clock menu more-arrow no yes mobile

Filed under:

Islander Analytics 2014-15 Review Part 2: Brock Nelson

What should we take from the Rev's sophomore season with the Isles? Should we be impressed or disappointed? Let's look at the analytics to find out.

The Reverend had an up and down year.  But he's only 23 and should get better.  And he's already good.
The Reverend had an up and down year. But he's only 23 and should get better. And he's already good.
Anthony Gruppuso-USA TODAY Sports

Coming into this year, New York Islanders fans had high expectations for Brock Nelson. At age 22, in his rookie NHL year, Brock had what was considered by many to be a fantastic year. For one, he was a big possession driver in his rookie year - basically every player on the team had a better Corsi with Nelson than without, as the below chart from February of 2014 (so not the full season) showed:

The trend in that above chart continued through the end of 2014.

In addition, Brock put up 26 points and 14 goals despite spending a large amount of his time on ice with mediocre scoring linemates - his most-common linemate in 2013-14 was hot shooter Cal Clutterbuck.  At age 22, there was solid reason to believe Brock could turn out to be a 40-point, strong plus possession center which every team would like to have.

His 2014-2015 season however, was a bit of a mixed bag. Brock exploded scoring-wise at the start of the season, scoring 10 goals in the first 20 games, and then promptly scored only 9 goals for the remainder of the season. He scored 19 goals, but actually scored goals at a lower rate than in 2013-2014 at even strength. In addition, Nelson's possession impact dropped - he was still a plus possession player, although not by much, and a bunch of players had better numbers away from Nelson.

So what happened?  Should we be concerned?  

First of all, Brock Nelson is 23. So, he is likely not done improving as a hockey player - hockey players tend to peak around 25 - if you want to argue that peak stretches to 27, be my guest, but the improvement in GENERAL continues till around age 25. This doesn't mean that he will improve or that he won't improve past that point of course - but if he's like most players, he's still going to improve for at least a season or two.

Second, Brock's individual numbers were the result of a few factors out of Brock's control, that made his season look a bit disappointing in the end.  First of all, his hot start was the result of insane shooting #s more than anything: In his first 10 games, Nelson had a shooting % over 25%. No player in today's game can sustain shooting at such a high rate for long, and Nelson was not likely to be an exception.

Nelson even managed to break 20% shooting over a 10-game stretch around the 28 game mark as well, but his shooting % promptly collapsed.  You can see the 10-game rolling average of Nelson's shooting % over last season in the chart below:

Shooting Percentage is a statistic that analytics research has long stated basically regresses to the mean. While some players do show better shooting skill than others, shooting % over shorter periods (even as long as a whole season) is highly variable and nearly entirely random. After all, even when a player takes a shot from a dangerous area, that shot going in is based upon things such as: the goaltender's actions, the movements of players in the way of the shot, or just human variation in repeating an action over and over.

We look at shooting percentage because though random in the short term, it tends to regress to the mean over time.

All that means is that a player's shooting % is often just luck of the good or bad variety and over longer terms, it regresses toward the mean for that player.

A Tale of Shooting Percentage

In Nelson's case, Brock had insanely good luck for the first 33 games or so, and then promptly suffered from horrible luck instead, even going 16 games in a row without a single shot going in.  And during this time, Brock was actually taking shots at a respectable rate, as the below chart shows:

Basically Nelson's first few games featured him scoring on every shot he took, with him not taking many shots at all. But over time, he settled in at a solid rate of above 2 shots per game. That's not a first line rate, but it's an average 2nd or 3rd line shooting rate.

In addition, Nelson's power play time, where one gets the most shots, dropped quite a bit as the season went on, reducing his odds of getting shots (and high percentage ones at that).  At even strength, Nelson averaged 8.01 shots per 60, which was good for 91st in the league amongst forwards with at least 500 minutes played. Basically, Nelson averaged shots at a borderline first line/above-average-2nd-line rate.

So Nelson's shot #s got MORE impressive as the season went on, he just suffered horrible luck shooting, which eventually actually made his shooting percentage lower than it was in his rookie season by a smidge, and by a decent amount at even strength.  The odds of Nelson putting up a 20-goal, 40-point season yet again would seem pretty decent, even if he doesn't improve, although if he gets less power play time, that might hurt that effort.

A Possession Dip

Possession-wise, Nelson DID in fact get worse last year than in his rookie season. The following graph shows Nelson's corsi compared to the team's corsi without him on the ice over the last two years:

This graph, courtesy of the great Muneeb Alam, showcases how much better the Isles did with Nelson on the ice last year than without him - see those large peaks. But in 2014-2015, Nelson was much less consistent, starting by performing above average, and then performing below the team's average for a while, before moving above average near the end (as the team's average fell without him).

Nelson went from being a clear strong possession driver to a guy with ups and downs, and was overall just slightly positive.

Still, there are mitigating factors there too: Nelson's teammate quality did improve last year, but he was used in a far more defensive role than in his rookie season: he went from a 49.8% zone-start in his rookie season to a 43.2% zone-start last year, meaning he was starting a greater # of his shifts in the defensive zone last year than the year before, despite the team as a whole spending a greater percentage of its time in the offensive zone. Now statistically this shouldn't matter too much - the effect of zone starts is far less than what we previously believed. But it's worth noting that expecting a slight drop in Nelson's #s could've been expected.

In addition, again, Nelson is 23. He's just faced a season after the league first got a look at him and still came out roughly positive, even if not as positive as he used to be.  Odds are, he is going to improve possession-wise next year, just as he likely will as a scorer as well.


In short, disappointment with Nelson is mostly overblown. His scoring drop was pretty much entirely attributable to a change from good luck to bad luck and a loss of power play time. Meanwhile, his possession play did in fact drop off, but was still positive, and it stands to reason that will improve next year as Nelson hits age 24. Basically, there's a good chance he emerges next season as an average or above second liner, which would be big for this team given that he'll probably be playing 3rd line center.