Illuminating the Unseen: Statistics Show Point Guards Are Key in Lynx-Mercury Game

. Thursday, September 4, 2008
  • Agregar a Technorati
  • Agregar a Del.icio.us
  • Agregar a DiggIt!
  • Agregar a Yahoo!
  • Agregar a Google
  • Agregar a Meneame
  • Agregar a Furl
  • Agregar a Reddit
  • Agregar a Magnolia
  • Agregar a Blinklist
  • Agregar a Blogmarks

I normally don’t follow games that I can’t see, but last night I had the urge to follow the Lynx-Mercury game.

As I’ve written previously, I’m sort of rooting for the Mercury to make the playoffs because it would be a great story for the league if Olympian Diana Taurasi can lead her team to the playoffs. And hey, when they’re on, they’re one of the most exciting teams in the league.

So I decided to casually follow the play-by-play of the game at Yahoo Sports as I watched the Serena-Venus tennis match and later tried to deal with the reality that Sarah Palin is a major threat to the Democratic Party. I put the quarterly team stats into an Excel spreadsheet to calculate the team dynamics ratings for each team and after the game looked at the individual stats to calculate the credit each player deserved, as I did for the Olympics using a method created by David Sparks.

Yes, I know – I’m a basketball junkie, but you already knew that.

Anyway, two things stood out in the numbers – point guards Kelly Miller and Lindsey Harding were the big stories of this game and we see once again that the Mercury are better when they move the ball to complement their individual play.

Points aren’t the only thing to pay attention to in the box score

Normally when we see the headlines about games the top scorer is highlighted as the key figure in the game. Look at the box score and pick out the player for each team who seemed to have the best game. Having not seen the game, you’d probably pick Taurasi and Augustus. They’re both great players who we’d expect to play well and they scored a lot of points.

But as Dave Berri and others have suggested elsewhere, we tend to overvalue points as a means of evaluating productivity – there are other things that are equally if not more important, even in the box score.

So I have used David Sparks’ credit formula in the past to get a better sense of how to assign credit to players for wins. It was helpful to pick out Diana Taurasi’s importance to Team USA in Beijing, but last night it showed that Taurasi should probably have shared the spotlight. Here are the numbers for the Mercury:

Miller: 15.08%
Taurasi: 14.53%
Pondexter: 14.82%
Willingham: 12.35%
Smith: 7.59%
Farris: 2.62%
Mazzante: 2.61%
Pringle: 0.02%
Scott: -0.18%

So how is it that Miller deserves more credit for the win when Taurasi put up such gaudy numbers? Two reasons – this formula values assists more than points and missed field goals and turnovers are the two most harmful statistics. In other words, it (appropriately) devalues volume scorers who have to shoot a lot to score and puts a premium on efficiency.

Miller had 5 assists, only 2 missed shots, and 2 turnovers whereas Taurasi had 3 assists, 7 missed shots, and 8 turnovers. It’s not that Taurasi had a bad game, but in terms of doing the little things that help her team win, Miller did a bit more and put up 20 points as well. It was quite an impressive all-around game.

But the Lynx’s numbers are an even more stark example of how sometimes we overvalue points and the most valuable players aren’t the ones who score the most points. While Augustus and DeForge are probably the players who stand out most in the box score, it’s possible that neither deserves the most credit for the Lynx’s effort.

Harding: 14.17%
Anosike: 13.27%
Augustus: 13.05%
DeForge: 13.00%
Houston: 6.27%
Rasmussen: 3.14%
Wiggins: 2.51%
Ohlde: 1.48%
Quinn: -0.064
Moore: -1.25
Hayden: -3.45

This is really interesting because while Augustus and DeForge scored 27 and 23 points respectively, Lindsey Harding only scored 5. So why does she deserve the most credit for this game? First, she only missed one shot meaning she didn’t expend possessions with missed shots. Second, she had five steals, which are the most valuable statistic in the metric.

It might be hard to accept a system that is so dissimilar from how we normally think about evaluating a game, I also think this formula tells us quite a bit about the importance of efficient play rather than flashy play. It stands to reason that a player who scores a lot while missing a lot of shots can actually hurt a team as much as she helps. And the concept we can derive from this formula is that the most important thing in evaluating a player is how well they manage possessions.

But in this game in particular, I think the formula highlights the role the impact of the point guards on the game in a way that we might not notice just looking at the box score.

Miller Might be the key player for a Mercury playoff run

When looking at the team dynamics ratings for the Mercury, one thing that really stands out is that in the quarters when their synergy score or ball movement was high, they performed much better overall. And part of that can be attributed to the play of Miller.

Miller had a pure point rating (PPR)– a metric that measures a player’s ability to create plays for others – of 3.80. As a comparison, consider that Deanna Nolan’s 3.83 PPR before the Olympic break was third in the league among point guards. Miller’s PPR before the break was 0.16.

In other words, whereas she’s normally a pretty solid point guard who can bring the ball up the court and initiate the offense, last night she functioned as more of a combo point guard, able to score and distribute the ball. And it ended up being the third scorer for the team that Penny Taylor was last year.

In fact, the way Miller plays seems to have an impact on the Mercury’s wins and losses. In wins, Miller has a PPR of 2.65 and in losses she has a PPR of -1.46. That’s essentially the difference between being a creator and a player who’s not doing much to create scoring opportunities for teammates.

But I also looked at assist ratio, the proportion of a player’s possessions that end in an assist. Last night, Miller had an assist ratio of 26.99, which is right around her average for the season. But in wins this season, she has an assist ratio of 30, whereas in losses she has an assist ratio of 23.

To put Miller's assist ratio numbers in perspective using David Sparks’ player styles spectrum, that’s the difference between a scoring guard like Deanna Nolan and a distributor like Ticha Penicheiro. On a team that consists of Taurasi and Pondexter, a distributing point guard that can create scoring opportunities for others is a huge benefit. And I think last night demonstrates how valuable Miller’s play can be to the Mercury.

Transition Points:

It’s important to note that none of these numbers can be used in isolation, but I consider these a starting point for analysis. Clearly, by saying Miller deserves the most credit for last night’s game I’m not saying she deserves the majority of the credit or even that her performance was valuable independent of Taurasi’s or Pondexter’s. Basketball is a team game and lineup combinations as well as defensive match-ups have a huge bearing on player performance. But for a game I couldn’t see, I find this to be more effective than just looking at raw box score data.

"MVDee": The Mercury website has the Taurasi MVP campaign up and running with a few numbers to support their claim. For those of us that enjoy pontificating, what threshold should the Mercury have to break to consider Taurasi the MVP -- .500 or making the Western Conference playoffs?

Relevant Links:

Time running out for title-defense try
http://www.azcentral.com/arizonarepublic/sports/articles/2008/09/02/20080902merc0903.html