Bye Bye Burmi

Though his agent assured everyone it wasn't going to happen, the inevitable did as it tends to do: Alexander Burmistrov and the Winnipeg Jets were unable to reach an agreement on a new deal, and the Russian forward signed a 2-year contract with Ak Bars Kazan of the KHL. While it's hard to evaluate the Jets' decision to not sign Burmistrov without knowing what he and Gandler were asking for, we do know that Burmistrov was an effective possession driver for the Jets and that he'd performed well relative to his peer group. Assuming Burmistrov only sought a contract in line with players who performed similarly during their ELC's, then the team’s inability to sign him to a new deal constitutes a needless loss of a valuable asset, a player who would have likely provided significant value over his salary. Instead, the Jets have a ton of salary and term tied up in Pavelec, Wheeler, Bogosian, and Byfuglien. NHL GM's are highly educated professionals who know just what they're doing.

Burmistrov's Corsi Effect

While Burmistrov had not displayed high scoring numbers at even-strength in the NHL yet, he had been an effective possession player. Not only did he have the second highest Corsi among regular Winnipeg forwards, but with the exception of Evander Kane, Burmistrov's five most common linemates saw their 5-on-5 Corsi% improved by playing with him:

2012-13 Teammate	Corsi% w/ Burmistrov	Corsi% w/o Burmistrov	Difference in Corsi%
Kane	48.8	48.9	-0.1
Tangradi	52.9	48.5	4.4
Antropov	50.7	46.7	4.0
Santorelli	54.0	49.1	4.9
Wellwood	52.2	50.6	1.6

The same effect showed up in 2011-12:

2011-12 Teammate	Corsi% w/ Burmistrov	Corsi% w/o Burmistrov	Difference in Corsi%
Kane	55.0	49.6	5.4
Wellwood	56.7	54.9	1.8
Antropov	48.6	51	-2.4
Wheeler	54.9	53	1.9
Miettinen	51.4	46.4	5.0

Burmistov's been a pretty effective possession driver for the Jets, but how much of this was merely a result of the way in which he was used?  Probably some, but not all that much:

Forward	CorsiRelQoC	Forward	CorsiRelQoC
Little	1.209	Burmistrov	0.043
Ladd	1.703	Santorelli	-0.038
Wheeler	0.902	Jokinen	-0.064
Antropov	0.602	Wright	-0.100
Miettinen	0.421	Tangradi	-0.306
Kane	0.349	Slater	-0.630
Wellwood	0.149	Thorburn	-0.944

Burmistrov was in the middle of the pack in terms of Quality of Competition, playing against tougher competition than Santorelli and Tangradi, yet he still helped their Corsi%. While it is certainly possible if not likely that some of Wellwood's and Antropov's Corsi improvement with Burmistrov was merely a result of playing against weaker competition and not an effect of Burmistrov driving play, Antropov, at least, got slightly more favorable ZoneStarts (56%) than Burmistrov (52%) and seems to have been used in a more offensive role and with better teammates.  That context mitigates any Quality of Competition advantage Burmistrov had.

2010 Drafted Forwards

Burmistrov is a rare example of a non-elite player who played each of his three post-draft seasons exclusively in the NHL, not counting his time in the AHL due to the louckout. Rather than compare him to players who have completed their ELC’s, let us compare him to other 2010 NHL-drafted forwards and specifically look at their Corsi% effects. The forwards are listed in descending order of draft selection and all the statistics are for 5-on-5 play only:

2010 F Draftee	CorsiRelQoC Rank w/in Team	Weighted Corsi% Effect	ZoneStart	ESP/60 (Rank)
Hall	2	9.8	54.8	3.15 (1)
Seguin	4	8.7	53.9	2.27 (3)
Johansen	3	2.4	50.6	1.26 (12)
Skinner	2	-3.3	52.0	1.23 (9)
Burmistrov	8	2.7	52.0	1.04 (8)
Granlund	11	-10.1	54.0	1.19 (10)
Schwartz	8	-1.8	51.3	1.47 (9)
Tarasenko	9	7.3	67.4	1.92 (3)
Coyle	1	5.8	64.6	1.52 (7)
Etem	13	1.4	55.1	1.25 (12)
Gallagher	9	2.6	66.0	2.86 (1)

The forwards can be placed into a few different groups based on usage and performance:

• Driving possession against top lines.  Taylor Hall is a stud.

• Facing top lines, benefiting from linemates.  Seguin played mostly with Bergeron and Marchand.  His 8.7% Corsi effect is a bit misleading because Bergeron and Marchand saw more defensive matchups, i.e., more own-zone draws (Seguin had the highest ZoneStart% of the three) without Seguin and were on the ice more often when protecting a lead or after a Bruins penalty expired.  Playing in these situations would have decreased their Corsi% and likely explains some of Seguin’s substantial positive Corsi effect.  Sequin is clearly an effective player but his possession and scoring numbers were likely helped by his linemates and not necessarily vice-versa.  Charlie Coyle was used in a similar role:  He played mostly on Minnesota’s top line with Koivu and Parise and faced top-line competition. The drop in Koivu’s and Parise’s Corsis when separated from Coyle can be largely attributed to Koivu and Parise playing mostly with Heatley when away from Coyle, and that line’s Corsi% was significantly worse than with Coyle.  That is possibly due to a combination of (1) Coyle being a better possession driver than Heatley at this point in their careers (though with only one-half a season of data on Coyle it’s difficult to conclude that with much certainty) and (2) the fact that Minnesota became a much better Corsi team as the season went on, making the Corsi% of the Coyle-Koivu-Parise line, which formed later in the season and coincided with the team’s improvement, look disproportionately good relative to Heatley-Koivu-Parise.

• Facing top-6 competition.  Johansen and Skinner both played against mostly top-line or top-6 competition. Whereas Johansen slightly improved his linemates Corsi, Skinner’s linemates did slightly better without him.  However, Skinner played mostly with Jordan Staal, a noted possession driver, and Johansen played with R.J. Umberger and Nick Foligno, two productive NHL’ers but not necessarily possession stars.  It’s also worth noting that Skinner’s Corsi% stayed roughly the same when he played away from Staal, whereas Johansen’s Corsi% cratered (going from around 47% to 39%) when he was separated from Umberger or Foligno.  It appears likely that any contribution that Johansen made to his linemates should be more accurately thought of as an effect that each of his linemates received when playing together, either due to situational usage or some possible synergistic effect among them.

• Playing with and against the middle of the opposition’s lineup.  Burmistrov and Gallagher had roughly equivalent positive Corsi effects on their linemates.  However, Burmistrov played with a greater variety of linemates on scoring and checking lines, and he increased possession on all of them (with the exception of breaking even with Kane).  Gallagher, on the other hand, played mostly in a scoring role with Pacioretty and Desharnais, and while the three of them helped each other when together, it was arguably Pacioretty driving possession.   Pacioretty’s Corsi% was roughly the same with or without Gallagher, whereas Gallagher fell apart without Pacioretty.  Burmistrov’s Corsi% was far more consistent from line to line, and he did not see his Corsi% significantly dip regardless of the line on which he was used.  In fact, Burmistrov had the second highest RelCorsi on the team after Andrew Ladd.  Gallagher, on the other hand, was 9^th on his team.  Tarasenko and Schwartz basically switched roles after Tarasenko’s concussion, which derailed what could have been a Calder Trophy season.  He ended up being used more in offensive situations, i.e., playing more when trailing and receiving a high ZoneStart%, which likely exaggerated his Corsi%.  That said, Tarasenko appears to have the makings of a substantial possession driver if used in a wider role.  Schwartz played mostly with good linemates and although he struggled with Berglund and Stewart, he excelled with Backes and Oshie. 

• Fourth lines. Michael Granlund and Emerson Etem both played against weak competition.  Granlund got absolutely destroyed possession-wise doing so, but his negative Corsi effect was further hurt by the fact that by time Minnesota turned their season around, Granlund was in the AHL.  His teammates’ Corsi% looks better without him, but much of that can be attributed to that fact that Minnesota became a better possession team as the season went on.  Etem slightly improved his linemates’ Corsi% but played mostly with players who do not drive possession.  For example, he often played with David Steckel, whose Corsi% was a low 41% without Etem, but Steckel was likely playing harder matchups in those situations and Etem was sheltered away from them.

At this point, Hall is clearly the best player in this group, with Seguin and Skinner, who struggled this season but has shown an ability to score, probably trailing most closely behind.  After those three, however, the field is pretty open.  Coyle and Gallagher put up good numbers but did so while playing with good linemates and in offensive roles.  Burmistrov’s closest comparable within the group appears to be Johansen.  Burmistrov has had two seasons where he clearly drove possession among his linemates, whereas Johansen mostly kept his head above water while playing with Umberger and Foligno but really struggled when separated from them.  Johansen’s raw scoring rate was higher than Burmistrov’s, but relative to his teammates, Burmistrov’s scoring rate was higher (8^th on his team compared to Johansen who was 12^th).  I think the two players are fairly close in value at this stage in their careers and it will be interesting to see what Johansen’s next contract looks like.  It's not a direct comparison because Johansen will have had an extra year of development by the time his post-ELC contract starts, but I'd bet Johansen gets significantly more than what the Jets were offering Burmistrov.

The Second Contract

While Burmistrov posts quality possession statistics when playing with and against average NHL players, his even-strength scoring output has been a meager 1.17 ESPoints/60 over his career thus far.  This makes him a difficult player to value coming out of his ELC and into his second contract, an important contract for the club because the player does not yet have arbitration rights and is still four year seasons away from unrestricted free agency.  Let’s compare Burmistrov to other forwards with scoring rates between 0.9 and 1.8 in the final year of their ELC and who spent at least two full NHL seasons prior to the second contract:

Player (Season)	ESP/60 (Rank w/in Team)	RelCorsi Rank w/in Team	CorsiRelQoC Rank w/in Team	2nd Contract Cap Hit (Term)
Clifford 2012-13	1.78 (5)	6	7	1.075 (2)
Paajarvi 2012-13	1.54 (5)	7	8	1.2M (2)
Sutter 2010-11	1.39 (8)	11	3	2.07M (3)
Wilson 2011-12	1.38 (12)	2	8	2.0M (3)
Kruger 2012-13	1.25 (10)	8	8	1.325M (2)
Bailey 2010-11	1.09 (11)	8	8	1.05M (2)
Burmistrov 2012-13	1.04 (8)	2	8	KHL
Backlund 2011-12	.90 (10)	1	5	.725M (1)

Burmistrov’s performance compares well to Wilson, Bailey, and Kruger.  Though Burmistrov had a lower scoring rate than each of them, his scoring rank within his team is actually higher.  Like Wilson, Burmistrov was second on his team in Corsi, but he did so while playing with worse players than Wilson, who played mostly with Legwand and Hornqvist that season.  Sutter’s performance probably outpaces both Wilson and Burmistrov at that stage due to his scoring rate coming against strong competition.  Backlund missed half of his final ELC season with various injuries, which probably explains the low salary on his one-year deal.  He increased his scoring rate the following season (though again only ranked 10^th on his team) but missed significant time with injuries again and signed a two-year contract with a cap hit of $1.5M per year.  Clifford and Paajarvi posted higher scoring rates than the rest of the group, but their teams were able to sign them to relatively cheaper second contracts in part by signing them late in the summer and either threatening a trade, in the case of Clifford, or actually completing a trade, in the case of Paajarvi.

Burmistrov’s closest comparables in this group are Bailgy and Kruger.  They posted similar scoring rates against similar competition over the course of their ELC’s, though Burmistrov had more success driving possession. All three received limited power play time and were used more significantly as penalty killers.  In 2011, Bailey signed a two-year contract with a cap hit of $1.05M per year.  His scoring rates steadily increased during the life of that contract, and he recently signed a five-year contract with a cap hit of $3.3M per year.  Kruger recently signed a two-year contract with a cap hit of $1.325M per year.  Unlike Burmistrov, Kruger has had the opportunity to play and excel in the playoffs, and part of his value comes from playoff performance, an opportunity Burmistrov has not had. Burmistrov’s value is somewhere between Kruger’s and Bailey’s.  A two-year contract with a cap hit of $1.25M would have been a reasonable deal for both the team and the player.   However, because Burmistrov had a realistic option of signing with a KHL team – in fact, he did just that – he had more leverage than Kruger and Bailey and could have likely commanded $1.5M per year while still providing value over that contract.  However, the team could have also chosen to lock up Burmistrov for a longer term in order to avoid the type of bigger contract Bailey ultimately received.  A three-year contract for around $2M per year would have likely been reasonable, or the team could have tried for a five-year deal at around $2.75M per year, which would have covered one year of Burmistrov’s unrestricted free agency.  The team would have not likely wanted to sign Burmistrov to a four-year contract as it would have taken Burmistrov right up to unrestricted free agency.

It’s unclear if Winnipeg seriously considered any of those options.  I’d guess that they wanted to get Burmistrov signed to a one- or two-year deal for $1M or less per year.  And I’d guess that Burmistrov wanted to go to a different franchise and, particularly, to a different head coach.  Claude Noel had been critical of Burmistrov and had scratched him during the season.  Burmistrov’s agent’s comments earlier in the summer to the effect that Burmistrov would not sign with a KHL team and would be playing in the NHL next season were likely made to entice NHL teams to submit an offer sheet to him or to trade for his rights.  A team could have signed Burmistrov to an offer sheet with a cap hit of up to $1.6M while only giving up a third round draft choice, or an offer sheet with a cap hit of up to $3.3M while only giving up a second round draft choice.  It’s not clear to me that Winnipeg would have definitely matched an offer sheet, though a team could have almost certainly traded a 1^st or 2^nd round 2013 draft selection to obtain Burmistrov’s rights.  Instead, Winnipeg received nothing for Burmistrov and though they retain his rights, it seems unlikely that Burmistrov will ever return to the NHL.  He will likely make far more money in the KHL than the Jets or another NHL team would be willing to pay him after two years out of the NHL.  It was terrible asset management by Winnipeg as they have lose a player that was on his way to becoming a solid if unspectacular NHL contributor.  Burmistrov, who left Russia to play junior hockey in Canada and played in the AHL during the lockout, seems to have been eager to play in the NHL at the right price and in the right role.  Sutter, Wilson, Kruger, Bailey, and Backlund have all become value-providing NHL players on reasonable contracts, and I would have bet Burmistrov would have followed that path in Winnipeg or another NHL team.  It seems that we will never find out, and it is Winnipeg’s loss.