Ramblings: Montreal’s Latest Moves, Colin and J.T. Miller, and More (July 15)
Montreal’s Latest Moves, Production and Opportunity, a Duo of Millers
This is only tangentially related to fantasy hockey, but I thought there was some interesting news to come out of the Montreal Canadiens on Wednesday night. This type of information isn’t always big news, but in this case I think it is:
Pfeffer, who joined the Canadiens at the beginning of the 2015-16 season, was relieved today.
— Eric Engels (@EricEngels) July 13, 2016
Mr. Engels would follow a few hours later with this:
Montreal Canadiens clarify that Matt Pfeffer's 1-year contract was not renewed + that they will be pursuing other options re:analytics hires
— Eric Engels (@EricEngels) July 13, 2016
And so ended the one-year tenure of Matt Pfeffer as the stats and analytics department for the Canadiens.
For those unfamiliar with his work, Mr. Pfeffer’s background is included in this piece from Yahoo three years ago. Since, he has done work for the San Jose Sharks, and until recently, Montreal.
I first started getting into the advanced stats in hockey about five years ago (maybe a little less). At the time, Pfeffer’s name was one of the first people getting into stats would learn, along with names like Dellow, Charron, and Tulsky (all hired by NHL teams since, by the way). From those whose work I read, Pfeffer was highly-regarded, and his transition to the NHL seemed a matter of if, not when.
Engels has a piece on Sportsnet that indicates Pfeffer was not renewed after his one-year deal, at least in part because of his opposition to the P.K. Subban/Shea Weber trade. That would be pretty shameful by the Habs, but it’s not really the most concerning part. There are a few other problems that arise from this situation:
- The assumption is, given what we know of his prior work, that Pfeffer is really good at what he does. Eschewing his information isn’t a poor decision because it’s eschewing what is likely good information, it’s a poor decision because it’s eschewing a dissenting opinion. The point of advanced stats, and the resulting decisions, is to challenge preconceived notions. The point of advanced stats, and the resulting decisions, is not to affirm preconceived notions. It may in some instances, but it isn’t necessarily a goal. This indicates that nothing has materially changed with the way the Habs do business over the last several years.
- Giving an “analytics department” one year to influence decisions on a professional sports team is hilariously awful. One year of work cannot possibly provide the level of impact on decision-making that is necessary to determine the quality of information provided, whether good or bad. How long are general managers given to fully shape a team into their image? Three years? Five? It’s certainly not one.
- The Habs are now going blind through the summer. This is the time of year when analytics departments get their work done for the upcoming season.
There is a lot that we will never know about this entire situation. All we can do is try to piece together what is made public, and judge the moves that are made.
To circle back, this is a great anecdotal story for fantasy owners. It’s one thing to have information at the ready to impact decisions, but it’s quite another to actually use this information to impact decisions. What’s the point of digging up shots-per-minute rates only to forego them in favour of grabbing a favourite player off a hometown team. If just grabbing favorite players off a favourite team is the goal, then fine. If that’s all that a person wants out of their fantasy hockey experience, all the power to them. It is, after all, about individual enjoyment.
Pfeffer’s story is a cautionary tale for fantasy owners, though. Have the information, know what the information is (and isn’t), and use it consistently for years. Using new information for a home league one year, and not winning that league, isn’t an indictment of the information. There is lot of randomness built into sports – hockey in particular. It takes a commitment to change the entire decision-making process
Good luck to Mr. Pfeffer moving forward. Knowing the work he’s done in the past, it won’t be long until a smart team picks him up.
Piggy-backing a bit on what was written above, Mr. Engels did a lengthy interview with Subban, and it is available on Sportsnet’s site. It is recommended viewing for anyone looking for a bit more insight into Subban, the trade, and his time in Montreal.
Continuing along with some thoughts on advanced stats, it’s crucial to keep in mind that being able to translate the rate stats into what’s meaningful for fantasy is a necessary skill. Solid rate stats are a nice foundation, but if a player doesn’t have a featured role, or isn’t expected to get a featured role, rate stats are fairly meaningless in fantasy.
Last year, for example, I was very high on Tyler Toffoli. Of all the analysts ranked at FantasyPros, I had him ranked the highest going into the 2015-2016 season, and was just one of two analysts to have him inside the top 110 players (I had him at 77). The reason for this was two-fold: first, his rate stats going into last year were excellent, and secondly, I thought he would be given a bigger role on the Kings. This came to fruition, and in roto leagues, he had an excellent season. I had this same thought process for Vladimir Tarasenko leading into the 2014-2015 campaign.
This is how important it is in fantasy for players to have a featured role – among the 141 forwards that Hockey Reference’s Play Index has playing at least 1400 total minutes in either of the last two seasons:
- Just 7.8-percent (11/141) finished with 40 points or less.
- About 76.6-percent (108/141) finished with 50 points or more.
- A little over half (72/141) finished with 60 points or more.
Quite literally, blindly picking players a fantasy owner thinks will be A) healthy, and B) given a top line feature role gives them about a 50/50 chance at 60 points. That is pretty remarkable.
Conversely, again using Hockey Reference’s Play Index, of the 119 players to play at least 75 games and finished with 1200 minutes or less of ice time in either of the last two seasons:
- None finished with 50 points or more (Toffoli was closest with 49 in 2014-2015).
- A shade under 46-percent (54/119) finished with 25 points or less.
- About 6-percent (7/119) managed a 20-goal season.
Even in the spread between the two limits presented, so between 1200 and 1400 minutes, there were 165 players, and:
- Just 11.5-percent of them finished with 60 points or more.
- Under one-third (53/165) finished with 50 points or more.
- A shade over 40-percent finished with 40 points or less.
From a truly featured role (over 1400 minutes) giving about a 50/50 chance at 60 points, to a less-than-featured role providing a little better than a 1-in-10 chance at 60 points, it’s clear how much opportunity matters. So while writers/owners (myself included) will use rate stats a lot to determine a player’s prior performance, in the end, a lot of fantasy production comes down to the role a player is given. It is important to be able to distinguish the two, and identify situations where a player with solid rate stats will be given that opportunity. Opportunity often leads to production, even when the underlying asset (the player) isn’t deserving of the opportunity given.
One signing that caught my eye recently was Boston giving two years to restricted free agent defenceman Colin Miller with an average annual value of $1-million. Miller managed 16 points in 42 games for the big club, bouncing between the NHL, AHL, and press box.
Miller played out his full junior career, so he never hit the AHL until he was 21 years old. In his last two regular seasons for Manchester and Providence (he was part of the Milan Lucic deal), Miller had 64 points in 90 games. Despite Miller being a bit older for what we consider to be a prototypical prospect, his development was really just delayed. There is offensive talent here – read his Dobber Prospect profile.
The concern with Miller’s fantasy value will be – and this is the theme today – his usage. He was given fairly sparse time when he was with Boston last season, and found himself as one of the popcorn boys a few times as well. This is all part of bringing along a young, non-elite defenceman slowly, but let’s face it, this Bruins defence corps is awful. How many guys do they have signed not named Torey Krug can this team rely on next year?
Having a poor defence rotation should be beneficial to Miller, if only for power play time. Krug obviously will log heavy top power play minutes, and the Bruins will probably try to wring whatever value they can out of Zdeno Chara by using him on the second pair. Maybe Miller is the third option this year?
I’m not sure Miller has fantasy value right out of the gate in October, but I think eventually he presents himself as, at worst, the number-3 power play option. Heck, he could even end up on the second even strength pairing. He probably won’t be worth drafting in the majority of one-year leagues, but he is definitely a guy to put on a watch list
Another signing that was of interest, in particular because of the team he will be hitched to, was the Rangers coming to terms with RFA forward J.T. Miller. His AAV for the two-year deal is $2.75-million. In yesterday’s Ramblings, Neil Parker had a bit to say about the signing, and Miller’s fantasy prospects for next year.
Projecting Miller is going to be tough just because, in my opinion, projecting the Rangers as a whole will be tough. Outside their top-two centres, what is the lineup mix going to be? Will Chris Kreider be locked on the top line? Second line? How long is Rick Nash in a Rangers jersey? Is Kevin Hayes going to get a big bump? The biggest question of all: they’re not actually going into next season with that defence corps, are they?
The only reason the Rangers scored so well last year is the team, as a whole, shot nearly 9-percent at five-on-five. That’s two years in a row shooting fairly high. Now, it might be a pattern, but then again, as we’ve seen in the last four or five years from teams like Anaheim, Colorado, and Toronto, having a high shooting rate for a couple years doesn’t really mean a whole lot.
The Rangers didn’t generate a lot of chances last year, but finished a lot of the ones they did. I’m not buying it. This team has lost Keith Yandle, will probably lose Rick Nash at some point this year to a trade, and was struggling last year as it was to generate offence i.e. scoring chances, as it was. There are nice pieces among the forward crop, but they won’t be nearly as productive offensively as a whole next year. Unless Miller is consistently over 17 minutes a game, I’m not sure there’s much fantasy relevance outside of deeper leagues.
No data at this moment.