GVT – 3 x (CAP – 0.5)
While discussing Dustin Byfuglien’s contract recently, a question from fat_daddyo came up about how to "measure GVT (and by extension marginal wins) against salary/cap hit and see if a particular player is playing up to their contract."
My initial response was to use GVS: Goals-Versus-Salary, for which the formula is:
Why 0.5? Because that’s what a replacement-level player gets
Why 3? Because average GVT for a team is 120, and there’s roughly $40M to play with
We first came up with it in August 2009 over at Hockey Prospectus to provide us, at a squint, a rough idea of who was providing good value for the dollar, and by extension which teams were managing their cap well.
Due to its association with GVT, it can also be used to measure roughly how much certain contracts are costing their teams in goals, and by extension in wins. Since then we’ve used it to evaluate trades and free agent signings.
Also, for those who are interested, about a year after the original article we published some conclusions on how to build a team in the salary cap era.
In anticipation of further questions, let me quickly hit on five of the most common concerns.
1. GVS Penalizes Elite Players
If you were to assemble a team composed of players with the best GVS, you would probably not have a very competitive team, despite your efficiency. You’d probably have a collection of lower-paid bargain players, and lots of cap space to spare. You’re better off filling that cap space with superior players that are higher paid, but are also less efficient uses of dollars.
In fact, if you were to have a team without any replacement-level players (that’s highly unlikely), you’re obviously better off with a single star making $6.0 million than two 2nd-tier players making $3.0 million apiece, since he would take up less roster space.
To put it simply, in order to be competitive you can certainly afford to be as efficient as possible with your supporting cast, but pinching pennies with your star players isn’t practical. Therefore, when looking at GVS, try to avoid comparing apples to oranges.
2. Cap Hit vs Salary
Some people don’t consider cap hit to be the appropriate choice for this calculation, for instance Nashville Predators fan Sam Page
"For a team that regularly spends up to the salary cap, the cap hit determines a player's value to the team. A player's worth is determined by what else his cap hit could be used towards. For a team constrained by an internal budget, though, the concern is what else the actual money paid could be put towards."
It’s easy enough to change the formula to look at salary instead of cap hit when looking internally at a team that is constrained by an internal budget, like the Nashville Predators. In almost every case you’re probably just splitting hairs, but it’s still the correct thing to do when comparing Nashville Player A to Nashville Player B.
On the other hand, the moment you’re evaluating a player’s value within the context of the league as a whole, or between several different teams, it’s probably best to use the more common limiting resource, which will generally be cap hit, not salary.
3. What about the Cap Floor?
We often get questions about why we haven’t factored in the cap floor. The answer is simple: it really doesn’t matter. Wade Redden and Sheldon Souray don’t become more valuable to a team near the cap floor just because they’re overpaid.
If you’re spending below the cap floor and you need to acquire a player to get over, you still want to get the best bang for the buck. The calculation is identical.
4. Different Types of Contracts: ELC, RFA, UFA
The average GVS is going to be different depending on whether the player is a free agent (restricted or unrestricted), or on an entry-level contract. You can expect certain discounts, and we’re soon to publish a study quantifying the average extent of those discounts. In the mean time take our earlier advice and avoid comparing apples to oranges.
5. Are there other ways? Yes: PCValue
To double-check a GVT-based calculation, you can use Alan Ryder’s Player Contributions, which is a bit more precise especially with defensive contributions. The formula is:
PCValue = PC x $56,800 – CAPHIT
The only problem with PCValue is that it’s only available in Alan’s off-season report, but it’s still useful for finding the types of players that are unfairly penalized by GVS.
* * * * *
Given the infancy of the salary cap era, and the subsequent lack of maturity in our analysis of it, it’s a great time to ask questions and present alternatives, so please be generous with your comments. And thanks for reading.