There's an article at FiveThirtyEight.com that had me thinking. It was written about a year ago and it was called "The Hidden Value of the NBA Steal." The point the writer makes is that if you want the best indicator of value a player has to a game, it's to look at how many steals he has.
Through his math, he concluded that with all other box score stats being equal
* an extra rebound is worth 1.7 points
* an extra assist is worth 2.2 points
* (one less) turnover is worth 5.4 points
* an extra block is worth 6.1 points
* a extra steal is worth 9.1 points.
For example, if a player averages 16 points and two steals per game, this player has the same impact - all else being equal - as a player that averages 25 points per game but just one steal. If the first player and the second player are both missing from their teams, we expect the two teams to decline similarly.
The writer has no explanation for why an extra steal is worth 9.1 extra points. You can come up with all kinds of theories. Maybe a player who has more steals plays a lot better defense. Maybe they're just smarter. Maybe it's because that unlike rebounds, a certain number of which will go to a player because they happened to be in the right place in the right time, players have to make steals happen. Steals can't be "replaced".
So even though the results of this NBA study might not be transferable to the WNBA, I figured what the hell. Let's create an "average" WNBA player who has
This player will earn zero points in the system. This is the per/game amount scored by the average WNBA player, aggregated over all games.
For every rebound past 3 rebounds, we'll give the player 1.7 points
For every assist past 2 assists, we'll give the player 2.2 points
and so on.
So a player that had the same statline as above but one extra rebound would score a +1.7. A player who had one extra rebound and one extra assists would score a +4.1. (1.7 + 2.2.) If the player has the same stateline as above but one few rebound, she would score a -1.7.
Let's call these points Morris Points after the author of the article above, Benjamin Morris. So under this system, who are the highest ranking players per game and the lowest ranking?
(See table at bottom of post for results.)
The next question becomes, "Do the numbers pass the smell test? IF I were to present the first five or first ten players on this list as the WNBA's best players in 2015, would you believe it?"
It seems believable.
* It gives Candace Parker the highest rating at +44.22. Note that this is a "per game" rating, so it doesn't imply that Parker is the MVP, merely that the Sparks are hurt more missing Parker than Chicago is hurt missing Delle Donne. The second half of the season certainly proved that.
* At #2 is Delle Donne with +35.24. Delle Donne is the MVP.
* We see some familiar names - Griner, Moore, CAtchings, McCoughtry - that we think of when we think of the best players.
* It has Sancho Lyttle ranked as #5 on a per-game basis, which explains the tailspin the Dream went in when she was off the roster.
* It puts Skylar Diggins at #10, which explains how much Tulsa was hurt by not having her.
* It puts Kiah Stokes at #21, and Jewell Loyd at #72. The Morris Points system values Ramu Tokashiki over Jewell Loyd. It gives both Tokashiki and Loyd negative value. Loyd's turnovers and lack of steals really hurt her.
Does this "prove" anything? I don't know. But it is good to talk about during the off-season.
|2||CHI||Elena Delle Donne||35.24|
|34||TOT||Erika de Souza||3.70|
|69||CHI||Clarissa Dos Santos||-4.28|
|112||TUL||Amanda Zahui B.||-10.68|