Vegas' UFA Selections

June 16th, 2017

The Vegas Golden Knights have between the 18th and 20th of June to create their team out of the pool of players that did not receive protection. During these 48 hours they will have exclusive negotiating rights to the majority of unrestricted free agents. Most free agents are expected to be exposed for a couple of reasons. Teams do not want to waste protections on players that may want to sign somewhere else and who come with an unknown price tag. Washington will not protect Kevin Shattenkirk over John Carlson because they will lose the latter with no guarantee of retaining the former. And even if they were confident that Shattenkirk would resign, why protect him? If Shattenkirk intends on returning to Washington (which does not appear to be the case) he will simply refuse to sign with Vegas during the 48 hour draft period. It’s clear that whether Shattenkirk wants to leave or whether he is willing to stay does not change the fact that Washington should protect NIskanen, Orlov and Carlson on defense. This begs the question for Vegas – with an abundance of unsigned talent available, which free agents should Vegas target?

If you first run the optimization with the ‘Don’t Consider UFAs’ box checked, UFA-free protection lists will be generated for each team. (There are some exceptions. Calgary, for example, will protect either Chad Johnson or Brian Elliott, since their third goalie, Tom McCollum, must be exposed to satisfy the exposure requirements). If you then uncheck the box and run the optimizer again the tool will select a lineup for Vegas out of the pool of unprotected players from the last run, which includes the majority of 2017 free agents. This iteration of Vegas’ lineup (Figure 1) contains 10 UFAs - the maximum number of UFAs Vegas can sign during the draft - indicating that Vegas has better options from the pool of UFAs then from the signed exposed players.

(Figure 1)

It will be too expensive for Vegas to sign all 10 players, and there are plenty of good options for signed players in net and on defense, so let’s consider the situation where Vegas only signs forward UFAs. We can do this by checking the ‘Don’t Consider UFAs’ box again, and clicking ‘Keep’ next to the forwards we wish to keep. Amongst the forwards we are considering to keep, Oshie and Radulov are established players in their prime that will ask for long term contracts. Signing them brings long term risks to Vegas. On the other hand Vrbata, Versteeg and Vanek are all older players coming off bounce back seasons whose success is not guaranteed. If they sign short term deals with Vegas they bring short term risk. Vegas might want to balance these risks by signing a mix of players from these two groups. Let’s go with Oshie, Vrbata and Versteeg for the next iteration of this team. Keeping these 3 players and optimizing results in a team with only $5.5 million in cap space (Figure 2). This won’t be enough. We can rectify this issue by moving Vegas’ slider one unit left so that the optimizer selects a less expensive team. Running the optimization again results in a team with $20.4 million in cap space (Figure 3). This is more than enough to sign our three UFAs.

(Figure 2)
(Figure 3)

In reality, the free agent situation is even more complicated. What’s to stop Vegas from offering a free agent a signing bonus that compels him to sign on July 1st? This would allow Vegas to snag two players from the same team. They can also sign a player and consequently trade him to another team before the official free agency period begins, allowing other teams to get in on Vegas’ exclusive negotiating window. Clever managers take advantage of these types of loopholes, lending to an upcoming week that will be as exciting as it is unpredictable.

The Trevor van Riemsdyk Dilemma

June 12th, 2017

(Getty Images)

Trevor van Riemsdyk is Vegas’ optimal selection from Chicago if you run the tool using default settings. The optimizer picked up on the dilemma that Stan Bowman currently faces. The Blackhawks are completely saddled with contracts containing no-movement clauses. If they opt to protect 7 forwards and 3 defensemen, Seabrook, Keith and Hjalmarsson’s NMCs ensure Vegas has access to the young defenseman. Chicago can alternatively protect 4 forwards and 4 defensemen, allowing them to protect van Riemsdyk, but at the expense of exposing another valuable asset in Ryan Hartman.

The situation is not as bad as it looks. Blackhawks general manager Stan Bowman has been reported by TSN’s Fran Seravelli to be working out a deal with Vegas to make van Riemsdyk available in the draft only if Vegas agrees to also take on Marcus Kruger and his hefty contract. If the deal goes through, Bowman will have turned an inevitable loss into a possible gain. The freed up cap space can be put to good use in resigning players like Hartman and Artemi Panarin, whose contracts will expire over the next two years. Bowman’s leverage in this deal is his ability to trade van Riemsdyk to another team that can actually afford to protect him ahead of the expansion draft. Last year he unloaded Brian Bickell’s contract to Carolina at the expense of Teuvo Teravainen, setting the precedent for this type of exchange.

Theoretically, whether McPhee selects Kruger or van Riemsdyk in the draft is irrelevant to Chicago, as long as a trade is set up and Vegas acquires the outstanding player. However Vegas might have more to gain by drafting Kruger. Using the tool, we can test both scenarios. If Vegas drafts van Riemsdyk (default settings) their total team point shares value is about 114.1. Trading for Kruger, who has a point share value of 0.9, would raise their total point shares to 115. If you are unfamiliar with the tool you might be confused as to how to test the other scenario. You can reoptimize so that Vegas selects Kruger by clicking the 'Remove' checkbox next to van Riemsdyk's name in the Vegas lineup and reoptimizing. Brian Campbell will be the next choice. Remove him as well and reoptimize to get the optimal iteration of Vegas' drafted team that includes Kruger. This team has a point shares value of 113.1. If you then assume that Vegas would trade for van Riemsdyk and his 4.3 point shares, that bumps the team up to 117.4 point shares. This makes sense when you consider the fact that by drafting Kruger, Vegas can load up on one extra defenseman in the draft.

Good young defensemen are actually hard to come by in the NHL. Last year the Edmonton Oilers traded all-star forward Taylor Hall for Adam Larsson. Larsson’s point shares value last year was only 1.2 higher than van Riemsdyk’s despite the fact that the latter had missed 24 games. He would be a valuable asset to the lineup, but as the tool demonstrates, he is not the only good player coming Vegas’ way on June 21st. The Rangers, Islanders, Ducks and Penguins are a few other teams that stand to lose significant assets to Vegas and it will be interesting to watch how these clubs react in the weeks ahead.

Welcome to the NHL Expansion Draft Optimizer

June 11th, 2017


The upcoming expansion draft is the first one in 17 years. An entire team will be constructed out of the offcuts of the existing thirty. This is shaping up to be the most exciting offseason in recent memory and we all can’t wait to see how everything unfolds. But until it does, all we can do is speculate. Hockey is a chaotic sport, which is partly why it is so fun to watch, but there can be just as much chaos in management offices as there is down on the ice. We can’t predict what will happen during the draft because we cannot account for all of the factors that are at play. No one knows today who is going to waive his No Movement Clause, or who is getting traded, or whose contract is getting extended to help his team meet exposure requirements.

No one knows exactly how George McPhee values players either, but it is probably a mix of age, cap hit and on-ice performance. We know all of these player’s ages and contracts, and thanks to the work of Tom Awad, Justin Kubatko, and the EA Sports’ analytics department, we also have metrics that quantify their on-ice performance. All of these statistics are freely available through online resources such as Cap Friendly and Hockey Abstract. So while we don't have the information to say exactly how the existing team rosters will change between now and June 21st, we can tell you the best possible group of players Vegas can select if the draft were to take place today. That is the power of optimization, and this app puts that power at your fingertips to help you speculate like an expert.

The default optimizer is based on the assumption that all protection lists will be created so that the value of the best player available from each team is minimized. This assumption works reasonably well, but in select cases, such as Kevin Shattenkirk being protected over John Carlson, it is probably the wrong assumption. You can correct for this ‘wrong’ assumption by going to Washington’s protection list and specifically protecting John Carlson and exposing Kevin Shattenkirk. Another assumption is that Vegas will draft the group of players that give them the best chance at winning the Stanley Cup next year. You can adjust this assumption by moving the slider further towards the left, increasing their financial flexibility at the expense of their on-ice performance. Finally, if you dislike certain Vegas selections, you can remove those players from all future iterations of Vegas’ lineup.

What you’re left with is the ability to see what Vegas’ best possible team would look like if the draft were to take place at that moment under a set of specified assumptions. You can test different scenarios by adjusting these assumptions. The post below, for example, compares two possible scenarios involving Vegas’ deal with Chicago to acquire Trevor van Reimsdyk, concluding that if the Golden Knights trade for him instead of drafting him they can accumulate more value. You’ll soon find that speculating Vegas’ lineup is a lot easier when you can create different iterations in a matter of minutes!