Who are the best 'shot creators' in the WNBA?

In a previous piece about potential assists in the WNBA, I mentioned that I didn't have the numbers to calculate a metric SB Nation's Tom Ziller created called "creation ratio" (CR).

I took a bit of a break from writing in between the women's pro and college seasons, but I did get the numbers I needed in that time and thought I'd share what these numbers say about who the top "creators" are in the WNBA...and what that even means.

What is Creation Ratio?

Put simply, CR is a measure of a player's ability to create shots for themselves. It is the ratio of a player's shots created for herself or others compared to the shots others create for them. The actual formula is somewhat elaborate*, but is essentially a player's assists plus estimated unassisted field goal and free throw attempts over their estimated assisted field goal and free throw attempts.

Ziller uses his creation to take a close look at a recent debate about whether Oklahoma City Thunder forward Kevin Durant's ability to create shots for himself is truly important, but adding assists** to what it means to "create shots" is what stands out for me: creation ratio suggests that NBA point guards create three times as many shots as other players.

Ziller described how CR works further at SBN's Sactown Royalty: a player gets a low CR if she is set up on most of their shots, and if she doesn't create many shots for herself or her teammates. A player gets a high CR if she sets up her teammates or creates her own shots frequently, and if she doesn't rely on teammates to set her up. A creation ratio of 1 means a player creates as many shots as they are set up on (that turns out to be the median in the WNBA, belonging to Washington Mystics center Nicky Anosike).

As in the NBA, point guards are responsible for creating a higher ratio of shots than any other player in the WNBA - while 27 the top 30 CRs in the NBA are point guards, 13 of the top 15 CRs were point guards in the WNBA***. Also similar, point guards led the way with the highest average CRs in the WNBA this past season, though not quite tripling the highest average CR of the next closest position as their NBA counterparts did.

So with these numbers, it's fair to confirm the simple assumption I made about the WNBA based on the NBA CR numbers: point guards are responsible for creating more shot attempts than any other position, arguably making them - and their particular skillsets - more valuable than they're often given credit for.

Yet that doesn't mean those poor centers stuck inside waiting for someone to pass them the ball are bad players while guards are inherently good.  But rather than just ramble through that, let's instead ramble through an interesting comparison of two point guards to better understand what this metric means.

A tale of two point guards

Key factors for the creation ratio of two players. Click here for a summary of what those stats mean.

Player A would normally be considered the better player by metrics such as PER, WARP, etc and she has the fifth highest usage rate of any point guard with a solid shooting efficiency. However, she has a below average free throw rate and assist ratio for her position in addition to the second lowest unassisted field goal percentage among point guards.

So although player A touches the ball at among the highest rates per game in the WNBA and uses more possessions trying to score than most point guards, she ends up with the fourth lowest creation ratio of any starting point guard (although still above league average).

In contrast, Player B is quite clearly one of the purest distributors in the league with the highest assist ratio of any point guard and the fith lowest usage rate. But while her primary emphasis on the court is clearly setting up others, she has the highest unassisted field goal percentage of any point guard.

That low usage, pure distributor, high unassisted field goal percentage combination sort of makes sense intuitively - Player B is obviously relied upon fairly heavily for setting up her teammates, meaning there might not be a whole lot of people looking to set her up for scoring opportunities.

Yet that Player B has the highest CR in the WNBA as someone whose statistical profile almost epitomizes the anti-scorer is sort of counter-intuitive, even if it makes sense by the way this metric is constructed mathematically.

And putting names to the numbers might make things more surprising: Player A is Seattle Storm point guard Sue Bird while Player B is Los Angeles Sparks point guard Ticha Penicheiro.

So does that mean that Penicheiro, notorious for her reluctance to shoot, is not only a bigger offensive threat than Bird or the biggest point guard threat, but also the biggest offensive threat in the league?

Well, although this metric might seem to make that argument, it might also be an overstatement.

Why touches matter

Touches* are a good place to start in understanding what separates Bird and Penicheiro. Point guards obviously touch the ball more than anyone else giving them more opportunities to create shots: they bring the ball up court and therefore have the first opportunity to make something happen. That much is obvious. But Bird also gets about nine touches per game more than Penicheiro and what she does with those touches is important.

Just to put things in perspective, Bird had more unassisted shots made (98) than Penicheiro had total shots made (72). Equally interesting is where the two players get their shots from.

So although Penicheiro has one of the highest unassisted field goal percentages in the league, we know she gets the vast majority of those on shots right around the basket and - given her free throw rate - it's safe to assume that she often gets those off drives to the basket.

Most of all, anyone who has ever watched Penicheiro play a game at any time in her basketball career knows that she's not a player who looks to shoot often - Bird shot more threes than Penicheiro shot total. She wasn't much of a shooting threat outside the key (she shot 30% from the 6-20 foot ranges). And you're not going to find a whole lot of film of any team running her off screens to get shots either.

So her CR is high because she racks up assists at a higher rate than anyone, gets herself to the free throw line off drives, and is rarely even in the situation to score off of an assist; she creates more shots than get set up for her, but setting her up for shots just isn't what teammates are looking to do.

Bird, in contrast, was the Storm's primary scoring threat. It's not at all that she's incapable of creating shots - she had more unassisted field goals than all but the two point starting point guards in the 2011 WNBA Finals (Atlanta Dream point guard Lindsey Harding and Minnesota Lynx point guard Lindsay Whalen). The Storm just also set her up for a number of assisted field goal attempts - she had more of those than any point guard in the league, which is a skill unto itself that shows that made her arguably the biggest point guard scoring threat with and without the ball.

What Bird did for her team offensively creating shots and having shots "created for her", so to speak, is a large part of what made her a MVP candidate and almost indisputably what got the Storm to playoffs despite center Lauren Jackson's absence. On the other hand, what Penicheiro has done throughout her career - without putting much pressure on a defense as a shooting threat - is also quite remarkable, almost magical at times in how well she's able to create scoring opportunities for others while defenders sag five feet off of her. She doesn't have the ball in her hands often, but she has the ability to make things happen when she does.

But let's stay with the idea of touches and look at this thing a little bit differently.

Who creates the most shots?

Again, CR is a ratio of shots a player creates to shots created for them. But aside from the math of it all, it sets up results that can be hard to wrap your head around. Perhaps more interesting on an individual level is simply the top half of that equation: shots created.

Of course, we already have a widely used metric to look at shot creation: usage rate.

A usage rate tells us the rate at which a player puts up a shot while on the court. As is obvious from watching Penicheiro, her 14% usage rate is low; given how Bird plays, it would make sense that her usage rate is around league average. It also probably comes as little surprise that the highest usage rate of any point guard in the league belonged to San Antonio Silver Stars guard Becky Hammon, who shoots so often that people often debate whether she even is a point guard*****.

Usage rate is a reasonable approximation of a player's skill in creating shots (and, of course, lacking any conscience in some cases), which is valuable in women's basketball - ball movement is beautiful and all, but at some point someone has to put up a shot. Even in a league with less one-on-one scoring plays, having players that can get shots up helps an offense function.

"Shots created" as set up by Ziller tells us something different. Rather than telling us how often a player shoots, it more narrowly tells us how many shots a player creates independently. That is interesting, perhaps particularly in terms of the NBA/WNBA comparison again - whereas 15 of the top 20 NBA shot creators were point guards, only 8 of the top 20 shot creators in the WNBA were point guards******.

The problem with shots created is that players playing big minutes on fast paced teams put up more shots partially as a result of just getting more chances. Hickory High put together a complementary metric called "Individual Creation Percentage", which is an interesting way to look at what percentage of shots a player create independently.

But returning to the idea of what a player does when they actually touch the ball, shots created per touch might be a more interesting thing to look at - that would get us a bit closer to some understanding of which players are the biggest threats to create a shot with the ball in their hands, independent of pace of play or number of opportunities. The top 20 are as follows:

Obviously, some of these players played inconsistent minutes or played for poor offensive teams. And I probably need not help avid WNBA fans distinguish between those players who are good scorers and those who toss up shots indiscriminately.

But New York Liberty guard Essence Carson does stand out as a bit of a surprise: how many people would consider Carson a better shot creator than a player like Angel McCoughtry?

What separates Carson and McCoughtry is again unassisted field goal attempts: although McCoughtry does get to the free throw line far more often, Carson's unassisted field goals made percentage was 60% whereas McCoughtry's was 49.8%. Nevertheless, McCoughtry was in the MVP discussion for a reason: nobody took on a larger responsibility for their team's offense on a per possession basis (usage rate) than she did. And that's noteworthy.

So, again, what exactly does all of this tell us?

The best thing you can say about CR is what Ziller already said and what many basketball fans already probably feel but struggle to quantify - point guards are important to a basketball team's ability to generate offense. Whether in the NBA or WNBA, point guards create around 3 shots to every they get help creating. Since the WNBA is a different game, it's probably a stretch to say that the WNBA is a point guard's league - the versatility of its players shows up pretty strongly just in who the top shot creators are. And individually, it gets even tougher to make a definitive statement about who the best shot creator is.

It might be fair to say that when considering passing and scoring, Penicheiro truly is the best creator in the game in that she makes things happen for her team more often than most players. But that doesn't at all negate the notion that Pondexter is the arguably the most dangerous one on one scorer in the game - her creation ratio (3.51) is fifth in the league (the only non-point guard in the top 10) and she created the most shots in the league. From another perspective, Carson created more shots per touch than any player that played regular minutes. 

Still, what's frustrating about about all this shot creation stuff is the question of whether it really matters how a player gets shots - if you were looking for a scorer, would you choose Sparks guard Noelle Quinn (1.1 CR) over Lynx guard Seimone Augustus (1.06 CR)? Or would you choose Tulsa Shock guard Ivory Latta at point guard over Bird (26.65 shots created) if you wanted a scoring point guard? Probably not.

That Augustus has the most assisted field goals made after Connecticut Sun center Tina Charles and Chicago Sky center Sylva Fowles means something. Just getting open off the ball on the perimeter and hitting shots is a skill that Augustus has mastered unlike few in all of basketball, men's or women's, and that's a skill the Lynx won't soon want to part with. Something similar could be said about Bird's scoring ability relative to other point guards.

But what got me interested in this is that I ultimately agree with Hickory High's conclusion - these numbers do give us one more way to think about what it means to be good at "creating shots", which helps not only in thinking about a player's value to their team but also maybe some insight about what it means to build a good team.

We can probably agree with the basic principle that every team needs to strike some kind of balance between shot creators and those who are efficient scoring with the help of their teammates. These numbers at least suggest a slightly different way of thinking that than the norm.

Relevant Links:

2011 WNBA Creation Ratios

The Subtle Value Of Great Point Guards: Why Potential Assists Matter

Notes:

* What's "elaborate" about this formula is that un/assisted field goal attempts obviously aren't recorded - if the shot doesn't go down, no assist is awarded, and potential assists aren't counted. So un/assisted field goal attempts are estimated using a player's un/assisted field goals made percentage. In addition, the free throw attempts are weighted with a standard .44 coefficient but also adjusted to represent the fact that free throw attempts are only half as likely to be assisted. So essentially, the formula is this: (unassisted field goal attempts + unassisted free throw trips + assists)/(assisted field goal attempts + assisted free throw trips). To keep it simple, it's: shots created/shots others created for them.

** This brings up an interesting question about whether an assist should be weighted: is an assist really worth as much as a shot? John Hollinger suggests no - in the formula for his pure point rating, he only counts 2/3 of assist. You could argue an assist might be worth more in the WNBA. But the fact is that although weighting assists drops the average NBA point guard's CR from 3.60 to about 3.08 - meaning they don't create three times the shots of any other position - everyone else's drops too so NBA point guards remain way ahead of the pack. So in the interest of remaining consistent with the original articles for the sake of comparison, I'm not going to weight assists here.

*** To make it roughly equivalent, these "top CRs" are those in the 90th percentile or above based for players that played semi-relevant minutes in each league. Why is that at all relevant? Interestingly, the 80th to 90th percentile diversifies a bit in both leagues after those "elite" creators: 22 of the next 30 players (73.3%) in the NBA are point guards, while it's 8 of the next 15 (53%) in the WNBA. You could dig further into this matter, but it sort of further reinforces the notion that skillsets are distibuted across positions a bit more fluidly in the WNBA than NBA (which might be more accurate than saying the WNBA is more "pure").

**** The "touch" is yet another estimated number and that is no more apparent than it is with point guards - clearly a point guard touches the ball on almost every possession. This metric can only estimate based on quantifiable actions - if the point guard comes down the floor and just initiates the offense without scoring, turning it over, or picking up an assist, it's not counted. So it very likely underestimates a player's touches, but it also does so for everyone. So consider it a good way to roughly understand where players stand relative to one another, if not the exact number of times their hands are on the ball.

***** I'm not even sure these distinctions matter much - we could all just agree Hammon is very good and leave it at that - but you'll have a hard time convincing me that Hammon is not a point guard based upon her performane in 2011.

****** There's a lot that could be said about this too in terms of the blurring of positions in the WNBA - for example, the top 21 in the WNBA includes Cappie Pondexter (league-high 581.29 shots created), Katie Douglas, Diana Taurasi, Penny Taylor and Tanisha Wright. All of those players are responsible for running their team's offense in different ways at times. Then you have players like Matee Ajavon and Tamika Catchings who also took on ball handling duties for their teams depending on the lineups. So again, that sort of adds to the point that the game flow and patterns in the two leagues are different, for those looking for "objective" evidence.