Hey there, EOTP readers! Like most of you, I’ve been greatly enjoying this current midterm edition of Top 25 under 25. It’s great to see where our young stars and prospects rank compared to each other, and in my opinion, equally interesting to see where the EOTP writers tend to agree or disagree with each other concerning these rankings. I took a look at this year’s aggregate midterm T25U25 scoring, and compared those scores to how those same players were rated a few months ago, and to what extent consensus on these players’ scores varied.
First off, here are the current (i.e. most recent) Top 25 Under 25 players, ranked in order of the most recent series. Points represent the arithmetic mean of each of the individual ranks assigned by the EOTP writers; error bars represent the standard deviation: the larger the bars, the less consensus between contributors. Obviously, some players’ ranks went significantly up or down between now and the last series, and some players became less (or more) controversial.
To visualize the changes over the past few months a bit better, here are the same players but ranked in order of their relative change since last year:
As you can see, there were some big risers in the ranks (Reway, Bournival, De La Rose, Lehkonen, and Andrighetto are all unsurprising entries here) as well as some drops- none so large as our most recent first round gamble, Mike McCarron. Overall, you’ll see far more rise, on average, than falls- why is this? Two reasons. First, Max Pacioretty graduated, bumping everyone (except PK and Galchenyuk) up by one rank by default. Second, those who dropped off the Top 25 entirely (Patrick Holland; Mac Bennett) don’t show up, and were replaced by new entrants Tokarski and Reway, who show up as modest and very large gains, respectively. (Since neither Tokarski or Reway were ranked by the whole panel last time, I assigned them both a single value of 30th for comparison purposes.) Altogether, no enormous surprises- but it does highlight the 6 players who slipped in their rankings more than one rank slot (the one Patches left behind): Hudon, Dumont, Bozon, Dietz, Ellis & McCarron.
Now, for a more interesting graph. Here are the same players, but with graphing the Coefficient of Variance (SD/mean value) of their T25U25 scores and ranking them by who changed the most, consensus-wise:
A few things to note here. First of all, Subban’s CoV went from zero to zero (as he was unanimously ranked #1 both times), which is why he has no bars at all. The way CoV is calculated means that a spread in scores between, say, 2 and 4 (like Galchenyuk last time) will be more significant than a spread between 18 and 24 (a Dalton Thrower type, for example). This explains why Galchenyuk went from a relatively high CoV (above 0.3, good for fifth-highest last time) all the way down to zero, for the largest gain in consensus of the group.
The next three biggest gains in consensus belong to Artturi Lehkonen, Magnus Nygren, and Christian Thomas- which is interesting to me, as these three players struck me as having different sources of disagreement last time around. Lehkonen, as the undersized but extremely skilled recent draft pick; Thomas, the recently-traded-for and less well-known commodity; and Nygren, who’s been around for a little while but we still couldn’t seem to agree on. All three seem to have fallen right in line, as the T25U25 judges seem to have consolidated their opinions greatly around these three.
Overall, there was far more consensus between the judges this time around than last time. Why? Well, partially due to there being two less judges- although that shouldn’t affect the score very much. My guess is that this is actually an interesting form of regression to the mean. It’s hard to imagine that the judges’ votes this time around weren’t influenced by both their own as well as their fellow judges votes last time- and if you were to re-do this voting over and over again, most likely the scores would regress down to, or at least towards, a general consensus. That’s my theory, anyway. Along this line of thinking, note that Reway and Tokarski are both absent from this graph, as we don’t have a way to compare their two consensus scores- but I would expect them both to drop when the 2014 offseason series comes out.
As a matter of fact, only three players- Beaulieu, Tinordi, and Collberg- became more controversial this time around (although Beaulieu’s increase was very slight). This was quite interesting to me, as I would consider all three of them to be stud prospects that are being watched very carefully.
Finally, it’s worth noting that of the players who were ranked worse this time around (Hudon, Dumont, Bozon, Dietz, Ellis & McCarron), overall the worse a player got, the greater a consensus there was around that demotion. (Looking at you, Mike.) I guess it’s up to them to prove us wrong. Keep up the great work, everyone at EOTP- I’m looking forward to the next series!