Sunday, November 23, 2014

Weaver and Gilbert; by the numbers

This post will use data-generated scouting to quantify and compare this season's performance of Tom Gilbert and Mike Weaver. The recent arrival of 2 veteran defensemen into the Montreal Canadiens fold could point to the possibility of a veteran d-man not seeing everyday action on the Habs blueline. The eventual probability then exist that the Canadiens could sit one of either Mike Weaver or Tom Gilbert.

The numbers included here represent each player's performance at even-strength this season. All metrics are generated from the data I track during games. A list of all the events currently tracked can be found here.

Neither player is there to produce offense at even-strength. That said, it's important to note that Weaver helps produce 6.0 scoring-chances per-60, while Gilbert has helped produce 3.0. In terms of even-strength chances-against Weaver has been directly responsible for 11.4 per-60, while Gilbert has been responsible for 8.4.

Both players start a similar percentage of their shifts in the offensive-zone. As Weaver's OZstart% is 41.1%, while Gilbert's is 40.9. Weaver on the other hand, has faced a better quality of competition during his shifts.

One of the "big picture" numbers used to quantify a player's performance on the ice is their rating. A rating simply represents how many more successful plays than failed plays a player produces per-minute of even-strength ice-time. Gilbert has the higher even-strength rating at 2.09 compared to Weaver's 1.93. Another overall "big picture" number is ratio. This number represents how many successful plays a player makes for every 1 failed play. These numbers also show an advantage to Gilbert, as he has an ES ratio of 2.90, while Weaver's is 2.5. As mentioned, neither player is there for his offensive-zone play. That said, Weaver has the slightly better offensive-zone rating at 0.22 to Gilbert's 0.21.

Breaking down each player's rating by 10-game bands tell us that Gilbert has been improving as the season has progressed while Weaver's performance has declined. Gilbert produced a rating of 2.01 through the first 10 games of the season while Weaver's was 2.15. Over the next 10 games Gilbert's rating improved to 2.19, while Weaver's fell to 1.70. That said, Weaver's recent injury could explain the drop in performances.





The most significant big-picture numbers for defensemen involve the number of successful plays they make in the defensive-zone for every 1 failed play. This defensive-zone ratio shows Gilbert as having a substantial advantage, as he has produced 2.86 successful plays in the defensive-zone for every 1 failed play, while Weaver has produced 2.34. Over the last 10 games the advantage is even bigger, as Gilbert's d-zone ratio over that period was 2.84, and Weaver's only 2.04. As a point of reference, Josh Gorges defensive-zone ratio last season was 3.38.

Gilbert's defensive-zone puck-possession success-rate is 71.7%, while Weaver's is 68.2%. This tells us that Gilbert turns the puck over in the defensive-zone 28.3% of the time he has possession of the puck, while Weaver turns it over 31.8% of the time. That said, breaking those d-zone possession numbers down further we see that Weaver has an advantage when attempting passing plays with possession that move the puck out of the defensive-zone, as he has been successful with 70.8% of his outlet passes, while Gilbert has been successful with 65.6%. In terms of stretch passes, Weaver has been successful with 69.2%, while Gilbert has completed only 58.8%. It's d-to-d passes that Gilbert excels at; completing them at a rate of 94.2%, compared to Weaver's completion rate of 78.2%. That said, d-to-d passes are not possession-drivers, as they do not directly move the puck out of the defensive-zone.

Defensive-zone puck-possession tendencies tell us how often each player has attempted specific plays while in possession of the puck in the defensive-zone. Defensive-zone puck-possession tendencies are broken down into outlet passes, d-to-d passes, stretch passes, dekes along the wall, open-ice dekes, dump-outs, and dump-outs off the boards or glass. Gilbert has attempted a d-to-d pass in the defensive-zone 27.6% of the time he has had possession of the puck in the defensive-zone, while Weaver has attempted that pass 26.7% of the time. Gilbert's propensity for making d-to-d passes combined with his above-average success-rate is important to mention and quantify, as a d-to-d pass is often the first play made by a player recovering a puck off of a d-zone face-off. I mention this because only PK Subban has more loose-puck recoveries off of defensive-zone face-offs than Gilbert.

Gilbert has attempted an outlet pass 44.5% of the time he's had possession, while Weaver has attempted that pass 39.9% of the time. Weaver on the other hand, has attempted to dump the puck out via the boards 16.7% if the time he's had possession in the d-zone, while Gilbert has dumped the puck out off the wall or glass only 7.8% of the time he's had possession. 

This brings us to one of Weaver biggest struggle in the defensive-zone. When Weaver attempts to dump the puck-out of the d-zone he's been successful with only 59% of those attempts. Gilbert on the other hand, has completed 76% of his d-zone dump-out attempts. These numbers carry over to icings, as Weaver has averaged 2.64 icings from the defensive-zone per-60, while Gilbert has averaged 1.02.

Dump-outs are not included as defensive-zone possession-driving plays as this play does not allow team's to clear the puck out of the d-zone while in possession of the puck. In the defensive-zone possession-driving-play-metric Gilbert has produced 4 more d-zone possession driving plays per-60 than Weaver, and 6 more neutral-zone possession-driving plays per-60. If we simply look at which player is skating the puck out of the defensive-zone, we see that Gilbert carries the puck out of the defensive-zone an average of 4.8 times per-60 minutes, while Weaver only skates the puck out of the d-zone 1.5 times per-60. These metrics tell us that Gilbert pushes possession up ice at a better rate than Weaver.

Weaver's even-strength possession-rating is .903, while Gilbert's is  1.146. This tells us that in terms of plays that directly impact possession (offensive and defensive), Gilbert contributes 1.146 more successful plays than failed plays per-minute, while Weaver contributes .903; loose-puck recoveries are not included in this metric, thereby separating it from a player's rating.

Defensively, the advantage easily goes to Gilbert. Number 77 has been successful with 70.9% of his even-strength attempts to remove puck-possession from the opposition at even-strength. Weaver's even-strength defensive success-rate in all 3 zones is 63.9%. In the defensive-zone alone, Gilbert has a defensive success-rate of 70.3%, while Weaver's is 66.3%. When combining corsi-events with scoring-chances, we see that 14.9% of the shot-attempts by the opposition that occur while Gilbert is on the ice, and 16.9% of those that occur while Weaver is on the ice are scoring-chances. This also speaks well for Gilbert's defensive-play at even-strength, as he is directly responsible for a lower percentage of scoring-chances within this metric than any other Montreal d-man.

While Weaver does block slightly more shots at even-strength than Gilbert, it's actually the latter that has a better success-rate when attempting to block a shot. Gilbert does blocks more opposition passes in the defensive-zone, to go along with the better success-rate.

Neutral-zone passing numbers seem to point the advantage in Weaver's favour. Number 43 has completed 80% of his east/west passes in the neutral-zone, with Gilbert completing 59.1%. North-bound passing success-rates show Weaver as successful with 60.9%; Gilbert successful with 55.6%. That said, in terms of actual successful passes per-60, Gilbert has averaged 9.3 successful neutral-zone passes per-60, while Weaver has averaged 7.56. Defensively, the neutral-zone has seen Gilbert be successful with 81.8% of his attempts to remove possession from the opposition, while Weaver has a neutral-zone defensive success-rate of only 56.5%.

Twenty-two percent of Weaver's events occur in the offensive-zone, while 20% of Gilbert's are in the o-zone. As for the defensive-zone, 67% of Weaver's events occur there, with 66% of Gilbert's events occur in the defensive-zone. Possession-wise Gilbert contributes 15.7% of his events while in possession of the puck in the offensive-zone, while 16.3% of Weaver events occur with possession of the puck in the offensive-zone. Conversely, 23.1% of Weaver's events occur while defending in the defensive-zone, and 21.8% of Gilbert's events occur while attempting to remove possession from the opposition in the defensive-zone. These possession percentages line up well with each player's on-ice corsi numbers, as Gilbert has been on the ice for fewer shot-attempts for, while also being on the ice for fewer shot-attempts against.

When placing these two players side-by-side at even-strength, there are only two instances where Weaver would win out over Gilbert; both those instances involve passing. The first is attempting to move the puck out of the defensive-zone by way of a pass, while the second is moving the puck through the neutral-zone by way of a pass. Defensively, Gilbert has a solid advantage over Weaver, while Gilbert's ability to skate the puck up ice also helps diminish Weaver's passing advantage. Gilbert on the other hand, has been the more successful defensive-player at even-strength, while also limiting his defensive-zone turnovers.