As longtime readers know, I'm a proponent of using stats in many of my posts. I've delved into advanced stats to get a better handle on what's going out there on the ice beyond how many shots or how much ice time one has. Values like Corsi, which is the differential of all shooting attempts for and against a team when a player was on the ice at even strength, corresponds well to possession. Especially over a larger population of games. Being on the ice for more attempts against your team suggests that you got pinned back more. Being on the ice for more attempts for your team suggests you've done something right.
Vic Ferrari's Time on Ice has a number of scripts that generate many of the stats on a game by game basis and Gabe Desjarden's Behind the Net compiles all kinds of rate stats over a season. Behind the Net is now up and running for the 2013 season, giving everyone a chance to get ironically excited over three games' worth of stats. Ferrari is apparently working on updating the scripts for this season, so that's why you haven't seen the Corsi charts or head to head ice time charts I usually link in my recaps. Today, I'd like to make up for that with Corsi charts from the first three games by the New Jersey Devils this season.
I owe a lot of thanks to Robert P. of Jewels from the Crown. He came up with an Excel template where a user can filter out all Corsi events - shots, goals, misses, and blocks at even strength - from a game's official play by play log and counted both for and against a player. It takes a little work, but it's not terribly difficult to get the desired results. Robert gave me this template and helped me out a bit with dealing the code and so I have charts from all three games. Again, these are all shooting attempts at even strength, non-empty net situations. The Devils haven't had to deal a lot with empty net situations and the last two games featured quite a lot of special teams play. Regardless, it's the most common situation in hockey and if we want the Devils to continue to get results, then we must see whether that's reasonable based on their even strength play.
Game 1: Devils 2-1 Win At Islanders
Team Corsi Results: New Jersey led 45-31. The Isles came the closest in the second period, but the Devils led then too at 18-16.
Devils Players Corsi Chart: This chart and the next one for the Islanders does not include any 5-on-6 events.
Notes: For this chart and the other Devils' charts, I sorted the players out by position. I did this to make the defense stand out. Adam Larsson has yet to play a NHL game this season. If he's going to play, then someone has to sit, so organizing the players like so makes it easier to identify who has and hasn't been doing so well on defense. I've complained about how soft they've been in their own end at times, but these numbers should show how serious that has been at least at even strength.
Against the Islanders, mostly everyone did well. Bryce Salvador and Marek Zidlicky saw the most shooting attempts against at evens, but Salvador was present for more productive Devils shifts than Zidlicky. Henrik Tallinder was also right up there in terms of Corsi events against along with a low differential. Tallinder played about four minutes less than Zidlicky and Salvador at evens.
This chart also shows how great the line of Mattias Tedenby, Patrik Elias, and David Clarkson did in the season opener. Corsi provides further proof of Tedenby having a good game as well as a little evidence that Clarkson can perform on the second line. Without knowing the match-ups, it's hard to say whether they just beat up on the weak depth of the Isles or did well against top players. Either way, that's all good.
Only three forwards ended up negative: Stephen Gionta and Ryan Carter ended up at -1, which is in line with what I saw that night. A line that did some good things, got pinned back a few times, and ultimately it all washed out. Curiously, their usual right winger Steve Bernier ended up at +4. The third negative forward was Ilya Kovalchuk at -3. We know that Cam Janssen was benched early and Stefan Matteau's time was limited, so Kovalchuk took plenty of extra shifts. While I don't know how to filter out events with his usual line from his extra shifts, a good rule of thumb is to look at Travis Zajac's Corsi. Since Zajac was a +4, I'm led to believe Kovalchuk did well with him and suffered when he played with other forwards. That also fell in line with what I saw, some of the few Isles attacks featured two members of the bottom six and #17. It didn't hurt the team on the scoreboard, but it didn't lead to much in the way of possession.
Isles Players Corsi Chart: Mark Streit did whatever he could but he couldn't hold back waves of offense.
Game 2: Devils 3-0 Win Against Flyers
Team Corsi by Team: Flyers led 33-25. Given that the Flyers had to go against a lead for the majority of the game, this shouldn't be so surprising. However, the Flyers crushed the Devils 14-5 in Corsi in the first period when it was still a close game. That's quite surprising and just plain bad.
Devils Players Corsi Chart: Note, these charts do include some 4-on-4 events since that was at even strength. There was only one, though: a shot by Josefson.
Notes: Well, the defense mostly played in their own end at evens. Mark Fayne did the best at +5 and considering he played 15:33, the second most even strength minutes against Philly, that's quite good. Bryce Salvador just got above positive. Everyone else, well, not so well. Zidlicky, Anton Volchenkov, and Henrik Tallinder got beaten and didn't do much on offense. Zidlicky stands out the worst in my eyes since he also took three penalties, two at evens and one on a 5-on-3.
Speaking of offense, only Dainius Zubrus, Zajac, andKovalchuk ended up positive in possession. Considering the Flyers were chasing the game, that's pretty good. The Elias line didn't do so great but they generated nearly as many Corsi events for the Devils. The rest, uh, not so good. The CBGB line was straight up beaten. Cam Janssen's cameo was all kinds of wrong. Matteau and Josefson both got rewarded with more minutes in this game so they got to generate some more chances than the rest. Even they finished up on the wrong side of zero.
Mattias Tedenby truly was awful against Philly. No wonder he was benched. It's one thing for Ryan Carter to end up at -5, he's a fourth liner playing as a third liner. It's another when you're playing alongside Elias and Zubrus and you're supposed to be an offensive winger.
Flyers Players Corsi Chart: Claude Giroux did better than I thought, though Scott Hartnell and Jakub Voracek actually took all the shots. Also, Kimmo Timonen did a lot better than I thought.
Game 3: Devils 3-2 OT Win Against Capitals
Recap: My recap of the win was here. I thought the Devils didn't play a smart game. As great as it was to get the win in overtime, it was disappointing that they blew a 2-0 lead in third period to the Washington Capitals.
Team Corsi by Team: New Jersey led 33-28. Overtime was 4-1, so in regulation the Corsi events were closer at 28-27, New Jersey. The Devils only fell behind 9-10 in the first and did quite well in the second at 11-5. They were down 9-12 in the third, which isn't that bad. Although that period didn't have a lot of even strength time in it so the Caps made more of their limited time while down in the game.
Devils Players Corsi Chart: As with the Flyers game, this chart and the Capitals' chart does include 4-on-4 play such as in overtime.
Notes: Fayne went from having a good game against Philly to not so strong against Washington. The rest of the defense was surprisingly on the positive side of Corsi. Marek Zidlicky finished at +2 though he led in both Corsi for and against events. It's evidence he did better than his bad night against Philly. I want to say his presence in overtime helped get him into positive territory. Volchenkov didn't even play ten minutes at evens, which is why his counts were so low. But since he did so much on the PK, I still come away from this chart thinking Tallinder could become the odd man out.
Speaking of low, Mattias Tedenby got to play with Zajac and Kovalchuk for shifts in the first and second period and it went poorly. That trio did very little on offense together. Once Tedenby was benched, Zajac and Kovalchuk were able to be more productive. That said, those two still were out there for a lot of attacks by the Capitals. I will say that the extra shifts may have helped Kovalchuk's possession given Zajac's -5. Going back to Tedenby, I'm astonished at his one Corsi for event and five against. Even Krystofer Barch was on the ice for three Corsi for events and he played 3:28 last night! Between this and the Philadelphia game, I think Tedenby may only be effective against soft competition. In other words, he needs to be removed from the top-six.
They could replace his spot with Jacob Josefson. I liked a lot of what I saw from Josefson last night and finishing +1 while moving around in the lineup is pretty good. I should have appreciated the Zubrus-Elias-Clarkson line more as they came out positive. Moving on to other forwards, I'm a little surprised Matteau only got a little under six minutes given that he did pretty well. I guess DeBoer saw something he didn't want to expose him to? I'm a bit more surprised the third line didn't do as poorly as I thought given that two thirds of the line didn't play more than ten minutes. I'm starting to think DeBoer should have used that line a little more. It would have given a couple more breaks for Zajac (15:54 at evens) and Kovalchuk (18:57 at evens).
Caps Players Corsi Chart: It is here that I started thinking that maybe John Carlson wasn't so bad and I was surprised Karl Alzner was pinned back that much. To be fair, I didn't recall Alzner doing that much in the game. Incidentally, I wish the Devils had a guy like Joel Ward. He did well against NJ and he's the sort of player that would do well in a third line role.
With this information we can have a clearer look at how the team is performing. Three games isn't enough to come to any wholesale conclusions about various players, but the shooting attempt differentials do give us an idea of where the team stands. Here are a couple that I have come up with. Overall, the team's not that bad at possession, but they're not all that good either - they're somewhere in the middle. Also, the Devils haven't played from a deficit so far this season so we don't know whether they will take more possession as scoring effects suggest in general. Unless Mattias Tedenby is going to play against weak players or get much better really soon, he's likely going to be a liability in terms of possession. Ilya Kovalchuk has played a ton of minutes so far and therefore has been present for a lot of events. Yet, the extra minutes away from Travis Zajac has been a detriment to his possession so he ends up negative when he may be positive on the top line. I was surprised to see Bryce Salvador positive in possession in all three games, clearly he's been put out there a lot with the Devils' top forwards. In considering Corsi differential alone, it appears that Peter DeBoer should consider replacing Henrik Tallinder with Adam Larsson for a game. He should more strongly consider getting twelve forwards who can play for three periods, too. Outside of the top two lines, no forward has really done well for a significant amount of ice time save for the CBGB unit against the Caps, which was still less than ten minutes of ice time.
Again, I want to thank Robert P of Jewels from the Crown for the template and the assistance in using it. What have you noticed from these charts? What conclusions have you come to about the Devils? What changes do you think they should make to get more possession at evens? Would you like to see these charts in the future until Time On Ice scripts are active for the 2013 season? Please leave your answers and other thoughts about the Devils at Corsi in one or all three wins this season. Thank you for reading.