Hurricanes GVT Values and Expectations

With advanced stats in hockey growing by the second, there have been a number of new metrics devised to judge and predict player performance. One of the most popular is Goals Versus Threshold (GVT) which measures how many goals a player is worth compared to a replacement level player. This stat was developed by the people at Hockey Prospectus and is broken down into three different parts: Offensive GVT, Defensive GVT and Shootout GVT. All three areas are looked at because the point of GVT is to find out who is contributing the most to wins. Goals lead to wins, so it is necessary to look at a player’s performance outside of his goal, point & plus/minus total. Goaltending is also looked at, but that is a completely different system.

A closer look at GVT and how the Hurricanes players performed by this stat is coming after the jump.

A more detailed explanation of GVT and the formula can found on Hockey Prospectus’ web site, but I will do my best to explain how the stat is devised here. A player’s Offensive GVT is how many goals a player created above replacement level. Goals and assists are looked at with this stat and they are compared with what is considered the "threshold level" for forwards and defensemen to determine how much offensive value a certain player had. For example a player with an offensive GVT of 5 means that he scored five more goals than a replacement level player would if they got the same ice time.

To figure out a player’s Defensive GVT, the amount of shots and goals against a player was on for are looked at. This isn’t a perfect method because luck and goaltending both play a huge role in goals against. The situations a player is used in also plays a role in that along with the amount of shots a player is on-ice for. For instance, Tim Gleason is almost guaranteed to be on ice for more shots against than Jaroslav Spacek because he plays in more difficult situations and GVT might underrate his value. Gleason is also put at a disadvantage because he plays significantly tougher opponents than Spacek and GVT might overlook that.

This is my main issue with GVT as a "definitive stat." Hockey has so many things going on and it’s hard to encapsulate it with just one statistic, but GVT is a nice step forward in hockey metrics in terms of figuring out how much value a player has. However, some defensive forwards & shutdown defensemen tend to get undervalued by this stat. The same thing goes for players who had a luck-filled and have a low chance of repeating it. It goes to show you that, every stat, GVT needs to be taken with context and hopefully that will come in the not-so-distant future.

As for shootout GVT, it is a necessary evil because shootouts are part of the game and, unfortunately, decides who wins and loses some games in the NHL. Shootout GVT is devised by looking at the league average shootout success rate, a player’s shootout performance compared to that (shootout goals/shootout shot attempts) and whether or not the shootout goal was game-deciding.

The last thing to discuss with GVT is threshold value and how it is defined. Threshold value and "replacement value" are extremely similar and you will often hear it discussed more by baseball fans rather than hockey fans. A replacement level player is someone who can essentially miss the entire season and be replaced by an AHL call-up or someone making the league minimum without any issue. Brett Sutter and Derek Joslin would be examples of replacement level players on the Hurricanes.

Translating GVT into wins can be done by following the "3-1-1" rule from Rob Vollman of Hockey Prospectus. This means that three goals is worth $1 mil. and that is worth one point in the standings. Which means that a GVT of 6 is equal to one win and finding the exact amount of wins that a player contributed can be done by dividing his GVT by six.

Again, this method is far from perfect but I thought it would be interesting to see how much the Hurricanes players are worth through this method and see if we can find any discrepancies between GVT and our general thoughts and opinions. There are bound to be at least a few.

With all of that out of the way, let’s see how the Hurricanes look through GVT. I’ve separated the forwards and defensemen since the threshold value for each is different. What I will also do is compare a player’s GVT to what their projected total was by Hockey Prospectus in their annual season preview, which can be bought here.

Player	GP	OGVT	DGVT	SGVT	GVT	GVT Per 80	EOGVT	EDGVT	ESGVT	EGVT	EGVT per 80	Diff	Diff/80
Eric Staal	82	10	1.8	-0.3	11.5	11.22	10.9	2.3	0	13.2	13.57	-1.7	-2.35
Brandon Sutter	82	4.8	3	0	7.8	7.61	4.3	3	0	7.1	7.78	0.7	-0.17
Jeff Skinner	64	6.7	1.5	-0.9	7.3	9.13	13	2.1	0	15.1	14.86	-7.8	-5.73
Jussi Jokinen	79	4.4	2.6	0	7	7.09	7.1	2	0	9	10.81	-2	-3.72
Jiri Tlusty	79	4.5	2.6	-0.3	6.9	6.99	0.4	1.3	0	1.7	2.66	5.2	4.33
Tuomo Ruutu	72	3.3	2.1	-0.3	5.1	5.67	6.3	1.8	0	8.1	9.36	-3	-3.70
Chad LaRose	67	3.3	0.9	-0.6	3.7	4.42	3	1.5	0	4.9	5.78	-1.2	-1.36
Anthony Stewart	77	2.2	1.3	0	3.5	3.64	3.8	1.8	0	5.6	6.70	-2.1	-3.06
Drayson Bowman	37	1.4	1.2	0	2.5	5.41	-0.1	0.7	0	0.6	1.38	1.9	4.03
Tim Brent	79	1.5	0.8	0	2.3	2.33	0.4	1.8	0	2.1	2.69	0.2	-0.36
Patrick Dwyer	73	-1.9	2.3	0	0.4	0.44	0.7	1.9	0	2.5	3.17	-2.1	-2.74
Jerome Samson	16	0.1	0	0	0.1	0.50	0.3	0.8	0	1.1	2.65	-1	-2.15
Riley Nash	5	-0.2	0.2	0	0	0.00	0	0	0	0	0.00	0	0.00
Jeremy Welsh	1	0	0	0	0	0.00	0	0	0	0	0.00	0	0.00
Andreas Nodl	60	-0.7	0.6	0	-0.1	-0.13	1.4	1.9	0	3.2	4.33	-3.3	-4.46
Brett Sutter	15	-0.3	0.3	-0.3	-0.3	-1.60	0.4	0.6	0	1	3.02	-1.3	-4.62
Zac Dalpe	16	-0.4	0.1	0	-0.3	-1.50	1.2	0.7	0	1.8	4.30	-2.1	-5.80
Zach Boychuk	16	-0.6	0	0	-0.6	-3.00	1.3	0.9	0	2.2	4.63	-2.8	-7.63

OGVT = Offensive Goals Versus Threshold, DGVT = Defensive Goals Versus Threshold, SGVT = Shootout Goals Versus Threshold, GVT = Goals Versus Threshold, GVT per 80 = Goals Versus Threshold per 80 games, EOGVT = Expected Offensive GVT, EDGVT = Expected Defensive Goals Versus Threshold, EGVT per 80 = Expected Goals Versus Threshold Per 80 Games, Diff = Difference between actual GVT & projection, Diff/80 = Difference between actual GVT per 80 Games & Projection

Almost every forward underperformed their expectations going by GVT with Jeff Skinner actually being the biggest disappointment. I want to say that injuries held him back but GVT projected him to be even better than Eric Staal this year, which is a little steep to me. Skinner had a better rookie season than I could have imagined but at only 19 years old, GVT projected him to score 13 more goals than a replacement level player and 15 wins better overall. Players who had a GVT of around 15 include Logan Couture, Anze Kopitar, Jamie Benn and Valtterri Filppula, guys who had around 65-76 points. Skinner may have reached that plateau if he didn't suffer a concussion but I think Hockey Prospectus overrated him a little bit.

Staal is another player who finished below his projection, but not by a lot. If you look closely at the data, you will see that Staal's offensive production was in the same ballpark as where he finished and his "defensive value" knocked him down a little. I am willing to bet that was from the amount of goals he was on-ice for and his terrible +/- rating. The problem I have with this and plus/minus in general is that it assumes that every goal Staal was on-ice for was his fault, which likely isn't the case. Remember, GVT looks at the amount of goals against a player was on-ice for when calculating GVT and Staal was on ice for 74. Staal was likely at fault for some of them but it probably wasn't his fault that the Carolina goalies posted a save percentage of .897 at even strength when he was on the ice. Staal wasn't great defensively this year but I don't think he was as bad as it is indicated here.

Tuomo Ruutu and Jussi Jokinen were also listed as disappointments and it's mostly because they didn't score as much as they were expected to. That is definitely fair to say because Jokinen had only 12 goals and Ruutu had only 34 points. I am a little more concerned about Ruutu than I am Jokinen because the latter was still able to produce a lot of shots and scoring chances at even strength while not giving up much in his own end. Ruutu, on the other hand, wasn't as effective in that department.

Chad LaRose also finished below his expected total because of his poor defensive GVT and negative rating in the shootout. His defensive GVT being poor is likely for similar reasons to Staal where he was on-ice for a lot of goals against, but LaRose also had trouble with preventing shots against. He plays in tough territorial situations (45% OZ Start) but he was on-ice for nearly 30 shots against per 60 minutes. Although, LaRose also produced a lot of shots and scoring chances for the Hurricanes, so I would have thought that would at least make-up for his defensive issues but that isn't the case.

Another player who underperformed according to GVT was Patrick Dwyer and it almost entirely because of his horrific offensive showing. His defensive GVT is actually around what was expected but his lack of offense this season made him barely above replacement level. Dwyer is a defensive forward, so I don't expect him to put up points but I guess it's fair to say that I was expecting more than five goals and 12 points from him. Maybe he would have had more than 7 assists if the team didn't shoot lower than 5% when he was on the ice at even strength. Just a thought. Andreas Nodl also had this problem but GVT had much higher expectations for him, likely because he began the year on the Flyers and played in their top-nine last season.

The two forwards who outperformed their expectations were Drayson Bowman and Jiri Tlusty. Bowman wasn't expected to contribute much this year, but he proved to be a lot more valuable than most predicted and has a good shot of making the team next year. It is interesting to see that he was the call-up who produced the most when it was expected to be Zac Dalpe and Zach Boychuk.

Hockey Prospectus projected Tlusty to play only 51 games this year and be more valuable as a defensive forward. He ended up playing in the top-six, was the team's fourth most valuable forward and blew away just about everyone's expectations.

Defensemen

Player	GP	OGVT	DGVT	SGVT	GVT	GVT per 80	EOGVT	EDGVT	ESGVT	EGVT	EGVT per 80	Diff	Diff/80
Tim Gleason	82	0.7	5.8	0	6.5	6.34	0.9	2.5	0	3.4	4.18	3.1	2.16
Jay Harrison	72	4.3	1.4	0	5.7	6.33	-0.1	2.6	0	2.5	3.39	3.2	2.94
Jamie McBain	76	3.5	2.1	0	5.7	6.00	3.9	2.2	0	6.1	7.35	-0.4	-1.35
Jaro Spacek	46	2.9	2.1	0	4.9	8.52	0.8	3.3	0	4.1	5.88	0.8	2.64
Bryan Allen	82	-0.3	3.7	0	3.4	3.32	0.1	2.7	0	2.8	3.76	0.6	-0.45
Justin Faulk	66	2.5	0.1	0	2.6	3.15	0	0	0	0	0.00	2.6	3.15
Joni Pitkanen	30	3.2	-0.8	0	2.4	6.40	4.8	3	0	7.8	9.47	-5.4	-3.07
Bobby Sanguinetti	3	-0.1	0.1	0	0	0.00	0	1	0	1	2.58	-1	-2.58
Derek Joslin	44	-0.4	-1.2	0	-1.6	-2.91	0.5	1.7	0	2.2	4.35	-3.8	-7.25

The defense has a much more positive outlook as three players outperformed their expectations and leading the way is Tim Gleason. He proved to be very valuable as a shutdown defenseman this season and this was with him playing the toughest minutes on the team. GVT had him projected pretty low because shutdown defensemen usually don't rank that high unless they produce points, but Gleason managed to provide a lot of defensive value for the Canes. However, GVT does overrate him a bit because he had a high on-ice save percentage and I find it interesting that he ranked higher than Bryan Allen considering they were paired together for more than half of the season. Both played great this year but Allen had lower on-ice shooting and save percentages which may have led to him having a higher GVT.

The other player who outperformed expectations was Jay Harrison. He was expected to be a third-pairing defenseman this year but ended up playing top-four minutes and had a career-year offensively. Harrison actually had the highest offensive GVT of every Carolina defenseman, which is kind of surprising. Justin Faulk also outperformed his expectations but that was because Hockey Prospectus didn't think he would be in the NHL this season. That alone is an accomplishment but I think GVT is lowballing him in defensive value.

The biggest underperformer of the group is Joni Pitkanen and that is mostly due to injuries and poor defensive play. He would rank higher in the defense corps if he played for the entire year, but he would still finish as a disappointment. While I can't deny that Pitkanen did have a underwhelming season, his DGVT suggests that he is a defensive liability, which I do not think is the case at all. No, Pitkanen isn't known for his play in his own end but he isn't nearly as bad as it is indicated here.

Derek Joslin was the other letdown and that's pretty bad because his expectations weren't even high to begin with. Playing forward for half of his season probably had something to do with that, but he was still facing easier competition and all he had to do was not get hammered in his own end. According to the underlying stats and GVT, that didn't happen. Almost any forward from Charlotte would have been an improvement over him on the fourth line.

While I do have issues with GVT, there are things that we can determine from this. The first of which is that there were a lot of forwards who underperformed and this is a big concern heading into next year. Bad luck played a role in Jokinen's underwhelming season but the lack of high-end offensive talent is something that needs to be addressed this off-season. From the sound of things, Jim Rutherford will address this need in the coming months. The other thing that can be said is that the call-up stints from Boychuk & Dalpe were very disappointing and they will need to do better in the future.

The Canes are also going to need Pitkanen, Harrison or Faulk to really step it up defensively if Bryan Allen leaves because other than Gleason, there weren't many defensive stand-outs on the blue-line. I'm interested to see how Rutherford addresses that need, as well. The good news is that the Canes have prospects and guys in the AHL who they can use next season and may not need to overpay for a free agent to replace him. That might be what is done since the forwards appear to be a bigger issue at the moment.

In the next couple of days, we will use GVT to figure out who is giving the Hurricanes the most for their money by comparing their GVT rating to their salary.