Each year, the NHL calculates the salary mid-point. The cap is 15% higher, the floor is 15% lower. The difference can't be less than $16m or greater than $28m. Next year's cap is expected to be about $71m. That puts the mid-point at about $61.7m and the floor at about $52.4m. Top to bottom that's a little over $18m.
Since the 2004 lock-out, the top team has averaged about 115 points and +74 goal differential. The last place team has averaged about 62 points and -76. That's a difference of 53 points and 150 goals.
So let's line up Goals, Points, and Dollars. One rule of thumb is that 6 goals equals a win, or that 3 goals equals a point. Let's round those average numbers off a bit to make it line up better. So top to bottom is 50 points, 150 goals, and $18m. Top to middle is 25 points, 75 goals, and $9m.
GVT is “Goals Versus Threshold”. The other metrics I reviewed are based on average. I'm not sure where Threshold falls on this spectrum. A completely Threshold team probably finishes last in the league by a wider than normal margin. Plus, it is impossible to have a completely Threshold team. You have to spend to the salary floor, which means you have to wind up with some average and above-average players. I know where Average falls and that seems to be the place to start.
If you had an average team with a mid-point payroll, you would be paying your players $63m/23 = $2.74m each. You have about $9m to acquire 75 goals. So if you spend efficiently, you need to spend about $120k per goal. You don't have to be completely efficient. Some of your players are on entry-level contracts. For example, Tarasenko makes $900k and Schwartz makes $830k but both produce above average. That frees up some money. Low cost UFA players probably don't help free up money. Although they cost less than average they generally produce less than average. Although you have more money to spend you have more goals to acquire. Maybe, if you have enough entry-level players, you can free up another $9m. So $240k per goal is about the upper limit for efficient spending.
Goalies are probably the easiest players to calculate an expected value for GVA. For each goalie, I calculated how many even strength shots they are on track to see this season. I took their career even strength save percentage times that number of shots to calculate goals allowed. I also calculated how many goals an average 0.920 goalie would yield. Estimated GVA is the difference.
Halak is 0.925 ES for his career. He was on track to face about 950 ES shots this season. That puts him at about +4.75 EGVA. Reasonable salary for him would be $3.3m to $3.8m. Right on the money, as it were. Elliott is a little harder to estimate. Career he's 0.917 but with the Blues he's been 0.933. Let's say he's really also 0.925. He was on track for 577 ES shots. He calculates to somewhere around +2.88 EGVA. Reasonable salary for him would be $3.1m to $3.4m. He makes $1.8m. That's a pretty good bargain.
Lundqvist is going to make $8.5m next season. Previously I calculated him to be about +14 EGVA. His efficient spending salary range calculates to $4.4m to $6.1m. $8.5m is a massive overpay. The cap would have to be around $120m for this deal to make sense.
Pretty much all of the other recent goalie contracts are overpays. Jon Quick 0.922 +2.5 EGVA $5.8m Overpay. Corey Crawford 0.923 +3.6 EGVA $6m Overpay. Mike Smith 0.921 +1.7 EGVA $5.7m Overpay. Jimmy Howard 0.925 +5.7 EGVA $5.3m Overpay. Steve Mason 0.914 -8.5 EGVA $4.1m Hysterical overpay. Proper salary looks to be around $1 to $1.7m.
Interestingly, the new contract for Tuukka Rask looks about right. 8 years, $7m a year. Career he's 0.933. He's on track to see about 1500 ES shots this season. That puts him at +19.5 EGVA. Reasonable salary for him would be $5.0m to $7.4m. The salary cap doesn't have to go up a whole lot for this one to even out. Of course Rask has to stay at 0.933, but that's the risk you take.