News

Stats say Maple Leafs & Coyotes should look at trading down in draft

Dom Luszczyszyn
By:

Author: The Hockey News

News

Stats say Maple Leafs & Coyotes should look at trading down in draft

Dom Luszczyszyn
By:

Just like every year, the idea of trading down inside the top 10 has become a hot topic. Arizona says it's open to trading the third overall pick, while Toronto has been rumored to be interested in doing the same with fourth overall. Is it worth it, though?

Just like every year, the idea of trading down inside the top 10 has become a hot topic before the NHL draft. Arizona GM Don Maloney has stated he’s open to trading the third overall pick, while Toronto has been rumored to be interested in doing the same with fourth overall. Is it worth it, though? That’s likely the question both teams are asking themselves leading up to Friday and with each one taking the scorched-earth approach to rebuilding, whether or not to trade down is an important question to ask. Generally, the answer is yes, but it’s harder to answer the higher the pick is. It really depends on the context. For these two clubs, the context is the luxury of drafting a franchise-cornerstone prospect in the top five. That’s difficult to pass up, because those types of players are usually only found in that top five. At the same time, it's hard to ignore the assets they can get in return for moving down a couple of spots. Historically, that’s usually been a second round pick plus, but the price is likely higher for picks within the top five, especially considering this year’s crop of talent. What it boils down to is knowing how much a pick is worth and how much the prospects available are worth. That value really depends on what a team wants from a draft: high-success rate or high-impact players. The simplest method for that is going through past drafts and finding the likelihood that a pick turns into an NHL player (the common threshold is 200 games played) and then finding the average points-per-game against positional average for those players per draft slot. Points-per-game isn’t everything, obviously, but it’s likely the simplest way of showing impact. Here are the first three rounds of every draft from 1988 to 2007. (Click on the image to enlarge.)  
1
  Of course, that’s on average over a 20-year period. Every draft is different, which is why it’s important to look at the players available as well. There have been big strides made recently in draft analytics with the most important one coming out of Canucks Army, a Vancouver Canucks blog hosted by the Nations Network. What they did was create a network of comparable players from past drafts based on their age, height, scoring and what league they played in to get a basis of how likely a prospect is to succeed. The model, called prospect cohort success percentage, or PCS%, is not meant as a tool to replace scouts, but merely to produce supplemental information that aids the process in finding NHL caliber talent. PCS% is in its infancy and not publicly available yet, but it still gives a decent frame of reference for the current crop of talent waiting to be drafted Friday and Saturday. Comparing that to THN prospect-expert Ryan Kennedy’s draft list gives a decent landscape of the value available in the top 10. One caveat with the model is that it currently underrates a majority of players because it’s not era-adjusted (yet) for increases in scoring rates over time. I’ve added a small adjustment factor to put it on the same scale as the historic probabilities of success.  
4
  What’s immediately noticeable is how close the forwards and defensemen are in the 3-6 group, which is a big reason why trading down becomes an attractive option this year. Even at 6th or 7th overall, there will still be an elite blue-chipper likely available. In fact, it’s possible to check that too. A blogger going by the pseudonym ‘DTMAboutHeart’ recently unveiled a tool that estimates the chance that a player will be available at a certain spot based on an aggregation of draft rankings from industry experts and how right they’ve been in the past. Here’s how things are likely to shake up for those four.  
2
  It’s reasonable to assume Dylan Strome and Noah Hanifin don’t fall very far and the likeliest outcome from trading down is acquiring Ivan Provorov or Mitch Marner, each of whom still has a chance of being there as far down as seventh overall. That’s where it comes down to the scouts. If they view Marner and Strome as anywhere close to equals, or Hanifin and Provorov, then trading down becomes a no-brainer because it means more assets to go along with an almost equal prospect. Trading down any farther than that, however, is where things get problematic, as the rest of the group likely isn’t at the same level. What this all points to is a large inefficiency waiting to be exploited. A couple of years ago, former blogger and current Carolina Hurricanes employee Eric Tulsky looked at draft-pick value in the scope of marketplace perception from past draft pick trades. Comparing it to other models based on different kinds of value shows how large the gap is between perception and reality.  
3
  Anyway you slice it – points-per-game, minutes played, point shares, career games played – it’s clear that draft picks are being undervalued by GMs. The first team to catch on will likely reap the dividends – maybe that’s Arizona or Toronto this season. But if history is any indication, you can expect those two teams to stay right where they are. Whether that turns out to be the right move is something we won’t know for a while.
Comments
Share X
News

Stats say Maple Leafs & Coyotes should look at trading down in draft