrote about how the Devils were in a lot of close games last season. Obviously, the game has to be in a tie for 65 minutes for it to go to a shootout. However, they were not in low-scoring games. The ones that went to shootouts were typically high-scoring affairs. All nine involved either four, six, or eight total goals in the game. So the goaltending as a whole was not on-point. At least the Devils were scoring goals to keep pace?

In-Game Performance stats from NHL.com

I bolded the games where the Devils’ goalie put up an in-game save percentage above 90%. Kinkaid did it twice, however, he also had the lowest in-game save percentages in these nine games. Schneider was within a percent of 90% in his other three games, so he wasn’t that far off. The larger point is that outside of the bolded games, neither Schneider or Kinkaid had the greatest of nights. It does not seem that had anything to do with their performance in the shootout as both Kinkaid and Schneider had some good shootouts after some not-so-good games. Kinkaid did have a flop of a shootout after an otherwise solid game January 2; although St. Louis crushed all three of their shootouts last season (6-for-7 shooting!!). While more data would be ideal, I’m skeptical that a good or a bad in-game performance will necessarily lead to a good or a bad shootout performance.

Shootout Performances or When a Save Percentage Isn’t an Actual Save Percentage

This link goes to the shootout summary for both Kinkaid and Schneider at NHL.com. You will see that Kinkaid made 11 saves out of 17 attempts and Schneider made 19 saves out of 23 attempts. From that, they have save percentages of 64.7% and 82.6%, respectively. That should be enough to determine who made more saves and that’s usually the extent of any shootout analysis for goalies.

The problem is that these percentages are not accurate. In counting misses, posts, and lost pucks separately from saves, I found the following:

NHL.com

Last season, Kinkaid did not make 11 saves and Schneider did not make 19 saves. The scorer and/or NHL.com give credit to the goaltender for making a save when the player does not score a goal. Even though the Gamecenter at NHL.com may list an attempt as a “miss,” it’s counted as a save. Even though you clearly saw and heard a post being hit on a shot, it’s counted as a save. Even if the shooter loses the puck on a deke and therefore makes no shot on net, it’s counted as a save. In other words, the save percentages and save counts under Shootout Summary at NHL.com are only reflective of whether a goal was scored or not. It does not mean that the goalie actually stopped a shot.

To that end, I calculated the actual save percentage of both goalies. When you take out the posts, misses, and lost pucks, they go down. I did give one extra stop to Kinkaid as he did force one of those lost pucks. In the last shootout attempt of the season against the Devils, Craig Smith went in on Kinkaid with plenty of stickhandling dekes, went to the top of the crease, and ran out of room as Kinkaid went down. The shaft of Kinkaid’s stick knocked the puck away. There was no shot and I was not certain Kinkaid intended to do that, so I tagged it as a lost puck. You could argue it was a save, which would bump that actual save percentage to 60%. Either way, it does not make Kinkaid look that much better.

The good news is that despite how save percentages are calculated, the larger conclusion remains: Schneider stopped more pucks and at a higher rate than Kinkaid. Kinkaid gave up more goals, too. It could be argued that Schneider was a little luckier. No post saved Kinkaid from conceding a goal; whereas it helped Schneider twice, both in that super-long shootout against Boston. Those two plus one lost puck on a deke (Brayden Point, in the first shootout attempt against the Devils last season) meant Schneider was beaten three times without getting scored on. That only happened to Kinkaid once.

However, this is not to say Kinkaid had no fortune. Kinkaid did have two total failures by the shooter in that the shooter did not get a shot off at all. And his one “goalie beat, didn’t finish” result was thanks to Erik Karlsson narrowly missing the top right corner after Kinkaid bit hard on Karlsson’s fake shot.

Speaking of those aspects, let’s go into moves. I did not summarize all the different moves each goalie faced because, well, there were a lot of different ones. A popular one was to make one deke prior to a shot, usually at close range to the goalie in the hopes of getting past them. Like showing forehand towards the center and then cutting the puck back to a backhand towards the left post (this is something the Devils themselves also did, look for that in Part 2). But there were multiple variations involving direction, timing, and even whether they were successful. It is granular enough that a listing does not say too much.

That was only in the cases where an easily identifiable move was even made. Plenty of shooters did not make a clear-cut move; choosing to approach the goalie and fire a quick shot in the hopes of beating them based on what they saw. It helped them that the goalie - Kinkaid, Schneider, or the Devils’ opponent - would be active in their own right. It was common to see the goalie come out of their crease and move back (sometimes) when the shooter approached. It was also common to see goalies start to go into a stance (like dropping into a butterfly) or move a part of their equipment; all to possibly to bait the shooter into shooting in a certain way. When I started watching hockey as a kid, I thought goalies were meant to be patient and not make any move first in a one-on-one situation unless they want to be beaten or they were supremely confident. So I’ve had to wrap my head around that.

Kinkaid saw more moves against him than Schneider, both in total and as a percentage of total attempts against. In his defense, Schneider did have a whole bunch of Bruins take shootout attempts that they normally would not in that 11-round epic. Five guys in a row opted to skate in and try to beat Schneider on a shot. Two of them hit a post, so it was not a bad idea.

What does look bad are the results of the goals. Schneider managed to only react to six out of nine moves he saw and was beaten twice when he took those moves. For the plays where Schneider did not take a move from the shooter and stuck to what he wanted to do, he was only beaten twice. Despite not taking as many moves and only being beaten once when he did, Kinkaid conceded five goals without taking one. Of those five goals, four of them were just well-placed shots that beat him glove side that made him look like a shooter tutor. The fifth was Kyle Turris flicking his stick in the slot - which Kinkaid did not react to, so he didn’t take that move - shortly before releasing a quick one through his legs. I will give Kinkaid credit for not consistently biting on the shooter’s moves - three out of ten is better than six out of nine. That said, if a shooter has space to shoot at, then he just needs to shoot it quickly and well to score - and that happened to Kinkaid more than it happened to Schneider.

The difference is also reflected in the shot locations. This is relative to the shooter, so the right is the shooter’s right. For the attempts where the puck was lost, I was able to figure out what the shooter was trying for:

Shooters attempted to go glove-side or right-side more than the other sides on both Kinkaid and Schneider. It worked more against Kinkaid than Schneider, especially for that top right corner. Let’s break that down.

For Schneider, out of the five high right shots, one hit the post and two missed the net. He had some luck there. One save was a glove stop without catching the puck and one was a shoulder save. For Kinkaid, three people scored, two missed with Karlsson missing the open corner on his fake shot attempt, and Kinkaid made one glove save. Those three people were Sean Monahan, Matthew Tkachuk, and Vladimir Tarasenko; two of those are among the best shooters in the game today. Not that Kucherov and Eberle are scrubs, but getting beaten straight up by Monahan and especially Tarasenko is not a huge surprise either. That said, Monahan, Tkachuk, and Tarasenko saw the top corner and a Kinkaid looking seemingly frozen with the flicked the puck past them. While Kinkaid was clearly trying to be patient against good shooters,

Let’s go down in height. The medium right shots tend to be those past the glove or between the glove and the pad. Brayden Schenn torched Kinkaid for that one goal, right under the glove. It was similar to what Tarasenko would do after that attempt; just lower. At least Kinkaid denied Filip Forsberg on the other try. Schneider saw it more often and just made more saves; two where he caught it with his glove and one where he got his chest on it. Only Mika Zibanejad beat him under the glove and that was off a deke that Schneider didn’t take. Still, it’s clear Schneider was more effective for these shots.

At least the lower attempts - about pad level to the ice - were better with a lean towards Kinkaid. Kinkaid saw four of those, stopped three and benefited from Kevin Fiala losing the puck on a deke. Schneider was turned out by David Pastrnak on a move, but stopped the other two - which were all in that Boston game. The legs were working relatively well.

Thanks to observation of each play, it is not as jarring it appears. Kinkaid flopped in two of his five shootouts as he was beaten on all but one attempt between the Calgary game on November 5 and the St. Louis game on January 2. That crushed his numbers as Monahan, Tkachuk, Schenn, and Tarasenko made Kinkaid look like a statue on their attempts. But it is not as if the league suddenly found the goods on how to beat each goalie. Kinkaid and Schneider were not beaten more than once anywhere else. Perhaps Kinkaid could stand to work on his glove hand, particularly against left-handed shooters (all four skaters are left-handed shooters). And he could certainly learn to be more consistent when the shooter does approach. But he is not as catastrophic as a 50-60% save percentage may seem in shootout situations.

One final note, Schneider did have to stop a few more five-hole attempts than Kinkaid. I was pleased to see that almost all of them were stopped. Alas, the only one that did was a Brock Nelson shot that trickled through - which unfortunately also decided that shootout on January 7 against the Isles. Perhaps shooters should not try to catch him low so much? Or perhaps I should encourage it knowing that Schneider can turn those away.

General Observations & Thoughts for 2018-19

You know what I did not see from either Kinkaid or Schneider? Poke checks. Kinkaid came the closest when Smith saw the puck knocked away, but it was not clear as to whether it was intentional. It was with the shaft of his stick and it was with Smith amid his business while he was on the ice. That’s definitely not your traditional poke check from a goalie. It was refreshing to realize that, especially since the legendary goaltender of this organization was no stranger to attacking the shooter with one. When it worked, it was great. When it didn’t, the goalie was in a spot to be beaten. Both Kinkaid and Schneider had some attempts where they were aggressive in their positioning, but nothing that as far as that.

That said, I am not so worried about the goaltending when it comes to shootouts in 2018-19. Schneider seemingly has it down. A save percentage - actually based on shots or what NHL.com reports - in the high 70s to 80s is great for shootout situations. While opposing shooters tried to glove side or five-hole on Schneider, he was not all that vulnerable there. He had some post-luck, but the majority of attempts were legitimately saved. Kinkaid could do better, but his poor percentage is driven by getting torched in Calgary and a St. Louis team that made their three goalie opponents suffer badly in shootouts. That said, Kinkaid was a bit all over the place prior to shooters shooting, whether it staying standing up or going down at the first sign of trouble. His fundamentals are generally solid in-game; so some tweaking may be appropriate - particularly for that glove hand since shooters tended to go to his left (their right) more often than not. I would not expect massive changes since Kinkaid is 29, has over 100 NHL appearances, and is entering his fourth season entirely in the NHL. But some small adjustments could help him and, by extension, the team.

Outside of the Devils, I would love it if the NHL or somebody would count saves, misses, posts, and lost pucks separately for an actual save percentage. In the 65 minutes played before a shootout, a missed shot is not counted as a save. Neither is a post or no shot at all. I do not see why that should change. Either the people of NHL.com should call it something other than save percentage or provide an accurate one. Also, just watching 40 shootout attempts against the Devils gives me a new appreciation for the shootout. Not only do they count in the standings, being able to beat a goalie one-on-one is an achievement - especially after playing for the previous 65 minutes. It’s not that easy and there’s a lot more to it than making a move and/or taking a quick shot. I admit that I do not know enough about goaltending to understand the subtleties that could also explain why the Devils’ goaltending was what it was; but there could be something there for someone is both knowledgeable and able to break down tape.

Your Take

I focused on the goalies today. Tomorrow, it’ll be all about the shooters. In the meantime, I’d like to know what you think. What did you learn about Kinkaid and Schneider in shootouts last season? Are you surprised to learn that the NHL.com Sv% for shootouts includes non-saves? Please leave your answers and your thoughts about the goalies in the shootout in the comments. Thank you for reading and not commenting about how you don’t like shootouts.