What Are NBA Win Shares?


[embed]https://youtu.be/jCohKyjipUA[/embed] In the current 2015-2016 season, Kawhi Leonard is averaging about 21 points, 7 rebounds, 2 and a half assists, and nearly 2 steals. Gordon Hayward is averaging about the same at 20 points, 5 rebounds, 4 assists, and 1 steal. They’re very comparable players when you look at their stats. However, very few people would say that Gordon Hayward is on the same level as Kawhi Leonard, who’s already won an NBA Championship, along with Finals MVP and Defensive Player of the Year honors. But how do we know they aren’t on the same level? Do we just know, ‘cause we're all basketball geniuses? What if I didn’t give you the names of the players? How would you know which one was better? Believe it or not, advanced analytics can give a pretty clear answer.

gordon hayward vs kawhi leonard

Now, when it comes to analyzing basketball, there are generally two opinions: Numbers never lie and Numbers don’t tell the whole story.

For those who say that numbers never lie, well, they kinda just did. Just because Kawhi Leonard and Gordon Hayward average nearly exactly the same numbers does not mean they are equal players.

The people who subscribe to the second train of thought cling to what many call the “eye test”. This super-scientific method says that you can tell how good a player is by watching him play. Now, to a certain extent this is true and an important element of judging players. However, what we see is really subject to perception. The “eye test” should really be called the “I test”, because it basically comes down to what I think based on what I see.

So, what’s the point? The point is that you have to take both aspects into consideration. But you have to be looking at the right things. The right numbers, and the right film. Let’s look at one very important number: Win Shares.

What are win shares? In a nutshell, Win Shares is an advanced stat that calculates how much a player contributes to, well, winning. Because that’s all that matters, right? Teams recruit players because they think they can help them win. You play to win the game. Winning is the end goal. Steph Curry didn’t win MVP because he was a good 3-point shooter. There were players who shot better than him from the 3-point line. He won MVP because his team won games, largely due to his contributions on the court. Win Shares helps us measure that. In fact, in the 2014-2015 season, when he won league MVP, Steph Curry had the most win shares per 48 minutes.

Ok, so how are win shares calculated? First, offensive win shares and defensive win shares are calculated.

Offensive win shares takes into account the total points a player scores, how many offensive possessions that player had, the average league points per possession and points per game, and the team pace and league average pace.

offensive win shares

Defensive win shares takes into account a player’s Defensive Rating (which is a whole other story), the number of defensive possessions the player had, the league average points per possession and points per game, and the team’s pace and league pace.

The sum of a player’s defensive win shares and offensive win shares is his total win shares.

As you can see, Win Shares is not a one-dimensional stat like points per game or assists per game. That’s why it’s called an advanced stat, it takes everything into account and makes it all make sense. When people say numbers don’t tell the whole story, it’s because they’re not looking at the right numbers.

But does this actually work? Well, let’s go back to the two players we mentioned at the beginning of this video: Kawhi Leonard and Gordon Hayward.

gordon hayward vs kawhi leonard 2

At this point in the 2015-2016 season, Gordon Hayward has 7.2 win shares. Kawhi Leonard has 11.8. Gordon Hayward’s win shares per 48 minutes is .144 while Kawhi Leonard’s is .283. The numbers tell us what we already know: Kawhi Leonard is a better player because he helps his team win more games than Gordon Hayward does.

Now, you could say, “Well, the Spurs are a better team than the Jazz. They win more games so Kawhi Leonard gets more win shares.” Well, as you probably noticed earlier, Win Shares does not include anything that would reflect how well the team performs, only how well the player performs and contributes to team wins. For example, Kevin Love averages about the same amount of minutes as Gordon Hayward does, but for a much better team. Better record equals more win shares, right? Actually, no. Kevin Love has 7 win shares to date. DeMar Derozan also plays for a much better team, but only has 8.2 win shares to date. So, while Gordon Hayward may be limited by quality of his teammates, how much of the team’s losses fall on his shoulders and would he really be better on a better team?

Screen Shot 2016-03-17 at 7.15.31 PM

A classic example of why win shares matters in evaluating a player’s value is Chris Bosh. In Toronto, he was The Man. He averaged 22-24 points and 9-10 rebounds. His win shares per 48 minute was .158, much lower than your typical superstar on a championship team. Now, we all know Minnesota didn’t have great teams in those days. But even with no else good enough to compete for productivity, Chris Bosh still didn’t contribute much in terms of wins. Once he moved to Miami and was on a better team with better teammates, teammates that can compete with Bosh for those points, rebounds, and productivity, he no longer averaged 24 and 10. For those who kept track of his win shares numbers, that wouldn’t be a surprise since they would already know that he wasn’t the type of player that could significantly contribute to a winning team. It’s not called sacrificing numbers for the sake of winning, it’s called not being that good in the first place.

At this point, you may be wondering, ‘What qualifies as a great Win Shares average? What is a bad one or an average one?’ To answer that question, we looked at some of the all-time greats to see what their Win Shares were. Below is a chart of the top 10 in career Win Shares Per 48 Minutes:

ws48min c

There are three things to note here:

  1. This is a career ranking so it will include any seasons were the player suffered injury who whose quality of play took a steep decline due to age
  2. We can see how accurate the Win Share stat is. Outside of Neil Johnson and (possibly) Chris Paul, this could easily be anyone’s rankings of the top 10 greatest players; it’s not surprising that they also have the highest career Win Shares Per 48 Minutes
  3. This is per 48 minutes; Total Win Shares would favor players who had longer careers.

Taking point number one into consideration, let’s look at the single season leaders for Win Shares Per 48 Minutes:

ws48min s

As a general rule, a Win Shares Per 48 Minutes number of .2 or higher is “elite”. A Win Shares Per 48 Minutes number of .14-.17 is “average” and anything below .12 is no bueno.

Now that we have a clear understanding of Win Shares and what they mean, let’s see how the win shares stat can help find the truly valuable players. We compared 14 players on 12 different teams with different records; some good, some not-so-good. Then we ranked them based on their win shares per 48 minutes. Now, of course, players like Steph Curry, Kevin Durant, and LeBron came out on top, that’s not eye-opening. But we can also see how a star like Jimmy Butler, despite being on a team with a mediocre record, can shine, thanks to the win shares stat. Even players like Anthony Davis, Damien Lillard, and James Harden, who are on average or disappointing teams record-wise, show their true value. On the other hand, if you were wondering why the Kings and Knicks are struggling, it’s because their stars are not contributing to winning much at all.

win shares chart

So let’s recap:

  • Win Shares is an advanced stat that isolates a player’s individual production offensively and defensively and shows how much that player contributes to winning.
  • Win Shares can be used to evaluate a player’s true value, when basic stats and the all-inclusive “eye test” isn’t enough.
  • Win Shares isn’t the only stat or variable to take into account when determining a player’s value, but it’s a good place to start.