There are a handful of models that predict pick value. I back-tested two popular ones, The Athletic and Puck Pedia. There are a number of ways at measuring these models. This is simplest way I can put into an ACGM.
I took a bunch of recent trade downs where two picks were exchanged for one. For each trade, I locked in the two highest picks as constants, assigned them the appropriate values, and then used their model to predict the third pick.
For example, at the 2015 draft, Tampa Bay traded down from #28 to #33. Using the two models, they offered two suggested for what the Islanders should be adding to #33 to jump up to #28. The Puck Pedia model suggested the 72nd pick was required. The Athletic model said it took the #92 pick.
Remember the first picks are locked in as constant, so we are only predicting that third pick. In this case, the PP model got correct. The Athletic model was 20 draft slots off.
Here is another example from 2005 when Atlanta traded back from #12 to #16. PP Model said NYR had to add the 43rd to pick to jump from #16 to #12. The Athletic said all it took was 73rd overall.
The Puck Pedia model is much better predictor for these picks. On average, they were about 3 draft slots off, while the Athletic was about 28 draft slots off (almost a full round).
Drafting is even more of an imperfect science in the NHL than the NFL and NBA, just doesn't feel worth moving up most of the time, and this is coming from a team that benefitted quite a bit from a trade up (2020 for JJ Peterka)
Sound of hockey pick value chart puts the Chicago trade at #10OA being worth 354.44 and the collection of Hawks picks being 420.49. the NJ 1st for the Tampa 1st and Chicago 2nd are almost right on in their value chart. I think the higher pick should command a premium and "win" the trade, but probably not by this much. That said, I think the Hawks are open for business to consolidate lesser picks into better as they have a lot of picks (7 in the top 100 currently).
@nhlfan10506 always appreciate your analysis and envy your free time
FWIW that package seems steep to me from the STL side but 16oa is starting to look like no-man's land in this draft whereas a much-needed impact RHD should be there at 10. I'm thinking the Blues should move up or down but we'll see how it unfolds.
Sound of hockey pick value chart puts the Chicago trade at #10OA being worth 354.44 and the collection of Hawks picks being 420.49. the NJ 1st for the Tampa 1st and Chicago 2nd are almost right on in their value chart. I think the higher pick should command a premium and "win" the trade, but probably not by this much. That said, I think the Hawks are open for business to consolidate lesser picks into better as they have a lot of picks (7 in the top 100 currently).
I have never heard of sound of hockey.
But the whole point of this exercise was to ignore my gut was telling and jump at the data.
But the whole point of this exercise was to ignore my gut was telling and jump at the data.
Would have to back test the #.
Whenever I have played with pick values, most point me to sound of hockey as the chart that is most realistic and I tend to agree. Most other models overvalue later picks particularly late 2nd and certainly later. For example, if you use the Athletic chart, it values early 3rds at 1 pt and a late 1st at 3 (or as low as 2.7 for #32OA). I am pretty confident no one is ever trading even the lowest 1st for three 3rds. By contrast, the sound of hockey would require 9.5 of the highest 3rds to get the #32 OA. That part of their model makes complete sense intuitively if you conceed that most meaningful NHL players come out of the 1st or 2nd round and the odds are pretty long for lesser picks.
Jded liked this.
Reply
Holtz,
CapFriendly
CapFriendly