Hockey Twitter wasted no time getting back to work.
With the first piece to the Leafs 2016 draft puzzle put in place after the team won last night’s lottery, attention quickly turned to Toronto’s later 1st round pick via Pittsburgh as well as the team’s own 2nd round pick which sits firmly at 31st overall.
Specifically, the question was raised about whether or not this was the year it would make sense to trade up from those two spots in order to acquire another higher-end prospect, possibly a defenseman.
My initial instincts were to jump into the argument and tell the people that thought trading up was a good idea that they were wrong. After all, don’t the numbers tell us that it’s typically a bad idea?
Rather than get myself involved in the argument, I decided to do some fact-checking to see whether or not my initial instincts were right, and when (if ever) it’s a good idea to trade up (or down) in the draft.
Here’s where my research brought me:
ERIC TULSKY ON TRADING UP
If you’re well-versed in advanced stats you know the name Eric Tulsky quite well. If you’re not, just know that he’s a former blogger that wrote a lot of great articles that delved deeper into the use of numbers in hockey, and that he now works for the Carolina Hurricanes as a Hockey Analyst.
In this article, Tulsky compiled all 46 of the trades between 2006 and 2012 that involved only draft picks (so, no players) and developed a model to determine the value that GMs were placing on these picks. Basically, he assigned a numeric value to the perceived value of each pick by the NHL’s various GMs. Those values looked like this:
So basically, GMs place a lot of weight on really high picks.
The key finding from his article is summed up nicely here:
By this approximation, three
out of every four trades were roughly fair, meaning that the two teams’
packages were within 20 percent of each other in estimated value.
Moreover, almost all of the ones that were off by more were very minor
moves, where being off by 30 or 40 percent still isn’t very much in
absolute value. Two 6th round picks are probably worth a lot more than a
5th, but I doubt the Flames are losing sleep that they might have given
up too much when they made that trade in 2007.
If we cut our 46-trade sample
in half by focusing just on the bigger trades for a minute, 21 of 23
were fair to within 20 percent, which is pretty reasonable for a model
that doesn’t take into account things like the depth of a given draft.
In other words, most draft trade up/down scenarios end up being fair in the eyes of the GMs. That means the Leafs could trade up in the draft at fair value, right?
Well, not exactly. The obvious problem here is that these values are relative to the weight that GMs themselves put on them, which might not accurately align with the actual value of a draft pick. It’s definitely helpful to know what sort of value GMs themselves place on picks though.
STEPHEN BURTCH ON ACTUAL DRAFT PICK VALUE
Long-time fancy stats writer Stephen Burtch (I know you’ve heard of him) looked at all the draft picks between 1995 and 2007 in a 2015 article for Sportsnet that sought to place an objective value on each and every draft pick. Basically, using expected games played and points per games played, Burtch ultimately came up with the following graph:
Not too dissimilar to what we saw from Eric Tulsky: high-end draft picks are more valuable. That’s a good sign that NHL GMs have at least some clue of what they’re doing.
Here’s what I think is maybe the most important finding though:
Another way to think of this is the comparative worth of picks in
various ranges of the draft. For example the value of a late first-round
pick from a contending team is actually closer in value to any
third-round selection than it is to a top-five pick. Thinking of this
from a glass-half-full perspective for a team trying to rebuild, this
suggests that adding multiple second- or third-round selections is
likely an easier way to improve the roster of a team than holding out
for that tough-to-obtain first-round choice when the market is far more
This, to me, is why it usually makes more sense to trade down. NHL GMs place a premium on high draft picks, perhaps rightly so, but not all first round picks are created equal. Here’s a table from the same article that exemplifies this nicely:
Of course, it’s worth mentioning when looking at that table that when we’re talking about the Leafs moving up in the draft, most people aren’t talking about acquiring another top five pick. But, it’s also worth mentioning that given the value of a 3rd round pick relative to where the Penguins’ pick might ultimately land, trading down from that pick might ultimately end up being more lucrative.
MICHAEL SCHUCKERS’ DRAFT PICK VALUE CHART
A professor at St. Lawrence University and a statistical consultant, Michael Schuckers also crunched some numbers to objectively determine the value of an NHL draft pick. To do this he looked at all the draft picks between 1988 and 1997 and determined the expected value of a draft pick by taking games played into consideration*. Maybe not perfect, but nonetheless it gives us another set of data to work with.
Here’s his findings summed up in one glorious table:
To me this exemplifies why it’s often more efficient to trade down than to trade up. Let’s say the Penguins win the Stanley Cup (a very real possibility) and the Leafs’ two highest picks after 1st overall (damn that feels good to say) sit 30th and 31st overall. Let’s say the Leafs are talking with the Arizona Coyotes, who own a pick at 20th overall courtesy of the New York Rangers, and the two teams agree to a trade that would send the 30th and 31st picks in the draft to Arizona for 20th overall. Seems reasonable, right?
In this scenario, the Leafs would be trading two draft picks with a combined expected value of 520 points for a draft pick with an expected value of 350 points (note: when I say points, I’m not referring to expected NHL production out of that pick). So, the Leafs come out 170 points in the negative, which, according to the above table, also equates to an extra 3rd-rounder.
Is that worth it? Objectively, no. But what if that player is someone the Leafs – and everyone else – really likes? What if it’s someone that’s fallen past their expected draft position. Let’s just say it’s Michael McLeod of the Mississauga Steelheads. Does it make sense then?
I’m a lot more tempted to say yes. But at the same time, you have to be fully aware that what you’re making is an inefficient decision if you make that trade. That’s problematic, because it seems to me like the more efficient decisions you make as a management team the greater position you put yourselves in to succeed. Conversely, the more inefficient decisions you make, the greater position you’re putting yourself in to fail.
Let’s take a look at a couple more examples. In 2011 the Leafs owned the 30th and 39th picks in the draft and traded them to Anaheim for the 22nd overall pick. So, the Leafs traded two picks with a combined value of 474 points in exchange for one draft pick worth 324 points – putting them 150 points in the negative. So that was another inefficient decision on paper, and as most of you well know, it turned up being the wrong one in reality as Toronto drafted Tyler Biggs 22nd and Anaheim took Rickard Rackell and John Gibson 30th and 39th respectively.
Last year though, the Leafs decided to trade down. They owned the 24th pick in the draft courtesy of Nashville, but ultimately decided to trade down to 29th in exchange for that pick plus 61st overall. Then the Leafs traded down from 29 with Columbus in exchange for the 34th and 68th picks. In this scenario the Leafs turned a pick that was originally worth 308 points and turned it into three picks worth a combined 579 points – a massive net gain of 271 points, equal to an additional first round draft pick between 29th and 30th overall.
Now, out of this trade Philadelphia got Travis Konecny and Toronto got Travis Dermott, Jeremy Bracco, and Martins Dzierkals. Was it worth it? I’m actually presently leaning no, however, we can’t say with any sort of conviction whether or not it was the right move given it’s been less than one year since those players were taken. But what’s impressive about that trade is that it was an incredibly efficient decision, and the more of those you make, the better position you’re putting yourselves in going forward.
This gets back to what Burtch said about a late 1st-rounder being closer in value to a 3rd-rounder than a higher 1st-rounder. For example, the 30th pick in the draft is closer in value to the 61st pick in the draft (a difference of 90 points) than it is to the 19th pick in the draft (a difference of 99 points).
This is where things get tied in rather nicely, because we know from the Tulsky article that we looked at that GMs rightly place a high value on high draft picks, yes, but that they also appear to undervalue lower picks. Look at the two graphs again: GMs place almost no value whatsoever on mid-to-late third round picks. But we know from the research done by both Burtch and Schuckers that this isn’t the case: draft picks have at least some value as late as 150th overall, and arguably long past that.
This is exactly why it usually makes more sense to trade down than up: it isn’t so much that GMs overvalue high draft picks, but that they undervalue low ones. Thus, it may not even be that you’re giving yourself more kicks at the can that makes trading down a smart decision, but that you’re in many cases unknowingly increasing the overall value of your original draft pick the more times you trade down, the more picks you acquire.
For what it’s worth, my understanding is that Kyle Dubas is a big fan of trading down in the draft. In the OHL you actually can’t trade your first round pick, which was apparently much to his chagrin. That’s both good news and unsurprising given his interest in analytics.
Of course, sometimes it might make sense to trade up, like in the Leafs-Coyotes scenario I concocted above. After all, what if the right guy is available, and you’re really sure of yourself?
That’s a tough one. After all, it sure seems like the Flyers got a good player in Travis Konecny. Or that the Islanders came away as bandits nabbing Mathew Barzal last year 16th overall (though they only had to give up middling prospect Griffin Reinhart to do so, no draft picks necessary). But then, the Leafs sure seemed sure of themselves when they took Tyler Biggs back in 2011. Or Luke Schenn in 2008.
I guess this comes down to having really good scouts and trusting those scouts. If you’re going to trade up to nab a player, you better be damn sure in what you’re getting. The difficulty there, as history tells us, is that even the surest of things can be anything but. Call it a crapshoot, call it a lottery, call it whatever: while there’s unquestionable value in scouts and their ability to determine which young hockey players are good and which ones aren’t, the fact of the matter is that any draft pick, no matter where they’re taken, isn’t a sure thing. So for me, it usually doesn’t make sense to trade up. And as the numbers would say, not only that, but it’s often smarter to trade down, because right now a market inefficiency exists between the perceived and actual value of later round draft picks as it relates to NHL GMs. For this reason, it might make sense to trade up for the right guy, but realistically you should probably be happier if the Leafs elect to trade down.
*: I previously stated that games played, goals, assists, and +/- were taken into consideration, which has since been pointed out as being inaccurate. My apologies.