So as part of a larger project, I took the salaries of each Jazz player and divided their salary by WS (win share - the estimated number of wins that player contributed to the season). I also listed Offensive WS, Defensive WS, Offensive Rating, and Defensive Rating for each player (all from Basketball-Reference.com).
OffWS = offensive win share = originally developed by Bill James for baseball, but converted for basketball to measure the estimated offensive contribution a player makes toward a win. Key factors that go into to this calculation are total points scored by the player, offensive possessions for that player, and team pace.
DefWS = defensive win share = the estimated defensive contribution a player makes toward a win. Key factors that go into this calculation are minutes played, defensive rating, team defensive possessions, and team pace.
WS = win share = OffWS + DefWS
And as you may already know, ORtg is the estimated amount of points scored per 100 possessions, while DRtg is the estimated amount of points allowed per 100 possessions.
Team Contribution to Defensive & Offensive Ratings
Basketball is a team sport. For this main reason, I believe there is an overall effect that a team has on each individual player's ORtg and DRtg, which only makes sense, because a team has to work together otherwise they all lose together. For example, if one player is always playing bad defense and happens to be on the court the same time as another teammate most of the time, than that other teammates DRtg will also go down since DRtg is a measure of the points allowed while he was on the court. I like the On/Off splits as a better tell of how each individual player is doing, but I think it's always important to a.) know how these statistical numbers are derived, and b.) acknowledge that they might not reflect 100% of what they are suppose to be representing.
I was surprised to see Gordon Hayward's Defensive Rating was one of the worst on the team at 110, and he got substantial minutes, so it's not like it's a bad sample size. I don't normally think of Hayward as being a poor defender when I recollect the time he frustrated Kobe and all the blocks he gets on transition layups, especially with the length he has and the balls we've seen him knock loose. However, since the DRtg formula weights defensive rebounds heavily, and Hayward usually plays SG or SF, he doesn't play a position on defense that puts him in position to get the defensive boards I suspect he would if he were defending more in the paint. Usually we see him trying to defend pick-n-rolls or close out on shooters. He still has a higher ORtg though (113), so that's good. It's also one of the highest on the team, behind Carroll and Evans.
Al Jefferson has the highest Defensive WS on the team at 3.3. WTF??? To most spectators, this would seem contradictory, but let's examine the formula. If you go here, you will find the algorithm for calculating defensive win share for a player. You'll notice that this calculation is dependent upon the product of the ratio (player minutes / team minutes) and team defensive possessions. The problem with this is just because Jefferson was playing more minutes than everyone else on the team doesn't necessarily mean his presence or skills led to defensive possessions. So hopefully Dennis Lindsey or Ty Corbin were not weighing this column too much when considering Big Al's defensive influence on the team. This is just one of many examples of how the numbers can be deceiving, one of the reasons Dennis Lindsey preaches "not painting by numbers". On top of that, DRtg is also used as a measure within the algorithm, and while Jefferson's DRtg is 104, DRtg heavily weighs defensive rebounds over more important stats like points allowed in the paint (a sign of lack of defense) or defensive stops while on the court.
In contrast to someone who had the most minutes played out of total team minutes played, Derrick Favors, who only played 1787 minutes to Jefferson's 2578 minutes, had a DWS of 2.9, tied for 2nd on the team with Paul Millsap, another player who received substantially more minutes than Favors. This bodes well for Favors, but doesn't reveal anything new to us. Everyone knows Favors is our future key defender in the paint. It also makes sense that Favors DWS was higher than his OWS. His game starts with defense, and ends with a transition alee-oop dunk. Favors did have the lowest DRtg on the team at 101.
Cheapest Players Contributing Wins
So if we go down the last column, we see that these players cost the least for the amount of wins they theoretically contributed to the team during the season based upon the WS calculation:
DeMarre Carroll $285,522.58
Gordon Hayward $501,800.00
Randy Foye $675,675.58
Derrick Favors $1,080,300.00
Paul Millsap $1,132,056.97
While Foye wasn't much for defensive help, I believe his 3-point shooting is worth only $2.5 million a year. I think most people would agree the only thing done wrong with Foye was playing him too many minutes. He'd serve better as a step-in, off-the-bench role player for those drawn up plays in the time out, using him to help spread the floor, and can a few 3's per game.
I don't have anything else to say about this data. Except maybe it's time professional basketball organizations redefine some of the metrics used in measuring things like offensive rating, defensive rating, WS, etc.