January 31, 2011
(Ed. Note: Welcome to special edition of Stat Nerd Sunday on Monday, where we occasionally obsess over hockey numbers like a Dungeon Master obsessing over the level of his warrior elf. Here's Matt Barr, formerly of LCS: Guide To Hockey and Trolleytracks and now blogging hockey at Kertwang.me.)
This All-Star Game business, it's showcasing the players having the best seasons to date. Mostly. There's some other stuff going on, like getting every team involved one way or another, but the idea behind it is to get the best players so far this year together for a weekend of televised fun.
So, how do these best players in the first half do in the second half?
Let's take a look at every player who played in any of the last five All-Star Games. Because of Olympics and lockouts, those were during the 2008-09, 2007-08, 2006-07, 2003-04 and 2002-03 seasons. For each All-Star season, let's split their stats into pre-All-Star Game and post-All-Star Game.
Obviously, the "first half" is really longer than the "second half," when you use the All-Star break has your halfway point. So to compare pre- and post-All-Star Game stats, we have to break actual games played, goals scored, plus/minus and so on in each segment into rates. Since that gives us a bunch of fractions, what we'll do is prorate each "half" into an actual half a season. So what we're looking at is pre- and post-All-Star stats each expressed in 41 game segments.
Note that because of rounding averages, not everything in the lines below will add up.
If anything the consistency is remarkable. There's a very short, maybe half a shift, drop in ice time per game, which is probably enough to explain the slight reductions in everything else.
Is this any different than if the player just took the All Star break off, and went fishing?
One way we can look at that is by looking at players who both played in and didn't play in All-Star Games the last five years they were held.
I picked 10 players who: (a) played in either two or three of the last five All-Star Games; (b) were around in all five of those seasons (that is, didn't come up in the last few years or retire a few years ago; (c) were basically healthy for their non-All-Star seasons, not passed over because of a long first half injury, say.
The 10 players: Daniel Alfredsson(notes), Pavel Datsyuk(notes), Shane Doan(notes), Marian Hossa(notes), Ilya Kovalchuk(notes), Alexei Kovalev, Patrick Marleau(notes), Scott Niedermayer(notes), Brian Rafalski(notes) and Kimmo Timonen(notes).
Now, here's now those 10 guys did, pre- and post-, during their All-Star Game seasons and non-All-Star Game seasons (looking only at the years they held the last five All-Star Games):
There's a very slight drop-off during their All-Star years, which looks familiar. Again, the reduction is really negligible, and probably completely accounted for by the 31-second drop in ice time per game. The non-All-Star years? As identical as can be.
Now I wonder if this is the same for All-Star veterans and All-Star newbies. Let's clear the pool of players from the last five All-Star Games of everyone who played in three or fewer. We're left with eight players who played in either four or all five games: Todd Bertuzzi(notes), Zdeno Chara(notes), Jarome Iginla(notes), Vincent Lecavalier(notes), Nicklas Lidstrom(notes), Rick Nash(notes), Martin St. Louis(notes) and Joe Thornton(notes). How did these routine All-Stars do in each half of their All Star seasons?
Like our group of 10, these guys dropped off very slightly, and the whole thing can pretty well be explained by a small drop in ice time per game. How about 38 younger players new to the All-Star experience?
It's nice to be able to stop worrying that participating in this year's All-Star Game is going to wring the life out of some of these guys for the last 30 games of the season.
Oh, how about goalies? Same thing, really:
You have my permission to refer to this data as soon as a sportswriter tries to explain a late season slump of participation in All-Star weekend. We all have to stick together.
Do award winners need a big second half to win?
While I was at it, I wanted to take a look at whether eventual post-season award winners generally had a more impressive or less impressive post-All-Star Game season. I ended up checking out the non-goalie, non-Olympic year Hart (MVP) and Calder (rookie) winners since 1997-98, when ice time data becomes reliably available.
I didn't look at the Norris or Vezina winners. Any data on the Norris winners since 1998 would tell us more about Nick Lidstrom than it would about defensemen in general. Likewise with the Vezina and Marty Brodeur and Dominik Hasek(notes).
Anyway, the MVP generally cranks it up after the All-Star Game. The Calder winner, not so much:
Do league leads in goals and points at the All Star break hold up?
The post-All-Star segment of the season is like Double Jeopardy, Hans: the scores can really change. Leaders in goals and points at the break and at the end, last five All-Star Game seasons: