Visualizing Corsi and the Ottawa Senators

Recently, there was a great post over at Arctic Ice Hockey by Tim Bonnar on the Jets struggles. There are a surprising number of similarities between the two teams, mainly stemming from management, coaching staff, and some media suggesting that the team needs more from their best players. Bonnar suggested that the Jets problem is a lack of talent - whether that's not utilizing their current talent properly, coaching strategy, etc. From the Sens perspective, I don't think the problem has to do with production from their best players, seeing that Ryan, Turris, Spezza, and MacArthur are all on pace for 60+ points. However, there have been some worrying signs lately regarding certain players on this team when we look at shot-differential statistics. This, combined with the fact that: 1) the Sens are the healthiest team in the league and 2) have the 5th easiest schedule (according to Sagarin's method) leads to some questions about what has really been the problem ailing the Senators?

Now, I think this is a combination of a couple things, but in this post, we'll look at lineup utilization and some unexpected drop-offs in play, rather than a narrative-based explanation like "lack of compete level" and "they don't have the will to win", or other concerns like goaltending. By my indication, it looks (from the outside at least) that the Sens don't know WHO their best players are, and if they do, they certainly aren't allocating ice time (reward for players playing well) appropriately.

The Chart: After reading the Bonnar piece on the Jets, Tyler Dellow and Cam Charron looked at the Jets and Cauncks in terms of team Corsi rates when players are on the ice with or without others.

It may look confusing at first, but hopefully after a walkthrough, it'll seem straightforward. The left-most column are the Sens forwards, sorted by 5-on-5 TOI/60 going from most-used to least-used, and the top-most row includes the Sens defencemen, sorted the same way (except left to right).

The numbers are the Corsi percentages of the forward with the corresponding defencemen, and are colour-coded. Red is when Corsi % is above 55%, orange for when it's between 50.0-54.9, or blue when below 50%. Corsi percentage is the rate of all attempted shots that were taken by the player's team when that player was on the ice, and a higher Corsi % (red, orange) means that you are driving possession, whereas a lower Corsi % (blue) means that you aren't.

For example, if Milan Michalek was on the ice for 10 shots for and 15 shots against, his Corsi percentage would be 10/(10 + 15) = 40.0%. Corsi has been shown to correlate with scoring chances, and time on attack, and is used as a proxy for puck-possession.

To add more context to this data, here's a table with each player and their individual Corsi %. Included here is Corsi Relative (Rel) %, which is the Corsi For % relative to the team's Corsi For % when that player is not on the ice.


Corsi %

Rel Corsi %

Mika Zibanejad



Clarke MacArthur



Derek Grant



Kyle Turris



Bobby Ryan



Cory Conacher



Erik Condra



Chris Neil



Jason Spezza



Colin Greening



Zack Smith



Milan Michalek




Corsi %

Corsi Rel %

Erik Karlsson



Patrick Wiercioch



Marc Methot



Eric Gryba



Chris Phillips



Jared Cowen



(Note: the with-or-without-you stats were found at Hockey Analysis, and the TOI/60 stats + Corsi Rel. numbers were found at Extra Skater - both are amazing resources so definitely check them out if you're more interested!)

Example: Now, let's give a concrete example of the type of information we can pull from this data. Mika Zibanejad ranks 9/12 in the Sens forward group with respect to TOI/60. When he's on the ice with Erik Karlsson, the two have a 57.1% Corsi%, so they're generating approximately 57.1 of every 100 shots (or 5.71 of every 10). This number is higher than both Zibanejad's individual Corsi % (bottom table, 55.9%) and Karlsson's (53.0%), so this seems like a good combination to have on the ice together. In contrast, Zibanejad has a 50.0% Corsi% with Eric Gryba, which is lower than both their individual Corsi% numbers (55.9% for Zibanejad, 50.8% for Gryba). Considering Mika has better numbers with every other defenceman, it seems wise to continue to play him with those players instead.

Now, keep in mind that this is still a very small sample size, as some of these duos haven't even played 60+ minutes together, but it's still a good representation for what we're here to look at. In his piece, Dellow proposes his theory about lineup optimization using the chart above:

If we assume that the coach uses his best players the most, we would expect that the Corsi% for when the most heavily used defenceman and the most heavily used forward are on the ice to be the highest on the team, barring something really unusual - a defenceman who plays exclusively in the defensive zone or something... If my theory's right, things should get worse as we move from more frequently used to less frequently used players.

Both Dellow's look at the Jets, and Charron's look at the Canucks, fit well with this theory. However, if you compare both of these charts against the one I created for the Ottawa Senators, there's a marked difference. What does this mean?




To me, this looks like MacLean isn't allocating his TOI appropriately. For example, if you look down Jared Cowen's column, you can see that he has sub-50% possession numbers (blue) with everyone except when he's on the ice with MacArthur, Turris, Grant, Zibanejad, or Conacher - all in the upper echelon of Ottawa's best possession forwards, which tells me that it's the forwards, not Cowen, driving the bus in that scenario. Patrick Wiercioch, who was Cowen's press-box buddy earlier in the year, looks to have rebounded nicely, and has posted positive possession numbers with all the forwards except Spezza, and Michalek - both of whom are undergoing some trouble of their own. I'd argue that Wiercioch and Phillips, who's playing too many minutes right now, should switch spots, in order to let Phillips settle down in a more comfortable role in the 3rd pairing, where he has had success before.

Spezza, Michalek, and Greening all have more ice-time than Zibanejad and Conacher, with Mika joining Turris and MacArthur as the only 3 forwards to have positive possession numbers with every defenceman on the team, and Cory only sub-50% with Chris Phillips on the ice. Judging by these numbers, Zibanejad and Conacher should be getting significantly more ice-time than some of the other forwards on the team, and at least in Mika's case, it seems like MacLean has started to use him more in the top-6, especially after comments by GM Bryan Murray stating the very same thing. I'd personally play both of them with one of the team anchors in Jason Spezza, who has had an unexpected drop-off in play where he's gone from being a 53.3% Corsi player, to a 49.8%. MacLean has recently tried to insulate him with two of the team's best defensive forwards in MacArthur, and Condra, but I wouldn't mind giving Conacher + Zibanejad more responsibility alongside Spezza for an extended period of time.

Thus, unlike the case in Winnipeg or Vancouver, the "depth"/"support" forwards in terms of ice-time and defense don't seem like the problem for the Senators. Instead, some of forwards with lower ice-time (Zibanejad, Conacher) have been great possession players, and warrant an upgrade in ice-time. In contrast, Greening, Smith, and Michalek seem to be getting a bit too much ice-time, and are only been kept afloat by Erik Karlsson (who is great with everybody). Granted, those players have also played well with other puck-movers in Corvo and Wiercioch (maybe highlighting that their main struggle has been in getting the puck out of the zone), and could still be paired up with the latter two defencemen in lesser minutes.

Thanks for reading!

This FanPost was written by a member of the Silver Seven community, and does not necessarily reflect the beliefs or opinions of the site managers, editors, or Sports Blogs Nation, Inc.