What Does NCAA Point Guard Performance Say About Pro Potential?

. Wednesday, April 15, 2009
Make a comment!

If you’ve ever paid any attention to any draft in any sport, you’ve probably figured out that drafting is by no means a scientific process – people often make arbitrary decisions or use shaky evidence to justify decisions.

I would argue that there is no position more difficult to unscientifically evaluate than that of the point guard, a position that people struggle to evaluate to begin with. I think a lot of people have gut feelings about what a good point guard “looks like” but have a much more difficult time describing what a good point guard “does”, much less what differentiates a good point from a great one.

Along these lines, I find Chicago Sky coach Steven Key’s comments about his selection of Kristi Toliver over Renee Montgomery particularly interesting.

"We knew we might have the option [Toliver or Montgomery]," Sky coach and general manager Steven Key told the Chicago Tribune. "We all just felt Kristi on the next level had more of our style of play. If you put them head to head, the numbers are pretty much the same. I think Kristi has a little more of that drive, that oomph and the ability to create her own shot."
Key’s comments seem to embody everything that makes drafting difficult, especially when drafting point guards.

First of all, there really are no strong predictors of how well a player will make the transition from the NCAA to the pros. There are so many intangibles like work ethic, adjusting to new coaching styles (not to mention lifestyles) and how well a given player “fits” with their team, not to mention factoring in differentials in college competition between any two given players.

Second, it’s not as if there’s one way to play point guard – there are many styles of point guard play and the extent to which one is more effective than another depends as much on the system and one’s teammates as it does on a given player’s skills. So the team’s style of play becomes as important as the player’s style.

Third, as a result of the first two challenges, people resort to statistics, which are always misleading for point guards if you look at the basic points, assists, rebounds, and steals, and not much more clear if you make the step to assist-to-turnover ratio.

Last, when the statistics fail to help in evaluating a player, the next move is to focus on intangibles that aren’t only difficult to measure but difficult to even define (I’m not sure exactly what “oomph” is, though I might agree that Toliver has it…I think…).

So in trying to figure out how well Toliver fits with the Sky’s system, I decided to take on these challenges, which I will follow up on throughout this season (thus embedding motivation to write more frequently into my own analysis).

The question: what, if anything, does NCAA point guard performance say about a player’s potential to be an effective pro point guard?

Here’s how I tried to answer this question (acknowledging that this is not necessarily statistically sound):

1. I compared the senior year numbers of this year’s group of rookie point guards to one another using the point guard statistics that I used to rank WNBA point guards last summer.
2. I compared this year’s group of rookie point guards to rookies from the past.
3. I tried to identify patterns (not correlations and certainly no causal relationships) in college performance that might serve as indicators of pro performance.

Ultimately, I came up with the following response to Coach Key (which I would love feedback on from readers):

1. Toliver and Montgomery are not “pretty much the same” statistically as point guards.
2. Based on her senior year statistics, if history serves us well, Toliver might be a solid WNBA player, but might be more effective as a scoring guard than the type of player who can run a team.
3. Statistically, based on past performances of WNBA point guards, Toliver might not even be a better point guard prospect than Briann January (who was drafted #7 by the Indiana Fever).

So obviously, I’m violating one of my fundamental rules – I’m looking at statistics without really observing the players extensively. Nevertheless, the statistical story here is actually pretty intriguing.

1. Toliver and Montgomery are not “pretty much the same” statistically as point guards.

When Coach Key made the argument that Toliver and Montgomery are “pretty much the same” statistically, it’s likely that he was looking at basic averages (pts, rebs, asts) so it’s not that his claim was way off, though each player clearly provides different strengths:

Montgomery: 15.8 ppg, 6.03 apg, 1.93 rpg, 44% FG
Toliver: 18.4 ppg, 5.48 apg, 3.35 rpg 44% FG

Since we know that the basic statistics say very little about the quality of a point guard, it seems to make sense to at least attempt to compare the two as point guards since that is the argument being made (again, I acknowledge that there really is no way to evaluate a whole host of intangibles a point guard brings to the court).

So even if we are to start with assist-to-turnover ratio – a very flawed statistic – we see increasing separation between Montgomery and Toliver:

Montgomery: 1.99 a/to
Toliver: 1.34 a/to

But given the arguments against that statistic, it may serve us to move beyond assist-to-turnover ratio. And it sort of starts with the fundamental question of what it means to be a good point guard. Subjectively, you want someone who can handle the ball, get the ball to her teammates, run the offense, make good decisions with the ball.

Last year, I used a set of statistics to rank point guards, four of which are attainable for college players (descriptions and formulas available in the RB statistics glossary):

Points per zero point possession (p/zpp): how often is a player individually responsible for scoring possessions and non-scoring possessions? (in other words does a player make good decisions with the ball).

Assist ratio (ast rat)
: how often does a player create a scoring opportunity for others (as opposed to scoring or turning the ball over)?

Pure point rating (ppr)
: how well does a player create scoring opportunities for others?

Usage % (usg)
: what percentage of a team’s plays does a player “burn” – either scoring or turning the ball over? (can be a good indicator, along with the other statistics of whether a player is “ball dominant” or really moving the ball around)

So what happens when we look at those stats? We see that Montgomery and Toliver are actually different types of point guards.

Montgomery: 2.64 ppr, 2.11 p/zpp, 23.16% ast rat, 25.50% usg
Toliver: -1.03 ppr, 2.01 p/zpp, 20.12% ast rat, 25.42% usg

The number that stands out is the huge gap in pure point rating. I will confess it is not a perfect statistic (which is why I don’t like using it alone) but what we see here is that Montgomery was much better in college at creating scoring opportunities than Toliver.

We also see that there is a differential in assist rate, though it might be hard to put that in perspective without a comparison to other guards. So to help put that in perspective: last season in the WNBA, the average assist rate for the point guards in my top 25 was 26.82 as of August 3 (Sue Bird, Kelly Miller, and Lindsay Whalen were all about average). I don’t have the average for NCAA point guards, but neither Montgomery nor Toliver should have been expected to have a high assist rate in their senior year as they also did a considerable amount of scoring.

So if we take it a step further and apply the point guard styles framework that I created last year, Toliver is something of a scoring point guard whereas Montgomery is closer to being a distributor who can score. To summarize these statistics, they might both be good WNBA players, but Montgomery seems to be much better at running a team, which I argue is what the Sky currently need.

2. Based on her senior year statistics, if history serves us well, Toliver might be a solid WNBA player, but might be more effective as a scoring guard than the type of player who can run a team.

Yet despite the statistical evidence, Coach Key seems to have a different perspective on Toliver:
“Toliver is a true point guard,” said Sky General Manager and Head Coach Steven Key. “Our point guards in the past have done a great job for us but I think Kristi will make us that much more versatile and allow us to move Dominique Canty over to the shooting guard role as needed. Canty is very good at getting to the basket; adding Kristi and some shooting to the team will help create some space for our inside players in Sylvia [Fowles] and Candice [Dupree]. It’s a great selection for us!”
But how can we really determine whether Toliver will be effective at running a WNBA team? Or better yet, how can we determine whether she might be better than Canty, Jia Perkins, or KB Sharp?

Well, one thing we can do is compare her statistics to those of past college point guards…but I do that with an obvious caveat.

Clearly, college statistics are difficult to compare because players have vastly different schedules, play with a wide array of supporting talent, and are asked to do very different things within the system. For example, an argument could be made that Toliver was not very focused on creating for others because she was asked to score for Maryland. However, even with that caveat, I think there’s a convincing argument that college point guard styles do have some bearing on how a player will play at the next level.

So without further delay, here are the senior year statistics for the point guards from my top 25 in WNBA compared to Montgomery, Toliver, and the rest of this year’s rookie point guard class. As I’ve done with these stats in the past, I simply ranked them by category just to get a sense of their relative value, which by no means implies that there might be some objective standard.

Name PPR Rank Pts/ZPt Rank Ast Ratio Rank Usage Rank Total
Bird, Sue 5.22 22 2.45 24 0.30 22 20.92 9 77.00
Mitchell, Leilani 4.46 20 2.06 16 0.30 21 25.97 18 75.00
Whalen, Lindsay -0.38 11 2.40 22 0.21 12 30.88 23 68.00
Renee Montgomery 2.64 16 2.12 19 0.23 15 25.51 17 67.00
Taurasi, Diana 2.72 17 2.24 20 0.23 14 25.49 16 67.00
Miller, Kelly 3.31 19 2.49 25 0.25 16 19.70 7 67.00
Briann January 3.25 18 2.08 17 0.28 18 22.40 11 64.00
Sharp, KB 5.34 23 2.02 14 0.28 20 19.68 6 63.00
Wiggins, Candice 0.78 14 2.43 23 0.14 2 30.44 22 61.00
Johnson, Temeka 7.59 25 1.71 7 0.38 25 19.28 4 61.00
Shalee Lehning 4.51 21 1.66 5 0.38 24 20.28 8 58.00
Whitney Boddie 6.87 24 1.64 4 0.36 23 19.60 5 56.00
Harding, Lindsey 1.00 15 1.97 11 0.20 10 23.45 12 48.00
Perkins, Jia -2.74 2 2.40 21 0.11 0 31.92 24 47.00
Johnson, Shannon -2.95 1 2.08 18 0.13 1 34.06 25 45.00
Quinn, Noelle -0.37 12 1.60 3 0.21 11 27.18 19 45.00
Kristi Toliver -1.03 7 2.02 13 0.20 9 25.42 15 44.00
Sha Brooks -1.51 5 1.75 8 0.20 7 29.24 21 41.00
Canty, Dominique -3.11 2.05 15 0.16 4 27.76 20 39.00
Blue, Nikki -0.83 9 1.40 1 0.26 17 22.06 10 37.00
Camille LeNoir -1.25 6 1.70 6 0.20 8 25.40 14 34.00
Nolan, Deanna 0.24 13 1.99 12 0.18 5 18.89 3 33.00
Latta, Ivory -2.43 3 1.81 9 0.19 6 23.98 13 31.00
Moore, Loree -0.88 8 0.99 0 0.28 19 14.43 0 27.00
Brown, Kiesha -0.47 10 1.86 10 0.14 3 17.72 2 25.00
Bobbitt, Shannon -1.84 4 1.59 2 0.22 13 16.34 1 20.00

Here are the two patterns that seem to stand out:

First, having a college pure point rating above 2.5 seems to mean something. All of the point guards in the top of the list -- who are mostly initiators, distributors, facilitators, and highly effective “combo” guards – had relatively high pure point ratings. The exceptions are Whalen and Candice Wiggins who you may remember were also big time scorers in college (both with usage percentages of 30% and both among the most efficient scorers on this list).

Conversely, having a pure point rating below 2.5 does not bode well for one’s effectiveness managing a team. The exceptions are Loree Moore (ACL surgery before here senior year) and Shannon Johnson (who like Whalen and Wiggins was relied upon as a big time scorer for her team and had a usage percentage of 34%). Deanna Nolan did not really play point guard in college (as far as I know, given the presence of Coco and Kelly Miller) so it’s hard to hold her accountable for point guard statistics.

The lesson: if you’re going to have a low pure point rating, it helps to have a high usage percentage to demonstrate that you were relied upon as a scorer.

Second, the average assist ratio among this set of point guards is 23%. If you’re above that (with the exceptions noted previously) it seems like you’re in the company of playmakers. Fall below that, and it appears you’re in the company of shooting guards trapped in point guards’ bodies. Montgomery seems to be on the border, Toliver seems to be among the scoring crowd. Toliver might be a better playmaker than Canty, but neither (nor Perkins) appears to be the type of point guard you would want running a team consistently.

So there might be some hidden qualities within Toliver that make her able to manage a WNBA team, but the combination of college statistics she accumulated doesn’t seem to bear that out.

3. Statistically, relative to the past performances of WNBA point guards, Toliver might not even be a better point guard prospect than Briann January (who was drafted #7 by the Indiana Fever).

Would I draft Briann January ahead of Kristi Toliver?

If I was looking for a point guard to run my team, then yes – I would just trade down and pick up January. The odds are that a player with January’s statistical profile will be more effective at running a team than Toliver, if you take into account their statistics relative to others. Toliver might turn out to be a great player, but based on her college numbers, she seems destined to replicate what the Sky already get from Canty and Perkins at the point guard spot.

Again, a caveat: January and Sharp have very similar profiles, which means that Janurary may be nothing more than a glorified initiator. Really, Montgomery is somewhat hard to figure based on these statistics as well.

Furthermore, I realize I’m comparing across styles in this analysis. It’s quite possible that certain styles fare better than others. But there’s still an interesting little tidbit that may seem superficial but seems to stand up to a test of reason.

If you look at the total points in these rankings that I constructed, there are two gaps: Sue Bird and Leilani Mitchell were both pretty outstanding college point guards in a class of their own. There is a drop off and then a second tier between Whalen and Whitney Boddie. Then there is another drop off and a third tier between Lindsey Harding down to Shannon Bobbitt.

Given the caveats for Shannon Johnson, Loree Moore, and Deanna Nolan, is there really a point guard in that third tier that you would anoint as your starting point guard? Dream fans might point to Ivory Latta, but she has not proven herself just yet as a consistent (winning) starter at the point.

You may also be wondering, what about all the point guards with great college stats that just didn’t make it in the WNBA? Were there other anomalies?

I went through the first round of WNBA drafts back to 2001 and identified all of the players that came out of college as point guards and ended up playing point guard in the WNBA. Here’s what I found (in no particular order):

PPR Pts/ZPt Ast Ratio Usage
White, Tan -4.95 1.90 0.12
Swanier, Ketia 3.09 1.62 0.33 15.52
Baker, Sherill -1.32 2.44 0.15
Thorburn, Shona 5.19 1.53 0.32 21.35
Haynie, Kristin 3.81 1.82 0.30 23.98
Wright, Tanisha -1.61 1.86 0.14
Jacobs, Amber -1.52 2.07 0.22
Creamer, Molly -4.28 1.92 0.17
Brown, Coretta -1.56 1.71 0.21
Curtin, Allison -4.27 2.00 0.16
Miller, Coco 0.62 2.20 0.15 23.34
Dales, Stacey 0.00 1.77 0.18

So while those with stats in that third tier seemed to maintain the third tier pattern, there does seem to be some variation in second tier players – most notably, Kristin Haynie, Ketia Swanier, and Shona Thorburn.

Swanier is an interesting case not only because she spent her rookie year on a very deep Connecticut Sun team that changed rotations a number of times, but also because she was not a full time starter in her senior year of college. When you look at her low usage numbers, it’s possible that she simply didn’t do enough statistically in college to evaluate her based on these stats. In any event, it’s hard to say she’s a failure yet.

I looked up the history of Thorburn and it appears that she had injury problems more than performance problems.

So that leaves Kristin Haynie. She is a major anomaly, even moreso than KB Sharp. There is nothing that really stands out in the statistics or her history in the WNBA as an explanation.

Wishing Toliver the Best of Luck

Regardless of whatever the statistics say, I do hope Toliver does well, if for no other reason for the sake of the Sky. I like them. I want to see them win. I have nothing against Toliver. It just strikes me that she might not provide the Sky with what they need.

I'll look forward to hearing more critiques and observations...

Transition Points:

I was unable to find turnover and minutes statistics for Ticha Penicheiro, Vickie Johnson, Becky Hammon, and Katie Smith, all players I had in my WNBA point guard rankings last summer. If anyone has any leads, let me know. I’d be curious.

Comparing Toliver’s stats
to those of past prospects was interesting – the best player with a similar statistical profile seems to be Lindsey Harding whereas the bottom seems to be Amber Jacobs.

What, if anything, do these stats say about Shalee Lehning and Whitney Boddie
? Not sure, but Bob Corwin seems to believe that at least Boddie has a shot at being a solid backup for the Monarchs, whereas Lehning might be caught in a numbers game.

Leilani Mitchell
was an impressive college prospect. GMs really snoozed on her.

I chose to focus on senior stats because they're what jumped out at me. But here are some other angles I thought about taking: career statistics; the rate of change between years; strength of schedule; some way of figuring out how much the player contributed to team wins.

John Hollinger has suggested in the past that steals are a strong indicator of guard college to NBA success as a proxy for athleticism. I chose not to do that here, but it would be interesting to add to the mix.

Continue reading...

Kudos to Women's NCAA BB for Actually Graduating "Student-Athletes"

. Tuesday, April 14, 2009
Make a comment!

The other day I posted about how the WNBA's age limit is a good thing in terms of keeping basketball in perspective and making sure female basketball players stay in school relative to their male counterparts.

Tangentially related (though not necessarily related at all) is this report from the Institute for Diversity and Ethics in Sport:

...the numbers in the 2009 Graduation Rate and Academic Progress Report re-emphasize the results of the smaller compilation. Sixty one women's teams (98%)graduated at least 50%, compared to 40 men's teams (63%.) 97% (60) women's teams graduated at least 60%, compared to 30 (48%) of the men's. Fifty women's teams (81%) graduated at least 70%, only 22 (35%) of the men's teams could point to similar results.
I think education is good. So this is impressive. Hope the women's side continues this trend.

Continue reading...