Update, 1:45 pm August 5th - Nashville announced that they had settled before Dante Fabbro’s arbitration hearing. The Predators awarded the defender a two-year, $2.4 million AAV deal.
The past few weeks of Predators news have been consumed by discussion of salary arbitration, prompted by Dante Fabbro and Juuse Saros electing for arbitration hearings as restricted free agents. While some argued that Fabbro’s choice was an inevitable business decision, others saw the situation differently. There were even some advanced visualizations offered up on the matter:
Both Fabbro and Saros have hearings scheduled: Saros on August 18th, Fabbro on the 20th. To get some perspective, let’s take a look into recent history when it comes to arbitration.
Author’s note: Goaltenders have been included in counts below for arbitration, but models for predicting outcomes are restricted to skaters only.
The NHL Collective Bargaining Agreement (CBA) outlines the rules regarding restricted free agents, but to spare you some reading time and summarize, RFAs can be offered a Qualifying Offer (QO) that is based on number of games played in the NHL and the player’s previous salary. So, while the amount offered to Dante Fabbro - $874,125 - seems low, it’s the league-mandated amount.
RFAs can then either accept the qualifying offer and sign the contract, or file for arbitration (the team can also elect for arbitration). Arbitration offers a third-party review of player performance and awards payment based on that evaluation. There’s not much out there in terms of guidelines that are used in the arbitration hearings, but we can guess (more on that below).
Another possibility for players electing to go to arbitration is that they can settle with the team before the hearing. In fact, settlement has been by far the most common option in the last five years.
One difference this season is that the new CBA updates (item 59, CBA memorandum of understanding) say that settlement can no longer occur after the hearing (Thanks to Adam Vingan for the heads up).
Players and teams have historically preferred to agree upon a contract on their own, rather than elect for the awarded amount from the arbitration hearing. The Nashville Predators are no exception:
There are other rules pertaining to arbitration, and the folks at Cap Friendly summarize them below:
Arbitration information courtesy of Cap Friendly:
- Player & Club can settle on a deal at any point prior to the official ruling
- Once the hearing has taken place, the Salary Arbitration decision must be issued by email within 48 hrs of the closing [CBA 12.9(n)(i)]
- Arbitration awards can only be 1 or 2 years in length [CBA 12.10(a)&(b)]
- The party (Player or Club) who did not elect for Arbitration decides on the awarded term [CBA 12.10(a)&(b)]
- Players who are in their final year Restricted Free Agency are only entitled to a 1 year term.
- Club cannot walk away from a Club elected Arbitration Settlement [CBA 12.10(e)]
- Player elected Arbitration Settlements of 1 year and greater than $4,538,958, Club can walk away from the awarded salary, making the player a UFA [CBA 12.10(a)]
- Player elected Arbitration Settlements of 2 years and greater than $4,538,958, Club can walk away from the second (2nd) year of the awarded salary, making the player a UFA at the end of year 1 [CBA 12.10(b)]
With all of that out of the way, the next topic is projecting the contract awarded to a player if they do not settle outside of arbitration. Usually projection models are complex and require use of several different input variables, but this is the NHL we’re talking about. I suspected the arbitration committee did not have some advanced metric or combinations of metrics to perform their evaluations, but I did not expect the existing correlations to be so obvious.
In the last five years, the statistic that is most strongly correlated to the awarded arbitration amount is simply points. For all skaters the correlation is strong (r2 = 0.85, with 1.0 being perfectly correlated), but for defensemen, the relationship is even stronger (r2 = 0.98). I’d like to reiterate that correlation does not equal causation, and results could be due to overfitting. Still, based upon the samples we have, I would say that using a skater’s career point total prior to arbitration remains the best predictor of what the committee will award them.
In the case of Dante Fabbro, I wouldn’t expect the awarded amount (if the hearing takes place) to deviate too far from his AAV last season ($1.4 million). Despite that, the CBA says that if the amount awarded from arbitration is accepted, the contract can only be either one or two years long. If I’m General Manager David Poile, I would use the awarded amount as a guideline (likely between $1.2-1.6 million AAV), and then negotiate a longer deal in order to lock Fabbro in at a reasonable value.
So, don’t fret that Fabbro and Saros went to arbitration - goaltending contracts are weird and hard to predict regardless of the circumstances surrounding them. Arbitration is just another business decision, and one that can benefit both player and team. The process isn’t necessarily to make sure a player “gets a deal he deserves”, but instead to evaluate the player by a third party. One would hope that the committee serves as a fair evaluating source, but when the results are so strongly correlated with point totals and little else it’s difficult to make such an assertion.