True, this article is a bit late, now that we’re 13 games into the season, and three games into November, but it’s still an issue worth looking into: does Dan Hamhuis take a while to get revved into full gear during the season?
The question comes from when Botchford dropped this little nugget in an edition of his Provies a few games back:
"“Hamhuis has been struggling, as he often does to start a season.”"
It got me wondering about just how true this could be–was this a case of Vancouver media’s short memory, or an actual long-term trend? I fully remember last year being a rough go for Hamhuis in the earliest games, but I mostly chalked that up to adjustments to Tortorella’s system, and guessed that this year was more of the same. But what about the seasons prior to that? I had to admit, I couldn’t remember much at all about Hamhuis’ early play since joining the Canucks, whether horrendous, terrific, or just average.
So to find out if Botch was right, I did some investigating into the first ten games of each season Hamhuis has played since joining the Canucks. And now I’m back here to give you the findings. Turns out, the answer is…(drum roll please)…both yes and no, but in a much truer sense, not at all.
Allow me to explain, through use of this lovely chart:
What we learn from Hamhuis’ basic counting stats (+/-), is that his Octobers (and one January–thanks, lockout!) have been way less successful than the rest of his career. Of course, +/- can be a suspicious stat, because there are so many factors outside of a players’ skill that can go into a goal. Bad bounces, unskilled teammates, poor goaltending etc. can all unfairly influence a player’s +/- stat line. However, over the span of a career, you’re going to find that the players in the +’s are generally better than the players in the -‘s, because they find a way to overcome those things over a large sample. Looking at his career Canuck numbers, this is true as well for Hamhuis. Of all the games he’s played in a Canuck uniform, he’s posted a + 51, which corresponds with the general perception that Hamhuis is a good player. This is in direct contrast to the relatively horrid -17 he’s been stuck with if you only count his first ten games in all seasons. This is probably the most significant reason Hamhuis has been perceived to start his seasons slowly: he’s been on the ice for way more Goals Against in Octobers than he has been on ice for Goals For, meaning that he’s more often than not been seen at the scene of the crime when something has gone wrong.
But what about the underlying numbers? If we look further down the chart, we can see that, while Hamhuis’ +/- is skewed against him in the first ten games, his Corsi For% is almost indentical in his first ten games to his career average. (Off only by 0.977%, which, while not nothing, isn’t terribly significant). This means Hamhuis has always been driving possession from start to finish in a season, regardless of what month it is. This suggests that the gap in his +/- and perceived poor play is actually a mirage.
This is backed up by the last group of stats in our chart. PDO is a stat that measures puck-luck, by adding a team’s shooting % and their save percentage. By this measurement, anything over 100 is lucky, and anything under is unlucky. Over a long enough stretch of time, any lucky or unlucky play usually regresses back to 100. Now, in using this for players, this means the shooting and save percentages of the team when that player is on the ice. PDO can fluctuate more with players than with teams, but the principals are basically the same–luck tends to regress to 100 over a long enough time. As we see with Hamhuis, over his career, he’s actually been a bit lucky at 101.5. The same can’t be said about his play in the first ten games. For whatever reason, Hamhuis has a miserable PDO of 96 to start seasons, a mark which is over 5 points worse than his career average.
So what does this all mean? First of all, it shows us that Hamhuis has actually been basically the same player he’s always been. He’s just been consistently victimized by weird bounces in small sample sizes during times of year when more people are paying attention, than say, a random ten game stretch in February. Overall, this shows us exactly why +/- is a stat worthy of wariness, because of all the other factors that can play against our eyes: we see that Hamhuis has continued to generate shot attempts and possess the puck at the same rate he always has, and we see that the random outside factors that have plagued him in small but visible doses have eventually evened out over the course of his career.
Mostly, this is another great reason to dig deep into the why’s and wherefore’s of any assessment made on a player’s performance. More often than not there are more factors at play than we can perceive with our eyes or even our memories at one time, and that’s why having so many great stats, tools and resources are valuable when trying to determine the true nature of a player’s performance.
Either that, or Hamhuis accidentally got some kind of Halloween curse and he’s just permanently doomed in the year’s spookiest month. Either way, I’m sure we’ll all be glad for November and a return to the old, slightly-lucky, totally awesome Hamhuis we all know and love.