I often make reference to the value of a win – here's where those numbers come from:
| Salary | Salary | Win | $/Win | |
| Season | Total | Above Repl | Above Repl | Above Repl |
| 2011-12 | 1782.86 | 47.35 | 19.5 | 2.43 |
| 2010-11 | 1699.55 | 44.58 | 19.5 | 2.29 |
| 2009-10 | 1647.79 | 42.85 | 19.5 | 2.20 |
| 2008-09 | 1615.07 | 41.76 | 19.5 | 2.14 |
| 2007-08 | 1404.75 | 34.75 | 19.5 | 1.78 |
The way this works – the entire NHL spends a certain amount on players (which has risen from $1.4B to almost $1.8B in four years) while the total number of available wins is fixed. But the price of a win is not just total salary divided by total wins. Here, we need to introduce the concept of a "replacement level" team, which is roughly what you'd get if you signed the 23-best free agents you could find for the league minimum (the cap prohibits such a team – that's why this is a thought experiment!) The salary above replacement is the amount spent above the $12.65M required to put that replacement team together – $47.35M per team this season.
The next question is how well such a team would do. The answer is a bit approximate – roughly as bad as any post-lockout team could be. I like using Tom Awad's GVT system, and he puts replacement around 52 points, so we'll go with that. The average team in the league puts up 91 points, which is 19.5 wins above replacement.
So now we can divide salary above replacement ($47.35M) by wins above replacement (19.5) and we get a price per win ($2.43M.) A one-win player would earn roughly $2.98M (minimum salary + $2.43M.) Keep in mind that this is the price per win for all players, including those who are young enough to not have reached unrestricted free agent status. Those players – on entry-level or RFA contracts – play at a significant discount, which drags down the average price per win. A team signing a UFA will have to pay more per win than the $2.43M we've found here.
