Kukla's Korner

The Puck Stops Here

Sophomore Jinx Candidates

There are many reasons why players suffer sophomore jinxes in the NHL.  One of the more predictable reasons is that the player was treated with kid gloves in his rookie season.  His team was careful not to put him in a situation where he would be “over his head”.  Now in a second season, he has to play a bigger role with his team and he is no longer as well protected as there is a new rookie to play the protected role.

One way to identify protected players is to look at offensive zone starts.  Some players were played preferentially in offensive situations and not in defensive ones.  Hence they will have a significant imbalance of offensive zone starts over defensive ones.

Two rookies placed quite high on the offensive zone start list.  Derek Stepan of the New York Rangers placed ninth and Kevin Shattenkirk of the St Louis Blues placed tenth.  This makes them strong candidates for a sophomore jinx.

Stepan finished fourth in scoring with the New York Rangers.  He scored 45 points.  He did so by playing a relatively easy offensive role with many offensive zone starts and a poor quality of opposition.  The New York Rangers will likely need him to play a more significant role next season and if they do, they may find he is not ready.  They will likely find that he can’t maintain his offensive numbers playing a tougher role.

Shattenkirk was the highest scoring rookie defenceman in the league last year.  He scored 43 points.  He split his time between the Colorado Avalanche and the St Louis Blues.  He too played a relatively easy role with many offensive zone starts and a poor quality of opposition.  Likely St Louis will use him in a more significant role next year and he will struggle to reproduce his offensive numbers.

Given how soft the roles Derek Stepan and Kevin Shattenkirk played in their rookie seasons, their offensive numbers were likely better than they would have produced in more standard roles.  Now that they are about to enter into sophomore seasons in the NHL and it is likely that they will be asked to fulfill a more difficult role.  This will be a tough adjustment and will likely affect them statistically.  Derek Stepan and Kevin Shattenkirk are both prime candidates to suffer a sophomore jinx season.

Filed in: | The Puck Stops Here | Permalink
 

Comments

Avatar

Hey, remember when you said that Kyle Clifford was the worst player in the NHL? Yeah, great insight there.

Posted by scooter from saskatoon on 08/05/11 at 03:24 PM ET

J.J. from Kansas's avatar

Who are the new rookies in to play the protected role in New York and St. Louis?

Posted by J.J. from Kansas on 08/05/11 at 03:35 PM ET

PuckStopsHere's avatar

Who are the new rookies in to play the protected role in New York and St. Louis?

We will find out in hindsight exactly who gets the protected roles on these teams next year.  At this point we can only guess.  It is a very safe guess that Stepan and Shattenkirk will not be as protected now that they are no longer rookies.

Posted by PuckStopsHere on 08/05/11 at 04:22 PM ET

Avatar

Hey, remember when you said that Kyle Clifford was the worst player in the NHL? Yeah, great insight there.

Posted by scooter from saskatoon on 08/05/11 at 01:24 PM ET

Maybe he was thinking of Westgarth. I always confuse those two for some reason.

Posted by steviesteve on 08/05/11 at 04:26 PM ET

Avatar

I’m curious—how did Shattenkirk’s offensive zone starts and quality of competition compare with Alex Pietrangelo’s?  Pietro was a rookie in all but name and ended up with the same number of points. 

As a Blues fan, I’m much more concerned about him having a sophomore slump than Shattenkirk.  I should re-phrase that.  I don’t think Pietro is as likely to have one, but it would be a far greater blow to the team if he does…

Posted by chapelhillblues on 08/05/11 at 05:20 PM ET

PuckStopsHere's avatar

Pietrangelo was third in offensive zone starts on the Blues (behind Shattenkirk and Berglund) and faced a roughly average quality of opposition.  He wasn’t as sheltered as Shattenkirk, so that probably puts him in a better place for next year.

Posted by PuckStopsHere on 08/05/11 at 05:27 PM ET

Avatar

For New York, Richards.

Posted by Ralph on 08/05/11 at 05:41 PM ET

Avatar

I believe I remember his calling Brad Richardson the worst player in the NHL as well

Posted by Brad from Calgary on 08/05/11 at 05:49 PM ET

Avatar

Doesn’t this assume that the players in question have not or will not improve? I can see this indicating they might get off to slow starts, but won’t they improve? Plus, isn’t Stepan like, third, on the depth chart at center behind Richards and Dubinsky? A lack of playing time would be more likely to lead to a drop in production.

Posted by Colin from Fredericton on 08/05/11 at 05:59 PM ET

PuckStopsHere's avatar

Doesn’t this assume that the players in question have not or will not improve?

That isn’t the way to look at things.  It shows that they were not as good as we thought they were last year because they played very sheltered roles.  They will have to improve to keep their numbers the same.

Posted by PuckStopsHere on 08/05/11 at 06:35 PM ET

EDJ's avatar

Who are the new rookies in to play the protected role in New York and St. Louis?

Posted by J.J. from Kansas on 08/05/11 at 01:35 PM ET

As J.J. said, if there are no rookies to replace them they may have similar protection by their coaches, if not as much as last year. On the other hand, this also ignores that players can be added ahead of them on the depth chart, also sheltering them from better quality of opposition. I’m not certain but if I remember correctly Stepan is a center, and now that the Rangers have added Brad Richards, Stepan may not have to face a better quality of opposition because Richards is taking some of that ice time. I think it’s more likely Stepan faces a worse year because he will not have the same quality of linemates that he played with last year.

Posted by EDJ on 08/05/11 at 07:16 PM ET

Avatar

I really don’t understand this guy’s blog - this and his corsi ratings are a perfect example of how data is misused in research and how he misunderstands (or simply ignores) the game of hockey in his interpretation. First and foremost, the data does not account for variation in players (ice time, experience, role, etc) and does not take into consideration how these factors change both within the same game and across games. His analyses also do not take into consideration the level of the opponent - either line combinations within an individual game or line combinations or entire teams across games. Anyone with any real understanding of the use and manipulation of data can tell you that this guy makes WAY too many assumptions - most of which are not consistent with what actually takes place on the ice - or with any consideration about how these factors change throughout the game or across games. I admire your attempt at providing some additional understanding or rating individual teams or players, but lets be fair about: identify your limitations with the data and do not write your analyses in a manner which suggests any certainty. I guarantee you that whatever (incorrect) models or algorithms that you are using are associated with tremendous error and your assumptions are not based on grounded theoretical assumptions derived from the game of hockey. Again, I admire what you are doing but if you are going to represent yourself as a researcher or even a novice that uses data to support his personal opinions - then conduct appropriate and sound work, be honest about what you can and cannot say empirically, and make sure that your results are supported by what actually takes place on the ice.

Posted by dr. no from calgary on 08/05/11 at 07:34 PM ET

PuckStopsHere's avatar

Dr No

You are totally wrong.  I believe you are working from the assumption that you don’t want numbers to be able to teach you anything about hockey.  That is a common place people who don’t like math start from.

Instead of attempting to discuss what is being - or likely even trying to understand them - you take the approach of saying that they cannot capture everything that happens in hockey and of course they can’t.

I am taking into account things like ice time, quality of opposition and linemates, experience and the role of the players and you claim I am not.  When your very non-specific argument is factually wrong, you are not doing much to earn any credibility.

Posted by PuckStopsHere on 08/05/11 at 07:43 PM ET

J.J. from Kansas's avatar

That is a common place people who don’t like math start from.

And some people apparently don’t like English much.

*snicker*

Also, I don’t see anything mentioning ice time, quality of opposition, or linemates in the above article.

Posted by J.J. from Kansas on 08/05/11 at 07:50 PM ET

Avatar

puck - its a blog post and honestly I’m not going to spend a great deal of time picking through all of your empirical and substantive limitations - many of which are evident by your interpretations without even seeing your algorithms. first off, I have a Ph.D. in statistics and about 20 years of hockey playing and coaching experience from which to start from. The fundamental flaw of your analysis - this is just the first of many that jump out - is that you assume that all games are equal and that each shift on the ice within those games are equal. The the first shift in the game carries the same weight as any other shift in the game. This also does not account for the individual game - whether it is the first game of the season, the third game of a 7 game road trip, the last of a home and home series, or any game in the last month of the season when teams are still playing for a playoff spot - not to mention the score of the game! Any player, coach, or fan of the game can tell you that there is absolutely nothing further from the truth. The differences in these games is tremendous: the ice time of your best players (which increases their opportunities, faceoffs, chances, etc), how much your role players sit, shifting of line mates, match ups against other teams, last change by the home team) - all of these things impact your Corsi ratings, your face offs, your shots on goal, your opportunities - strategies change and when they do, so does the fundamental role of each individual player. You are right - we cannot model the game perfectly but if you are going to attempt to make inferences about a very complicated game with significant variation within games and across games, you must be much more empirically savy than what you are presenting. Like I said, I admire what you are doing - and empirically testing these types of models are very complicated and the average person does not understand - and in most cases doesn’t want to understand - what you are saying. they simply trust that you know what you are doing. All I am saying is that those of us who use data to support our opinions and theories have a responsibility to be honest about what we can and cannot say, to do the best job we can, and attempt at every opportunity to present a model that best mirrors what is actually taking place on the ice.

Posted by dr. no from calgary on 08/05/11 at 09:15 PM ET

PuckStopsHere's avatar

The fundamental flaw of your analysis - this is just the first of many that jump out - is that you assume that all games are equal and that each shift on the ice within those games are equal. The the first shift in the game carries the same weight as any other shift in the game. This also does not account for the individual game - whether it is the first game of the season, the third game of a 7 game road trip, the last of a home and home series, or any game in the last month of the season when teams are still playing for a playoff spot - not to mention the score of the game!

False.  I assume that these things for the most part average out over the course of a season.  If a player has an imbalance of offensive zone starts over the season, he has an imbalance of zone starts in the first game of the season, the third game of a 7 game road trip, the last game of a home and home series or any gane in the last month of the season when teams are still playing for a playoff spot - nopt to mention the score of the game.

I don’t think these things are particularly important to this analysis in that they do not have a significant bias toward anything.  They even out over the season.

You could use exactly the same half-baked argument to argue that anyone using statistics like goals, wins etc. suffer the same problem for the same reasons that you list - but for the most part it all averages out over a long season.  If it doesn’t and it turns out that the effect of playing in the third game of a 7 game road trip significantly affects data, we can correct for it.

Anyone with a PhD in statistics (and not the phony kind that only exists in a blog comment) would know this.

Posted by PuckStopsHere on 08/05/11 at 09:30 PM ET

Avatar

False.  I assume that these things for the most part average out over the course of a season.  If a player has an imbalance of offensive zone starts over the season, he has an imbalance of zone starts in the first game of the season, the third game of a 7 game road trip, the last game of a home and home series or any gane in the last month of the season when teams are still playing for a playoff spot - nopt to mention the score of the game.

I don’t think these things are particularly important to this analysis in that they do not have a significant bias toward anything.  They even out over the season./quote]

If you think none of the factors in this are important, and even out over the course of a season, it is likely that you have never played any sort of competitive level hockey where these types of factors are a regular occurrence.

Playing the back end of a 3 in 4 nights with travel can be a challenge for any player. If this happens to be at a time where of the year where it is a busy schedule, your body reacts differently than at the start of the season when you are fresh.

I am a numbers guy for a lot of things in hockey, but you cannot base a player on these stats alone. You have to play in these scenarios before you can ever try to analyze them with solely statistics because hockey can never be transferred to sabremetrics with so many other factors affecting a player/team.

Posted by Brad from Calgary on 08/06/11 at 05:32 AM ET

J.J. from Kansas's avatar

The best hockey sabremetricians understand the limitations associated with what they’re trying to do and do a very excellent job of socially navigating the way they write their findings so as not to put a definitive stamp on their conclusions, but to understand that what they’re doing is still in the realm of “fairly close” and “better than a lot of other options.”

The worst of them say things like “everything evens out over the course of a season” (which is horribly untrue), attack the intelligence of those with whom he disagrees, ignores salient points, and engages in pedantic arguments which either consistently move goalposts or create strawmen.

However, let’s be fair as it relates to this specific article. Stepan and Shattenkirk are prime candidates for a sophomore jinx.  You could base that solely on offensive zone starts differential, or you could base it solely on the fact that they were #5 and 7 respectively on the list for rookie scorers and, since the guys at the top have a larger scoring amount to live up to, it will be the most difficult for them to live up to it.  Rounding out the top ten in rookie scoring, I believe that the two measures combined makes the argument even stronger.

1. Skinner
2. Couture
3. Grabner
4. Ennis
5. Stepan
6. Eberle
7. Shattenkirk
8. Hall
9. Marchand
10. Fowler

When you look at zone starts by percentage, Stepan had 64.1% of his starts in the offensive zone and Shattenkirk had 60.1%.  The next closest was Ennis at 54.5% - that is a very large gap.  Grabner and Eberle both started less than half of their shifts in the O-zone (indicating that they are already among the least protected of the rookies).

When you get to looking at the rookies’ QualComp and Qualteam scores, there isn’t a grand amount of differentiation there, but there are four players who had a negative QualComp and a positive QualTeam.

1. Skinner (-.023 QC .013 QT) -53% O-zone FO
2. Ennis (-.008 QC .130 QT) - 54.5% O-zone FO
3. Stepan (-.038 QC .038 QT) - 64.1% O-zone FO
4. Shattenkirk (-.033 QC .034 QT) - 60.1 O-zone FO

While QualComp and QualTeam stats aren’t great for comparing players among different teams, there is something telling about the numbers that indicate each of these four players were the most protected (and productive) rookies in the league last season.  I believe any of these four players would fit well into the category of prime candidate for a sophomore jinx (or slump).

Naturally, team effects are going to change this. Skinner’s Hurricanes are going to rely heavily on him in the coming season, especially after losing Erik Cole.

Tyler Ennis’ Sabres are going to have more ability to shelter him in the coming season with what I feel is the best team of the four included on this list (although it’s important to note that the Sabres are still currently $3.5M over the cap and will have to move at least one asset who would help protect Ennis in his sophomore season). 

Stepan’s Rangers are, in my estimation, just below the Sabres for the best teams and the Rangers have two centers on the depth chart higher than Stepan. I believe he’s also going to be able to keep a decent level of protection in his sophomore season.

Shattenkirk is not well-protected by the Blues’ defensive corps and could find himself having to work tougher minutes than he did last year. However, there will be some level of protection for him as I see he’ll likely get a lot of power play minutes for the team and will be able to help pad his stats there.  (The same could be said for Skinner in Carolina as well).

Based on all of that, I believe Jeff Skinner, as the highest rookie scorer and the winner of the Calder is going to have the toughest time among all of last year’s rookies to live up to the production he posted last season.  Behind him Stepan, then Shattenkirk, and finally Ennis.

As that ties back into the article above, I’d say that TPSH and I are in agreement.  All he said was that Stepan and Shattenkirk are prime candidates for a sophomore jinx; I completely agree with this.

(as an aside, of the top rookie scorers I believe are least likely to suffer a sophomore slump, I would put Couture, Fowler, Eberle, Grabner, Hall, and Marchand in that order)

Posted by J.J. from Kansas on 08/06/11 at 12:17 PM ET

Avatar

Stepan wasn’t really treated with kid gloves last year.  Tortarella played him in lots of different situations on lots of different lines.  Also, he was pretty lousy at faceoffs, so his “o-zone starts” would even out with his low FO%.

Posted by cb1 from sea, wa on 08/06/11 at 02:35 PM ET

Avatar

Perhaps you should pay more attention to the game and spend less time counting “offensive zone starts”!  Have you even seen Stepan play?  He played everywhere!  He played on the PP and the PK!  And Torts varied his lines almost nightly!  He centered a line with Gaborik and played wing on Prospals line.  The kid can play every role!

Posted by Jimmy from Ohio on 08/06/11 at 03:28 PM ET

PuckStopsHere's avatar

Perhaps you should pay more attention to the game and spend less time making insane rants.  Stepan played against weak competition and was shielded from defensive responsibility last year.  The kid didn’t get a chance to play tough roles.  When he does, I think his numbers will take a hit.

Posted by PuckStopsHere on 08/06/11 at 03:48 PM ET

J.J. from Kansas's avatar

Derek Stepan played just under 37 minutes of PK in a season where he played a total of 1,348 total minutes (2.7%).  Let’s not overstate what he did as a penalty killer.

Posted by J.J. from Kansas on 08/06/11 at 04:11 PM ET

Avatar

JJ from Kansas, yours is an excellent example of what I would consider an appropriate use of the data: supporting your opinions and hypotheses but not relying on the data to make your argument. I’m glad that there are other readers here that understand this and value the contribution that data “can” provide when used appropriately and honestly.

Posted by dr. no from calgary on 08/06/11 at 04:48 PM ET

Avatar

What I find hypocritical about this discussion is that the haters come out in droves when the author has a conclusion they disagree with, yet the dissenters can’t be found in the comments of a piece illustrating why, for example, the Sedins aren’t gods. With some exceptions, some of you try and find fault with the argument rather than the conclusions drawn. Just because the arguments are numerical in root does not mean you can’t look at them and interpret them differently. If you want an example, look at Kent Wilson’s piece at Flames Nation on Zanon.

Dr no: um, these are observational in nature, not experimental. The existence of confounding variables does not invalidate the study. It’s not like these numbers always lack context, it’s just that sometimes the context doesn’t add anything or it’s presented in words (e.g. quality of competition) rather than numbers.

JJ: Which variables don’t even out all that well? QualTeam, what else? I think for the basics, on the same team with similar sets of games played in a given year, it evens out decently.

Posted by Ralph on 08/09/11 at 01:37 AM ET

Avatar

Enough for our purposes, at any rate. I mean, we’re not the ones with sixty million to play around with.

Posted by Ralph on 08/09/11 at 01:39 AM ET

J.J. from Kansas's avatar

JJ: Which variables don’t even out all that well? QualTeam, what else? I think for the basics, on the same team with similar sets of games played in a given year, it evens out decently.

As far as comparing the rookies to each other for behindthenet’s advanced stats?  I think they all work somewhat well, just none of them great.  It’s a pedantic difference, but I’m happy with the way I worded it.  Generally, team effects have enough of a hit on Qualteam and even Qualcomp that they lose some of their power trying to compare across players on different teams and especially through different conferences, but overall they can’t just be thrown out altogether because they’re imperfect.

Posted by J.J. from Kansas on 08/09/11 at 02:26 AM ET

Avatar

Yeah, that’s tough. Looking at a guys rank on his team and the spread of QualComp scores on his team should get you a good idea, though

Posted by Ralph on 08/09/11 at 12:07 PM ET

Add a Comment

Please limit embedded image or media size to 575 pixels wide.

Add your own avatar by joining Kukla's Korner, or logging in and uploading one in your member control panel.

Captchas bug you? Join KK or log in and you won't have to bother.

Smileys

Notify me of follow-up comments?

Feed

Most Recent Blog Posts

About The Puck Stops Here

imageThe Puck Stops Here was founded during the 2004/05 lockout as a place to rant about hockey. The original site contains over 1000 posts, some of which were also published on FoxSports.com.

Who am I? A diehard hockey fan.

Why am I blogging? I want to.

Why are you reading it? ???

Email: y2kfhl@hotmail.com