Friday, August 14, 2015

You Want More? You'll Get It.

Eventually.

The media has really caught on to the idea that there are certain repeatable skills in the NHL. Measuring these skills has been labeled "analytics" and the only ones who are really making decisions based on things unrelated to analytics are NHL General Managers. Oops!

Luckily for us, not everyone gets it. A few people think they do, but it isn't the "Summer of Any Stats Ever lol", it's the "Summer of Misunderstanding Analytics."

Quick, what do Jeff Skinner's rookie campaign and his 2013-2014 season have in common?

Shit, I dunno. I'll just take a dumb guess. Uhhh...he had a shooting percentage above his NHL career average.

If you guessed that he had a shooting percentage above his NHL career average, you win the prize.

Finally, I can retire. Good night, internet!

Skinner has a career shooting percentage of 10.5%. During his rookie year he shot at a 14.4% rate. When he next potted in excess of 30 goals, he was shooting at a rate of 12%, still a good 1.5% above his mean. What about Eric Staal's two best scoring years? Yup, you guessed it. His shooting percentage was well above his career average too.

First, two of these seasons were cut short; one by injuries and one by greed. In each season, he projects for 25 goals across 82 games. Is that a below average season? Is that a bad goal-scoring season? No, it isn't. It isn't cherry picking his 2 best seasons, but we're only 6 goals off.

Second, do you know how many more goals he would've scored if he'd shot his mean this year? 6 more. Six. Or about the difference between his pace in those shortened years and his "career best."

When conducting research into shooting percentage it is easy to make some assumptions and some generalizations.

Exactly. Ok, maybe I've been too harsh. I'll do a quick CTRL+F here...




Well, that was unexpected.

They sometimes just don't hold water. However, increases in shooting % generally lead to improved offensive years. More specifically, when a player shoots above his career average, he tends to have a year of high offensive output. Conversely, when players shoot below their career average, their offensive output can be characterized as "disappointing".

Do I really have to parse this? I mean, c'mon.

Another generalization is that players that shoot over the average league shooting percentage tend to be better scorers.

Get out.

Here's something fun. In 2013-14, the average shot percentage in the NHL was 8.9%. You have to go all the way down 89th in the goals that year to find someone who shot below average. Brad Richards, Dustin Byfuglien, Erik Karlsson, and Radim Vrbata all tied with 20 goals for 89th place. Two of them were defensemen.

Players who score more will have higher shot percentages and higher point totals. This is not worth mentioning.

Getting back to young Jeff, much can be concluded about the "what", but not as much about the "why". That is more based on educated guesswork. Consider a couple of extrapolations. Last year his shooting percentage was a dismal 7.7%. For the last two years, the league average was about 9.1% (which includes the historically lower percentages of the blueliners). If Skinner merely returns to his average, he likely scores 25 goals.

I already mentioned this. Do you read my articles about your articles before you post your articles?

And, I'll check again but...




Just making sure I didn't spell anything incorrectly.

With his lifetime assists to goals ratio of .88, one can estimate that Jeff ends up with around 47 points. That is a significant improvement over last season's 31 points.

Are we allowed to do that? Just add assists willy-nilly? I expect Danny Briere to have a career year in 15-16, based on his lifetime assists-to-goals ratio.

Jeff Skinner had teammates that sucked. He skated most with Victor Rask, Elias Lindholm, Alex Semin, and Riley Nash. Is that a list of good players in 2015? No. It isn't. It's a miracle Jeff Skinner had 13 assists. What a stud.

Another important and related statistic is the number of shots taken. Simply put, scorers need to be taking a lot of shots. Skinner's NHL average in this area is 234. In 2013-14, he took 275 shots and, of course, he scored 33 goals that year. His rookie year he only took 215 shots but some of that can probably be attributed to "new kid" nerves and, frankly, he was making the most of his high quality opportunities.

Superb undermining of your own argument. Taking more shots is important, except for when it isn't and then we need to use clichés and hyperbole because we just can't explain why sometimes, gosh-darn-it, the puck just goes it for some crazy reason rofl!

What about those guys that would be considered comparable to Jeff Skinner? Again, normalizing for partial years and extracting the strike shortened season, the same assumptions seem to hold true for them. Consider these three players: Tyler Seguin, Logan Couture, and Kyle Turris. All 1st round draft picks, all in the league about the same amount of time, and all considered goal scorers.

Are they? I would consider Couture a goal-scorer and Seguin is an all-around scoring star, but Turris? Kyle Turris is about 60-40 assists-to-goals and his career high in goals in 26. Even Seguin seems just as content setting up a play than he does scoring himself.

First, let's look at Tyler Seguin. He's got a career shooting % of 11.7% and averages 237 shots per season. His best years were 2013-14 and 2014-15 when he had 84 points, a shooting percentage (SH%) of 12.6% and 77 points with a SH% of 13.2% respectively.

And also got traded to Dallas and played with ART ROSS WINNER JAMIE BENN. Jamie Benn vs. Victor Rask, Alex Semin, and Riley Nash. If you ADD ALL THREE their point totals fall short of what Benn put up last year. Unreal.

Logan Couture's lifetime average is 11.7% with an average of 249 shots per season. His 2nd and 3rd best scoring years had him shooting at nearly a full percentage point above his career average. In 2013-14, shooting well below his average led to his lowest goal scoring output (23). Last year was one of those outliers where he shot below his career average (10.3%) but still saw him put up a career best 67 points. Those point totals did include a higher proportion of assists however.

He skates with Patrick Marleau and saw over 11% of his ice time last year on the power play with Joe Thornton and Joe Pavelski. Over 12% of the time, Couture was on the ice with 2 of Pavelski, Thornton, and Marleau while the other team had fewer players and often he was with all 3.

Jeff Skinner didn't play 11% of the time on the power play period (Is Bill Peters drunk?), let alone with 3 other good forwards. Not that he'd have the chance anyway. It's looking more and more like Jeff Skinner's 18-goal season was phenomenal.

Lastly, Kyle Turris, historically not a high volume shooter, exceeded his average shots per season in the last two years by nearly 60 shots (215 vs. 158). Accompanying this his shooting percentages of 12.1% and 11.2% easily bested his 10.2% career average. His point totals were clear beneficiaries at 58 and 64 points in those last two years. In each case these young scorers seem to confirm that an improved shooting % leads to better scoring performances.

How about just getting older and better? These are dudes hitting their prime years and they are improving in the points department. They are older so coaches trust them more in scoring situations and with more ice time.

Also, in Turris' case, if in 14-15, he would've shot below his career shot % (and we'll be fair and give him the same deviation, dropping his career mark by 1% to 9.2) he would've had four fewer goals.





4.




And still would've set a career high for points. Do I have to continue? This is torture.

Does this same pattern hold true for more veteran scorers?

God, please kill me.

The research does indeed show a similar correlation.

Prove it.

Using the Eric Staal examples mentioned above, he has a SH% of 10.9%. His best years he shot 16.1% and 12.3% scoring 100 and 82 points respectively. Last year he underperformed in this category with a 9.4% rate. If Staal performed at his career average and using the same extrapolation that was used with Skinner earlier, he scores around 64 points. To add to last year's woes for Eric, he shot nearly 30 less shots than his career average of 272.

Fewer. You meant to type fewer.

What about players comparable to Staal? Two of Rick Nash's three best offensive years had him shooting at a 15.2% and 13.8% clip, well above his career 12.5% mark.

Just going to ignore that other season from Rick Nash? Well, I'm not.

I guess the author just sorted by points. The difference between Nash's 2nd "best" season and his 5th is...three points. 3. points.

I have to think at this point in the article, you stopped to think just for a moment, "Championing shot percentage might not be such a hot idea. Maybe I'm learning something new here and shot percentage really doesn't mean all that much because it's so full of noise and context. Eh fuck it, no one reads this shit."*

*my actual thoughts when making fun of strangers on the internet

Corey Perry, a career 13.4% shooter, posts 98 and 82 point seasons when he shoots at a rate of 17.2% and 15.4%, well over his 13.4% average. Similarly, Joe Pavelski's two best offensive seasons came when he shot an incredible 18.2% and 14.2% (79 and 70 points). Again, this was well above his lifetime NHL average of 11.5%. To reiterate, players, especially goal scorers, who exceed their historical shooting percentage averages tend to have excellent overall offensive years.

I'm not repeating myself. See: everything I've written.

There are, of course, the Jiri Hudler's of the world whom are emblematic of the SH%/scoring relationship. This past year he had one of his best, shooting at an incredible 19.6% clip, scoring 76 points. Sporting a 15.1% career average, Hudler is clearly either a sharp-shooter or a "right place/right time" kind of player. Another mold breaker is Alex Ovechkin. He led the league with nearly 400 shots, scored 53 goals, while sporting an impressive career SH% of 12.4%. The other "superstars" like Crosby (14.4%), Malkin (12.8%), and Getzlaf (12.4%) all exceed league averages by a large percentage.

How does this make Ovechkin a "mold-breaker"? What mold? He shoots a lot. He scores a lot. He shot percentages a lot. This is proving your argument.

Without a doubt, there are those scorers, often higher volume shooters, where this relationship is more tenuous. Take Henrik Zetterberg for example. In a number of years he has shot well below not just his personal career shooting percentage but also below league average and still posted 70, 80, 69, and 66 point seasons. Just this past year Claude Giroux had a pretty successful 73 point season shooting at 9%. Even Pavel Datsyuk is a bit of a riddle with his career SH% of 14.4%, but his two best years he shot below his career average at 11.7% and 12.9% (hitting 97 points in both of those years). He was, however, well above league averages.

These are set-up men. These are players who enjoy and will occasionally employ an action that can take place during a hockey game called a "pass". Often, passes can lead to goals. There's no riddle to it. I know you understand this. I'm losing my mind. I'm going to break out in hives.

The shooting percentage statistic seems to be a pretty fair indicator of individual success but less a measure of team success. Take the Hurricanes two most recent post-season years. During the Cup run they had the second highest shooting percentage in the league at 11.2% while the league was averaging 10.3%. However during the 2008-09 season the Canes shot a mere 8.8% vs. a league average of 9.6% (good enough for 25th in the league). Yet, in the last 10 years the Hurricane's team shooting percentage has only been in the NHL's top 10 twice, the previously mentioned Cup year and 2009-10 when the Canes were 10th. Could there be a correlation to team success there?

I'm going to paraphrase so that maybe I can cook up some logic:

-Shot % does not tell you how well a team did.
-When the Hurricanes were good, they had a great shot %.
-In the smallest season sample size of 1 season, they had a low shot %, but made the playoffs.
-In a WHOLE FUCKING DECADE, they've shot like blind grandmas playing croquet and have been a crap team for nearly the entire duration.
-Of course there is.

BONUS FUN!
2014-15 Calgary Flames 2nd in the league in shot percentage
2013-14 Colorado Avalanche 2nd
2012-13 Toronto Maple Leafs 1st

These teams all sucked, but made the playoffs because they shot at a percentage which is unsustainable compared to their possession numbers beyond 80-90 games.

I've concluded that shot percentage shows you who got lucky and who will probably not be so lucky next year. I mean, look at this.

And yet...




So what does all of this mean for the Carolina Hurricanes in 2015-16? Shooting percentage can indicate a lot of things: the quality of the chance, the accuracy of the shot, the movement of the puck, the quality of the playmakers, and the overall creativity of the team.

The 2015-16 Carolina Hurricanes: The NHL's First 5-on-0 Showdown.

There is a difference between playing the Oilers and the Rangers that might help or hurt your shot %. Maybe not. What do I know?

Here is one final set of thoughts to ponder. If Jeff Skinner and Eric Staal merely performed up to their historic career shooting percentages, they would have posted an additional 26 points between them. If Riley Nash hits his career average, he's a 30-35 point player.

His career average is apparently 5-10 points higher than his career best. Figure that out.

More importantly if Jordan Staal comes close to his career shooting % rate, he's at or around the 50 point mark. Those performances alone won't likely get the Canes into the playoffs, but it will make a big difference. Add in moderate improvements from Rask, Lindholm plus a full year of Nestrasil and this team can be offensively competitive. In the best Jim Mora voice that can be mustered, no promises were made about "Playoffs".

If Jordan Staal plays at all, it'll help. Not much, but he's better than Jay McClement.

There are two things that will help Carolina. Better players is one....hmmm...no matter how hard I try I just can't seem to search for the other.