Death by a Thousand Contracts

The Vancouver Canucks confirmed a John Buccigross report from earlier this week when they announced they’d come to terms with University of Western Michigan’s Griffen Molino to a two-year entry-level contract.

This marks the third time the Canucks have ventured into the unrestricted free agent pool to try and unearth young talent that’s slipped through the cracks of the NHL entry draft in this season and the seventh occasion since Canucks general manager Jim Benning took the job three years ago.

At a glance, one can certainly appreciate the Canucks not leaving any stones unturned in the pursuit of fresh young talent to invigorate the ranks. The beauty of adding an undrafted free agent is it doesn’t cost anything except cash and a standard player contract against a team’s upper limit of 50.

If you have the logistical flexibility, it can make sense to dip into that market as often as possible. That first point matters, though, because each free agent is another contract towards the maximum and they eventually add up. It’s a double-edged sword and the team that wields it has to be judicious in how often they swing it.

The most recent Canucks signing doesn’t seem to reflect that level of discipline. As Canucks Army’s Ryan Biech pointed out in his article on unrestricted college free agents, Molino’s meek statistical profile suggests he’s highly unlikely to develop into a professional NHL’er. In fact, at the time Biech profiled those collegiate forwards, Molino’s expected success rate by pGPS (Prospect Graduation Probabilities System) was the lowest among the entire crop of NCAA centres he bothered examining.

That wouldn’t be such a problem if we could write this off as one of a kind, low probability bet that can be explained by Benning’s scouting success as a counterweight to the quantitative landmine it seems they’ve jumped on with glee. I’m willing to turn a blind eye to this kind of contract if it’s the exception and not the rule. If that was only the case.

Now the pGPS system that we so frequently lean on in this space isn’t a perfect tool. While the research Jeremy Davis has done shows it to be a predictive metric for draft success, the very fact that it uses historical principle as a threshold means it’s imperfect in nature. A player can, hypothetically, have a zero percent expected success rate and make the NHL — it’s just supremely unlikely.

With that caveat, we can look into each of the Canucks’ undrafted free agent signings and the principle for success with each. And the picture is bleak. If you combine the success percentage of each undrafted prospect the Canucks have signed, the expected success rate is 37.25 — less than half of an NHL player. That’s seven players worth and not many reasons to expect more than the one NHL’er they’ve already squeezed from the lot.

It would be disingenuous if I didn’t add that one part of that sum, Troy Stecher, is in the midst of his first season in what looks like it will be a distinguished NHL career for many to come. I think everyone involved would suggest that Stecher’s expected success rate belied his potential, though. That’s bound to happen when one uses a metric based on historical principle, and to that exact end, we should allow a certain amount of wiggle room to the decision makers involved to put their acumen against this methodology or any of its kind from time to time.

How much wiggle room is a worthwhile question, though. Assuming the Canucks don’t add another undrafted free agent, Molino makes seven contracts added through this route. That’s 14% of the Canucks 50 contract limit, and six of the players that account for that number are even money to combine for 100 NHL games played.

Frankly, I’d set that number at one NHL game played if not for the Canucks burning the first of Molino’s two-year deal by playing him as early as this Friday — a savvy move, which effectively turns his contract into a one-year deal with a couple of weeks on top.

This might not seem like it’s worth worrying about, but there can be very real ramifications for throwing contracts around with reckless abandon. Each sound, high percentage play on the waiver wire requires the requisite room against the contract limit to work. Imagine a world where the Canucks can’t put a claim on the next Mike Hoffman, Reid Boucher, Magnus Paajarvi, Martin St. Louis or David Schlemko because they just had to get their hands on Molino or a player of that ilk? What of the number of other, readily available undrafted prospects with a significantly better precedent for success?

Some of these gains I’m pleading the Canucks grant themselves the flexibility to make may seem insignificant, but running an NHL hockey team is an efficiency contest where everyone works from a capped resource base. It’s not “just one contract, ah whatever” when it happens seven times over and the quantitative support for each decision is less and less solvent.

The Canucks have 49 contracts against their 50 limit as they enter the final week of their season. They’re already committed to 31 contracts for next season per www.CapFriendly.com, but that number swells to 40 when one accounts for the restricted free agents Vancouver is likely to sign, and that’s a fairly conservative estimate. Then there’s the matter of their own draft picks and adding players to fill the significant gaps in their NHL roster.

That doesn’t leave the Canucks a large margin for error. Then again, they seem to like working under those auspices. They’d be far more judicious in their approach at free agency otherwise.

  • GoodUserName

    Benning is winning me over. I like his moves lately, Im good with giving him the benefit of the doubt going forward when it comes to his scouting eye. I don’t think being close to the contract limit will negatively impact us that much. Deals get made all the time where an ahl’er is thrown in to make things work. You cant win the lottery if you dont buy a ticket. Jimbo is just buying a lot of tickets, and I think realistically if JB could have signed a better college prospect he wouldve.

  • TD

    I agree with the premise of this article that the Canucks need to make sure they have some contract flexibility to sign new players if the chance arises, but that is what they just did. If one more of the players turn into an NHL player then they will have hit two out of 7 which is the found money the CA writers always talk about. I don’t trust the pGPS stats. The NHL has changed dramatically recently with lots of smaller and faster players making it. That has to have skewed the numbers of the predictive models.

    I think the predictive models are also limited to pure stats and don’t assess any other attributes a scout might see by watching the player. An earlier CA article commented on what drafted prospects the Canucks are likely to sign this summer and I was surprised Carl Neill was not mentioned. I later found out he is considered slow and not likely to have an NHL career, but his size and scoring stats told a completely different story for the predictive models.

  • Killer Marmot

    Your analysis would be correct IF the only information at the manager’s disposal were the ones that went into the pGPS analysis. But if the manager brought other information into the decision then he could, in theory at least, defy the pGPS odds.

    For example, suppose a prospect had statistics that suggested that his chance of having a successful career were low. But then your scouts notice some things. They see that the prospect does a lot of the penalty killing, that his line mates are of poor caliber, and that he often takes the faceoff in his team’s end due to his excellent faceoff skills and defense. These extra facts suggest that player is better than is reflected in his offensive statistics, so that the pGPS underestimates his chances of success.

  • Bud Poile

    During the ‘Six lost Gillis years’ I can think of only Dane Fox as the one undrafted free agent signing.
    Knocking Benning for consistently evaluating,singing and giving promising young talent the opportunity to make both the player’s career and franchise prosper is negative,destructive and unnecessary.
    I would have explored writing this topic in a positive manner instead of what is being offered by the author’s negative stance.

  • wojohowitz

    This issue of the number of contracts is a non-starter. Benning is not going to wake up one morning and say; What have I done? If he does enter an area of questionable asset management you can be sure that one of his Directors of Player Development (Johnson and Smyl) will be letting him know. With the signings of newer free agents we can assume that a number of AHL players will be cut loose and who that will be has already been decided but won`t be announced until the AHL season is over. It is like you are suggesting Benning is shopping on an over limit credit card and doesn`t know it.

  • acg5151

    I don’t think it’s as big of a deal as it is being made out to be. If even one of these players pans out I think these gambles were worth it. We are going to see the expansion draft take some of the contract space off our hands, we are going to see several players let go this following season, and generally it’s not as difficult to dump a contract as it might seem, especially with teams often trying to bolster their minor league depth to get their prospects deeper into the playoffs – something seen as beneficial to the development of rookies. I find it hard to believe that we aren’t going to dump one or more of Alex Grenier, Skille, Cramarossa, Megna, Shore, Chaput, Philip Larsen, Chad Billins, Tom Nilsson, or Borna Rendulic. I think we’ll see a few of those guys leave, especially with Desjardins probably on the way out.
    Bud Poile, you are wrong as Dane Fox being the only undrafted free agent. Chris Tanev was taken using this approach, I think it was one of Mike Gillis’ best moves.

  • Carl Jung

    This is the usual blogger supremacy article.

    The reader must suspend disbelief and accept that the publicly available info CA offers for free is better than the info Canucks management has…which includes publicly available blogger nonsense such as this article.

    In any case, signing seven undrafted free agents and ending up with one top 4 defenseman seems fine.

    Are other teams doing better or worse in free agency?

    Did Evan Oberg or Lee Sweatt receive this much scrutiny?

    Or did CA bloggers accept that one is going to whiff A LOT in order to unearth a Burrows, Tanev or Stecher.

    All 31 teams sign players that are essentially AHL fodder such as Rendulic & Billins.

    This is similar to all of the middling assets the Canucks gave up for Baertschi, Granlund, Pedan, Etem, Vey etc.

    I haven’t seen a CA blogger demonstrate that the Canucks came out on the shorthand on this strategy COLLECTIVELY.

  • HockeyTruther69

    but does pgps take intangibles into the picture??? noooo. someone shoudl make a model that reflects intangibles…hmm this gives me an idea. look out canucks army!!!!

  • Canuck Beagle

    Not much to fret about here. There’s 31 contracts to start next season plus 15 RFAs, a number of which (Zalewski, Nilsson, Shore, etc.) will not be re-signed-let’s say 10 are re-signed. A few juniors will come on the books (Juolevi, Brisebois, Neil, Tate, etc.). Bennng will add a couple of low level UFAs. This leaves about 3-4 slots open next season. After the 2017-18 season, a bunch of these NCAA contracts will expire and go away if they don’t work out.

  • Benning signed Molino because he saw something he liked. SBNcollegehockey.com thinks he has 3rd/4th line potential which is like a having late round draft pick. What’s wrong with that?

    “Molino is a crafty forward with excellent agility. He’s a decent north-south skater, but is really good at attacking east-west, making him a tough match-up for opposing defenders. Molino also has excellent hockey sense and competes hard, playing much bigger than his listed size. His upside at the pro level is likely as a third or fourth line energy line player.”

  • Roy

    what’s wrong with stacking your farm team with young maybes and not resigning the lifers and/or trading them? Who cares if they never make the NHL. The waiver option is good too, this article is not terrible nor is it wrong to be careful, but if Benning wants to test the NCAA against waivers, go nuts.

  • Van94

    Tell us true though, if you had your way we wouldn’t have Stecher would we? As far as the rest of them you make it sound like Benning had this huge pool of talent to draw from and didn’t need to try for a few long shots. Which he’s done well at. It must be time for another Corrado rant.

  • Killer Marmot

    Analytics can be useful, but there is a danger that people respect its scientific-like authority too much without appreciating the assumptions and limitations behind it.

    • Killer Marmot

      Further to my point, what does the “2.7% expected chance of success” really mean? It does not mean…

      The prospect’s expected chance of making it in the NHL is 2.7%.

      Instead it means

      If you limit yourself solely to the information that is used by the pGPS, the prospect’s expected chance of making it in the NHL is 2.7%.

      If you bring in more information then the player’s odds change, sometimes for the better, sometimes for the worse.

      This is what I mean by appreciating the assumptions and limitations of analytics.

      • DJ_44

        Actually, as I understand the pPGS model, the percentage means that 2.7% of player that exhibited similar characteristics (which I believe are goals, assists, age, size and league played) have gone on to play 200 games in the NHL.

        I think JD is being a bit disingenuous…it is not the fact that GMJB is signing free agents that is objectionable, it is the fact he does not like free agents that were signed (they appear to be over 5’9″ and weigh more than 165lbs). ……. Ty Rattie where are you?

        • Killer Marmot

          the percentage means that 2.7% of player that exhibited similar characteristics (which I believe are goals, assists, age, size and league played) have gone on to play 200 games in the NHL.Yes, that’s probably a better interpretation of what it means.

          That seems like a more accurate interpretation than my version. My point being that going from that to

          Any prospect who Benning signs that has those characteristics therefore has a 2.7% chance of being successful in the NHL.

          is an unjustified leap in logic. The second statement does not follow from the first.

        • 5forBoarding

          Further to this, there are some glaring flaws in the analysis. First off why does every contract have to be someone who can turn out to be an NHL’er who will play 200 games? Perhaps Molino is being brought in to boost Utica, while being able to provide a couple of games of replacement level NHL play in case Vancouver is decimated by injuries and need a couple of games from deep down on the depth chart. The theory that every contract you sign must be in search of an NHL player with as high a chance according to pPGS of playing in the NHL (at least 200 games) is ridiculous. Sometimes a player is signed just because he can add depth to your farm, might work out as an NHL’er, but at the very least will be able to provide spot duty if you’re on your 50th forward in a given year due to injuries.

          The other problem I have with evaluating Molino is with the pPGS itself. It’s based on historical data, but I don’t feel has ever done a good job of factoring in major shifts with regards to they type/style of what a current NHL player is. I have read on many occasions that the NHL is becoming smaller and faster, favouring skill to toughness. This would, in theory, lead me to believe that for an undersized forward such as Molino that his pPGS is much higher than what is being shown, and that the pPGS of larger/taller players is often much lower than what is being stated.

        • sloth

          Yeah DJ_44 is correct about the way the number is calculated – it’s not really about “expected chance” or probabilities or likelihoods, it’s about the percentage of previous players with similar statistical profiles who have gone on to play in the NHL. Killer Marmot is correct that the number that pGPS gives is based only on the information available to the model.

          The thing that irks me about the way this blog uses pGPS is that it seems to me to be most useful as a drafting strategy tool when you’re sifting through a couple hundred prospects who you can’t scout in depth individually. Hockey is such a complex sport and there are so many variables that go into a player’s utility at the NHL level that are not captured perfectly by boxcar stats and body size, though they are undoubtedly valuable metrics to look at as proxies for skill, athleticism, attitude, and IQ. But the data is less valuable when applied to players following a less-common development path, as many of their peers are likely players who have already passed that window of NHL potential, so while Molino has 498 “statistical matches,” that doesn’t really tell us anything about him as a player, because I’d wager the majority of those matches were never really considered NHL prospects for reasons other than just their boxcar stats and body size.

          The problem is when we get to actual professional scouting and prospect evaluation, the importance of boxcars is diminished, and with that, the importance of other players’ past stats are reduced to virtually zero. We’re talking about a predictive model that uses imperfect proxy data from a questionably biased sample, and we’re treating it like some sort of proven scientific method of player evaluation. At this point Molino is now a Canucks player, and they apparently beat out other teams to sign him. I don’t know anything about this prospect and I don’t necessarily have high hopes for him, but I don’t just want to hear that it was a bad gamble because 498 former college players didn’t make it to 200 games in the NHL, because obviously there are a couple NHL scouting departments that like the way Molino plays and are comfortable with his boxcar production (or lack thereof). Why? What role did he play on his team this year, and who did he play with most? How good was his team? What was his ice time and zone deployment like? Does he kill penalties? Does he read the game well?

          I’m really not concerned about the contract limit, and it’s silly to make such a huge fuss about it in this post-deadline lame duck stretch. In fact, the flip argument is that we’re reaching a point in the year when the value of empty roster spots has decreased, with no expected waiver pickups and a ton of contracts expiring in the summer, this is the time to fill up to 49 (leave 1 empty for flexibility), and ride it out to the end of the season. Considering Zach Aston-Reese and John Stevens both signed (PIT & NYI), the Canucks options for making an addition at centre were Molino, Alex Iafallo, and Mike Vecchione. I might have preferred they take a flier on one of the wingers available (Spencer Foo would be my 1st choice, but it sounds like he’ll wait until the summer to sign), but I can’t really criticize this decision given the circumstances.

          • Dirty30

            Well said! We do currently have a small surplus of wingers and one top ten draft pick waiting in Utica. A good centre to replace Hank and even Sutter at some point is a priority. And centres can effectively play wing but seldom does vice-versa work as well.

  • Dirty30

    Isn’t this less ‘proof of poor asset management’ and more ‘proof that management just got very serious about rebuilding’? Given the number of contracts about to expire and the low likelihood of a good number of those not getting renewed, this seems to be exactly part of a strategy to obtain new players at little or no cost. No draft spot used, no assets traded, no big money or term contract and this player fits the age range to build a team for 3-5 years in the future.

    I only hope that JB does have the fortitude to indeed allow many contracts to expire and offer just a handshake and a thanks for coming out and trying.

  • Betty

    I love stats but this is just silliness. How many contract spots would you gamble for another Stecher, Tanev or Burrows? To rephrase JD, imagine a world where the Canucks don’t pull Tanev, Burr or Stecher out of the undrafted pool, instead use the PGPs system.

    The fact Stecher had a 6.2% chance of making the NHL, according to this tool kind of speaks to its imperfections.

    Not to mention, using their own tool, it looks like JB is seeing something the other GMs are missing as half of the non Stecher acquisitions have improved their odds since signing with the Canucks.

  • tyhee

    The main point of this article is perhaps getting hidden behind the imperfections of a tool CA uses with prospects.

    The main point of the article is that the Canucks have signed several low-ranked prospects recently. I don’t think that can be denied.

    The tool the article used to show the signings are low percentage players is imperfect-in fact, it is very rough. As I understand it all that is done is compare with a database of similar players with similar age, height and scoring in the same league and say how many of them did or did not, for whatever reason, become NHL regulars. It isn’t based on scouting and is only intended to be a rough glance.

    When Stecher was signed last season everyone was happy with the signing. Nobody was all that concerned about his pgps, which was low because one of the few variables it uses is height. He was a highly regarded, very successful college player with obvious assets.

    None of this season’s free agent signings to date have been as highly considered. There are other players that other clubs were interested in to one extent or another, but by no rankings are these players as highly rated or coveted as Stecher was a year ago. These are generally marginal prospects.

    To return to the point of the article, pretty clearly CA has taken the position that some waived players and some minor league free agents that will be available have better chances than at least some of thse signings and that with extending some players they really need to extend (Horvat, Tryamkin etc), fleshing out the NHL roster and getting some good veterans to give the Comets a competitive squad and serve as decent injury callups, the 50 contract limit places limited space for low percentage prospects. It is worse with overage juniors, only a year out of the draft and playing against younger players, than with college players who are often 2-3 years past their draft, with substantial improvement and having succeeded against players of similar age.

    On the good side of the most recent signing is that Molino isn’t a 19 or 20 year old kid. He’s a little older, been playing against other players in their 20’s, is reportedly a good skater and he was last eligible for the draft in 2014. He’s apparently improved considerably since then. There’s a reasonably good chance he might be helpful at the AHL level even if he doesn’t make the NHL. It is probably a better signing than, for example, Sautner was in 2015, one year removed from the draft and playing against younger players in an overage D+2 season.

    To switch back to the point quite a few are making, of course, any time CA uses PGPS it would be helpful to have some other sources indicating how scouts view the player.

  • Neil B

    The main point of this article is the Canucks Army editorial board needs to remind us that they don’t like the Canucks’ front office.

    Case in point: the 40 contract slots once our own free agents are re-signed.

    Firstly, 39. We will lose someone to Vegas. It is mandated by the NHL that we will lose a contract to them.

    Secondly, we have only 6 guaranteed picks this year (something the CA editorial board has complained about all season). So we could sign every pick to a contract, and still have enough open slots to pretty much hire a complete top unit. Thirdly, one interesting side note about Benning is the number of his picks that go the college route. My guess would be he adds another college-bound player this year as well; which gives us at least 2-3 years to forestall signing anyone. Fourthly, if the 7th-rounder is worth signing in his draft year, then start the fricken’ parade, dude.

    So, really, we’ll have room for 8 new FA signings, after all is said & done. Do we really think he’ll bring in that many new people?