Isles Analytics Live Stream #2: Isles First Line Issues

I had my second live streamed cast tonight, talking Isles Analytics and what it says was actually a problem with the Isles first line last year, and what could be a possible fix.  You can find the archive below:

Much happier with this second cast, as the stats are now visible on screen.  Let me know what you think!


Live Cast Monday 7/21 at 7:10 PM EST and Glossary is up!

Hey Everyone,

In case you’re checking the site, two updates:
First, today, Monday 7/21 at 7:10 PM, I’ll be having a live streamed cast talking the Isles and how things look with the analytics on my twitch account:  Hope to see you there and if you register an account you can ask me questions in chat – I’d hope to answer some of them.

Second, I’ve put a basic glossary of terms up on this site, please check it out if and if you have any further questions on the terms, feel free to comment.



On Kirill Kabanov and Asset Management.

The Islanders waived Kirill Kabanov yesterday, cutting ties to a guy once thought to be a top 10 talent amongst Isles prospects.  It appears that no one has claimed him – which isn’t really a surprise; whereas prospects with off-ice problems (in Kabanov’s case, reports of lazyness and missing meetings amongst other problems) OR prospects with high talent who never seem to put it together into results often get second chances from other organizations, prospects with BOTH problems are not exactly guys that teams really want to take chances on.  It is extraordinarily unlikely that Kirill Kabanov ever amounts to anything in professional hockey, nevertheless the NHL.

That said, the Isles cutting ties with Kabanov before his ELC is even up (buying out the last year of his deal) is a typical example of poor asset management that NHL teams exhibit all the time.   NHL teams have 50 spots for NHL Contracts (which don’t count sub age 20 prospects in juniors due to the slide rule).  Your NHL team only has room for 23 players (prior to the trade deadline, when it becomes unlimited).  That means you can have 27 players who aren’t in the NHL signed and playing in your AHL/ECHL affiliate.  These are spots for extra assets.

Note:  AHL players don’t need to be on your 50 contracts – many AHL players are on AHL/ECHL deals, and thus don’t count for the 50.  So you don’t need your extra 27 slots to fill up your AHL team.  And in fact, last year the Isles were at 43 contracts – they had 7 left to give. 

This year the Isles are at 48 contracts currently, with a 49th coming soon with Kevin Poulin’s arbitration hearing.  So Kabanov was likely cut for flexibility.  But think how those spots have been filled up:
Brett Gallant given another NHL deal (Qualifying offer)
Kevin Poulin qualified (going to arbitration)
Harry Zolniercyk given a 2 way deal (with a 300K AHL salary, a nearly 300% raise!)
Jack Skille getting a 2 way deal
Kael Mouillierat given a 2 way deal. 

Of those 5 guys, only Skille and Poulin have any business in potential on an NHL roster, and Poulin* is a beyond long shot goaltender who is 4th in line for NHL time (and that should’ve been the plan when they qualified him) and Skille is at best a fourth liner.  Zolniercyk is an AHL energy guy, Gallant is a goon, and Mall Rat wasn’t even good enough to get an NHL deal prior to this year.  None of these guys, bar Skille, are good enough to be worth a 50 man slot – you can get players of similar quality on AHL/ECHL deals – in fact that’s how the Isles previously had both Gallant and Mall Rat. 

*It’s possible they expected Poulin to follow Nilsson abroad and just QOed him to retain his rights, but Poulin didn’t have any connections overseas to make that seem likely.  A bad gamble, if that’s what they were going for.

Which brings us back to Kabanov.  Kabanov is a super long shot at this point, if any shot at all – but he’s a long shot at actually being a top 9er, not being a grinder.  Moreover, THEY WERE GOING TO BE PAYING HIM THIS YEAR ANYHOW – they saved like 30K by cutting him and are still paying most of his 3rd year of his ELC.  If he bothered the AHL coaches, they could’ve loaned him to Sweden or the ECHL again – and who knows, maybe he’d pull it together.  And if he didn’t the Isles could just not qualify him next year and no harm no foul.  Instead, the Isles have a bunch of mediocre AHL players on taking up spots of assets. 

This is what I mean by poor asset management.  And every team does it by the way – so it’s not just an Isles problem.  99.9999% of the time it won’t matter, but a team like the Isles, without great financial resources, shouldn’t be casting off any lottery tickets if they don’t have to.



Is Casey Cizikas really anything special?

The Isles announced the signing of two of their three remaining RFAs* today: Calvin de Haan was signed to a 3 year just under 6 Million dollar contract, and Casey Cizikas signed a 2 year 2 Million dollar deal.  Both of these contracts are reasonable for what the guys are:  de Haan’s 2 million per year is a pretty standard rate for a Bridge Contract given to an RFA in the NHL, and getting the third RFA year at that rate is a reasonable compromise for a guy with an injury history (instead of getting the first two years cheaper, the Isles get his long term contract started a year later, with him being an RFA after this deal – assuming he pans out again).  Cizikas’ one-way 1 Million per year rate is a pretty standard 4th line rate, and is only a slight raise over what he made last year (900K) although it’s no longer 2-way.  In short, neither deal should be considered a big deal either way, and both are pretty in-line with league prices.  Neither incidentally should commit the Isles long term to either player, although the Isles are more likely to be reluctant to cast off de Haan’s deal even if de Haan busts the next two years.

*Kevin Poulin is the last RFA given a QO and is awaiting his chance at arbitration.

That said, there’s a bunch of talk from Isles fans about how good deals both of these are.  I’m not going to talk about de Haan now (another time), but I’d like to talk about Casey Cizikas.  To a bunch of Isles fans, Cizikas is a pretty solid player – maybe even a pretty good one.  On twitter today, I mentioned that I didn’t see the need to add a second year to Cizikas’ deal, calling him a “mediocre 4th liner” and got a bit of pushback.  So it’s worth asking:  Is he really mediocre?

Looking at Cizikas’ fancystats (from the great


The Key #s here (I’ll refer to the other #s below, but these two are the ones you should make sure you understand – feel free to tweet or comment me with questions on the others!) are his CF% – which is his corsi – and his CF% rel – which is his relative corsi.  A quick explanation: 

Corsi%: % of all Shot Attempts (SOG+Missed Shots+Blocked Shots) taken by a player’s team while they’re on the ice on 5v5. A Corsi% of 50% means that each team is getting an equal amount of shots while a player is on the ice; above 50% means the team is outshooting the opponent, under 50% means they’re being outshot. Corsi has been found to approximate time of possession and scoring chance #s extremely well.

Relative Corsi%: Difference in Corsi% while a player is on the ice for a team and when he is off the ice. For example, a Relative Corsi% of 5% means that the team’s corsi with the player is 5 percentage points better with that player than without him – while a negative 5% means the team is 5 percentage points better in corsi with him on the bench. This measure allows us to separate a player’s performance from his teammates to a certain extent.

On CC, you basically have the picture here of a guy who in his first full season last year put up solid points for a 4th liner, but was losing the possession battle (taking into account zone starts only makes it slightly better) and this year his possession #s went off a cliff. When he was off the ice, the Isles were winning the possession battle and were spending more time in the O Zone than the D zone.  With him ON the ice, the Isles were greatly outshot instead.

Now here’s the thing:  4th liners are GOING to have negative relative corsi #s – relative corsi measures a player relative to his team, so naturally the 4th liners should be worse than the top 9 guys.  So how does CC compare relative to other 4th liners?   Well let’s look at this year and see similar forwards to CC. is the original hockey advanced stat resource (in fact the years NHL PBP allows us to obtain individual player fancystats is often referred to as the “Behind the Net” era) and while extraskater is most of what you need these days, it offers a level of customization that is currently unmatched.  In our case, we can use it to see how well Cizikas does in comparison to players with similar ice time.  I’ve generated a list of such players here. 

This list essentially includes forwards who played at least half the season last year, between 8 and 11 even strength minutes per game.  Those are your typical minutes given to 4th liners who aren’t face-punchers (Cizikas is just below 11 minutes per game, as is Matt Martin.  Colin McDonald misses the cutoff).  I then also removed players with extreme zone starts (players who started under 40 or over 60 percent of their starts in the offensive zone), who tend to be extreme defensive specialists (Manny Malhotra) or extremely sheltered players (Bryan Bickell) and thus aren’t comparable to Cizikas.  This left me with 54 players who fit the criteria, as you can see on that link.  In terms of relative corsi, or corsi relative to his teammates, Cizikas was 43rd of these 54 or fairly below average.*  Cizikas wasn’t much better in scoring per minute compared to these guys as well.  Even if you try adjusting for zone starts and competition, the season ranks toward the bottom of 4th liners, and remember our relative corsi metric isn’t exactly comparing him to a top 9 that ranked in the top of the league. 

*I’m not going deep into it because this post is already running way too long, but Cizikas does also play penalty kill.  Unfortunately, his PK results aren’t good either.

Now there are two elephants left in the room:  First, Cizikas’ teammate, Matt Martin, is also on the list BELOW Cizikas (Martin is 46th) and Colin McDonald would also be bad if he made the list.  How do we know Cizikas is the problem?  Well, we don’t – and in fact, I think it’s pretty clear from Martin’s play with other players, that he’s also a problem (more on that in a later post).  But Cizikas’ play without both of these guys – he played 240 minutes without Matt Martin – isn’t that much better than it was with Martin (See his #s with and without each linemates (WOWY #s) HERE). 

Second, Cizikas’ #s were better last year.  And in fact they wound up last year around average for a 4th liner.  So perhaps CC bounces back (he certainly will score a few more assists this year, as he had notoriously bad shooting luck from his linemates) a little.  But if his upside is “average 4th liner”, that’s not worthy of talking about much.  And the Isles do have FIVE other centers (although I’d rather Nelson and Strome play wing on the top 9), six if you count Bailey.  He may be 23, but there isn’t much growth likely here.

In short Cizikas isn’t anything special, and calling him mediocre is hardly wrong.  In fact, his #s compare unfavorably to one other Islander on the BTN list linked above: Peter Regin – who scored just a little less points but drastically was better at keeping the puck out of our own end.  And Regin was hated by fans and traded for peanuts (that wound up being traded for Halak).  But let’s stop talking about him as if he’s a key part of the lineup, who deserves to have 4C locked up.  If guys like Sundstrom or hell, even Halmo (who doesn’t usually play C), might be better, the Isles shouldn’t hesitate eating the 2M and switching over.  And late in games where the Isles are down, Cizikas should be getting as little ice time as possible.

Welcome to Islander Analtyics

Welcome to Islander Analytics! I’m garik16, and I’m a poster at Lighthouse Hockey and Hockey Graphs.  For a while now, I’ve been making posts at LHH in which I use hockey analytics – or advanced stats or fancy stats or whatever you want to call them – to take a different look at the Islanders.  I’ve attempted to write a primer on these stats on LHH (admittedly It’s only about 80% done) and have always attempted to explain them to readers such that they can use them themselves, if they so choose.

That said, it’s become apparent to me that while I like to think my posts have been helpful to Isles fans who are interested in analytics and learning what they say about the team (and about the rest of the league), that there isn’t really a resource where I can direct Isles fans to a broad spectrum of analytics-based posts – if you want to learn what they say about the Isles, you currently have to be linked to one of my posts, learn on your own, or google really well.

Hence this site.  From here on, I’m planning on cross-posting my LHH work over here, such that anyone interested in an analytics perspective on the Isles has one place to look.  In addition, I plan on bringing over shorter analytics-based posts over here, because usually I just explain on twitter and have people wind up misunderstanding due to the limits of 140 characters.

In time, I’d hope other Isles fans might want to host some analytics posts on this blog, such that it becomes one place for any Isles fan to go to for analytics, as opposed to just my work.  But for now, welcome.  Let’s see what we can learn together about this team as we enter a new era.