Team Synergy: The Value of Ball Movement in the WNBA

. Saturday, July 12, 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

If someone forced me to list the things that I value most in the game of basketball, ball movement would probably be at the top of the list.

It is probably why I put so much energy into evaluating point guards – the point guard would seem to be the engine on the floor that gets the ball moving and keeps the team in rhythm.

However, I also think ball movement is actually underrated although people give a lot of lip service to it. It’s scoring, rebounding, and -- to a lesser extent – individual assists that grab all the attention. Part of that might be that ball movement – an indicator of team chemistry – is very difficult to quantify. But I think APBRmetricians have made great strides in helping us with that.

During the Mercury's victory over the Comets on Tuesday, I noticed an assisted field goals (a/fg) stat mentioned at halftime. Essentially it boils down to the percentage of field goals that resulted from an assist.

After the game, I looked through some Mercury boxscores. There seemed to be some relationship between ball movement and Taurasi’s scoring in Mercury victories – in games when the Mercury moved the ball better than opponents and Taurasi scored above her average, the Mercury won all but one game.

This got me curious – what does this assisted field goals statistic really tell us? And how important is ball movement on teams that can effectively run offenses not conducive to assists (e.g. teams where one or two players bears the scoring burden)?

Since that Mercury game on Tuesday, I watched two additional games with an eye on ball movement trends – Dream vs. Lynx and Sparks vs. Monarchs. And I saw the same thing I saw in the Mercury game: if a team doesn’t have good ball movement, they need a star who can carry the team. But even more important is a team’s ability to disrupt their opponents’ ball movement.

In the process of trying to make sense of all this, I came across a simple NBA statistic created by Jeff Fogle that may be even more useful for the WNBA: synergy.

(I wrote this partially to document my own thinking...but it got kinda long. So if you want to cut to the chase, skip to the section titled: "Team Synergy: Ball movement and shooting ability". You've been warned.)

Ball movement theory

This whole endeavor ended up going way beyond the assisted field goals statistic, so I’ll start with a simple theory about ball movement that should seem more like common sense than any kind of real analysis.

It’s simple – even if you have a volume shooter who can score efficiently, if the team can’t move the ball and create a variety of scoring opportunities, the team gets predictable. What ball movement does is force the defense to work hard and eventually leave gaps that the offense can exploit. If the offense becomes predictable, the defense can focus all of their attention on stopping one (or two) players, which makes it difficult to score if a player has an off night or if the opponents have more weapons on offense.

(A great example of offensive ball movement was that New York blowout of the Phoenix Mercury broken down quite well at the X’s and O’s of Basketball blog.)

Conversely, if you can stop an opposing team’s ball movement with strong defense, you limit their scoring opportunities and force them to be dependent on one or two options. It’s not just about playing passing lanes, but also blocking shots and playing good on-ball defense.

A good example of this – if you’ll allow me to make an NBA diversion – is the 2001 Philadelphia 76ers. Some people might use that team as an example when ball movement didn’t matter because one player – Allen Iverson – took the majority of the shots. However, what people forget is that they were also able to shut down their opponents’ ball movement – they were among the best defensive teams in the league, if not the best.

Philadelphia’s center, Dikembe Mutombo, was NBA Defensive Player of the Year and led the league in rebounding and Iverson led the league in steals. So although they were a below average shooting team and had a high turnover percentage, they made up for it on defense.

So really, I think ball movement is about two things – establishing your own and disrupting the opponents’ offense.

Ball movement statistics: What is the value of assists in the WNBA?

An important preliminary question to look at before going any further is the value of assists to winning basketball games.

Go back to that Allen Iverson team again – certainly one could argue that Iverson’s individual field goal percentage on a given night would be worth more than team assists (they were near the bottom of the league that year in assists).

The Wall Street journal published an article earlier this year describing how assists in the NBA are in fact overrated in relation to team wins. The article suggested that actually, the differential between team assist percentage and opponents’ assist percentage is what really mattered. That’s similar to what I found just looking through the Mercury box scores. But to what extent is the differential in the WNBA important compared to the WNBA?

Kevin Pelton suggested in a comment on my rookie rankings post that perhaps, “creating shots for others is slightly more important in the W. If that's true, though, I doubt it's by much.” If that is so, not only are assists more important, but so is ball movement on the whole, no matter what system a team runs.

So there’s a basic way to cover some ground on this without taking a course in advanced statistics – comparing NBA and WNBA assisted field goal percentage. After looking at the numbers a bit, it was clear that assists per field goals attempted (percentage of assisted baskets out of all shots) seemed more descriptive than assists/field goals made (I’ll come back to that later).

Assisted field goal percentage: WNBA vs. NBA

So ast/fga numbers are actually quite close. But the fact that field goal percentages differ by a few percentage points may tell us more.

WNBA (2007 season) Ast/Fg: 25.03% FG%: 41.99%
NBA (2007-2008 season) Ast/FG: 26.68% FG%: 45.72%

Considering that teams shoot 100% on all assisted field goals (we know the ball went in because someone was credited with an assist), the differential in field goal percentage seems to indicate that WNBA teams are more dependent on assisted field goals to play efficiently than NBA teams. (Again, these are surface level boxscore statistics, so if anyone has done something more advanced to demonstrate this, I’d love to see it).

I know some people might use this as a knock on the WNBA, but I don’t see it that way. I actually think the increased dependence on team basketball is better to watch than the one-on-one showcases that we see in the NBA, best represented by Iverson’s 2001 76ers.

Anyway, here’s a look at the top 10 teams in the NBA and WNBA in assisted field goals differential, as suggested by the WSJ article:






































































WNBA 2007

A/FGA Diff. NBA 2007-2008

A/FGA Diff.
PHO 03.37%

PHO 10.49%
SAS 01.93% UTA 06.92%
IND 01.91% BOS 04.89%
CON 01.52% LAL 03.84%
NYL 01.06% SAS 03.57%
DET 00.76% DET 02.90%
CHI 00.18% DAL 02.78%
MIN -00.43% HOU 02.75%
HOU -00.88% NJN 02.18%
SEA -01.34% TOR 01.68%


Quick note: The reason that three teams in the WNBA’s top ten have a negative differential is that there were only 13 teams last year, so it’s natural for 6 or 7 teams – about half -- to have a negative differential. In the NBA, it’s the same – 15 of 30 teams have a negative differential.

So the first thing I notice is that whatever is in the water in Phoenix should be bottled and sold to every other team.

But seriously, this seems to work out somewhat well – conference finalists in both leagues rank in the top six and the championship teams in both leagues rank in the top three.

However, there are a few glaring flaws with this – the Phoenix Suns blew everyone away but lost in the first round of the 16-team playoffs. In the WNBA, three non-playoff teams – Chicago, Minnesota, and Houston -- end up ahead of Seattle, a playoff team. The New Jersey Nets ended the season 10th in the NBA’s weaker Eastern Conference. And the Sacramento Monarchs who came in 3rd in the Western Conference are 11th in the league in assist differential.

So these figures lead me to another point – that assisted field goals alone may not in fact the only number to look at. Even in a one season sample, that seems to make sense – this doesn’t take into account defense, rebounding, or most importantly bad shooting.

If a team moves the ball and nobody around can shoot it, does it make a difference?

Team Synergy: Ball movement and shooting ability

If I pass the ball to you 100 times and you only make 30 shots, while I could make 40 shots on my own, it would be perfectly reasonable to keep shooting the ball myself…for the good of my own stats and the team.

So a team like Seattle that has MVP Lauren Jackson to pass to is probably wise to do so as often as possible, whether there’s good ball movement or not (…and as a side note, both Jackson and Sue Bird were injured for parts of the 2007 season which undoubtedly had an effect on these figures).

Fortunately, Jeff Fogle, who writes the Stat Intelligence blog, has already tried to address this problem and created an NBA stat called “synergy”. He not only looked at offensive synergy, but a team’s ability to shut down another team’s synergy.

Synergy is a simple formula – assisted basket percentage plus field goal percentage. Pretty simple, something you can take right out of a box score, but it does have some descriptive power. Fogle writes:
Teams with low numbers tend to isolate one-on-one. Teams with high number pass the ball around to get open looks. Either approach can work. I’m more biased toward higher numbers…but teams have won championships with lower numbers. This is a “descriptive” stat more than a good/bad stat. Though, there’s actually a good correlation to success when you look at it from the defensive perspective. “Disrupting” synergy is a positive.

So my thinking was that if we looked a team’s offensive and defensive synergy scores and found that differential, we might have something even more useful. Here are the results (WNBA and NBA again):




































































































WNBA 2007


NBA 2007-08
Team A/FGA Diff
SynDiff Team A/FGA Diff SynDiff
PHO 3.37% 6.86% PHO 10.50% 14.90%
DET 0.76% 4.19% UTA 6.92% 10.59%
IND 1.91% 3.57% BOS 4.89% 10.50%
CON 1.53% 2.42% LAL 3.85% 6.96%
SAS 1.93% 2.09% DET 2.90% 4.98%
NYL 1.06% 1.36% DAL 2.79% 4.90%
SEA -1.35% -0.25% SAS 3.57% 4.88%
HOU -0.88% -0.50% HOU 2.76% 4.34%
CHI 0.18% -2.13% ORL 0.73% 3.57%
SAC -1.79% -3.91% TOR 1.69% 2.73%

So the first noticeable thing looking at the synergy differential (SynDiff) scores is that the WNBA finalists end up #s 1 and 2. Additionally, all 8 playoff teams are now in the top 10 and no non-playoff teams are in the NBA’s top ten (although three playoff teams don’t crack the top 20). This is an improvement and it’s fair to say that there’s something to synergy score, for the WNBA, if not the NBA.

The reason it’s not working for the NBA as well as the WNBA is not my concern here, though I have some ideas of why that might be. However, if we could just figure out what’s going on in Sacramento, it seems as though this is a useful metric for the WNBA.

Observations of synergy in games

So from the three games I watched this week here are the synergy scores:

Mercury (71.86%) beat Comets (90.30%)
Dream (60.84%) beat Lynx (63.48%)
Monarchs (74.61%) beat Sparks (74.61%)

As I described my summary of the Mercury Comets game, it seems like there are other factors at work that could be influencing this outcome – in all 3 of the games I watched, the team with a lower synergy differential won.

Part of that might be that the Dream, Monarchs, and Mercury used big runs to pull away in the second half and those runs were the difference in the game. In addition, at least the Dream and Mercury were sparked by one player having huge 4th quarters to win, meaning synergy was not the key element.

Although this seems to have some descriptive value at the team level over the course of a season, it seems to be less effective at the game level. So how do we make sense of that?

Balance is the key

To figure out what was going wrong, I started by trying to figure out why Sacramento ranked so low in 2007. I was stumped for a while – how could a team neither move the ball well nor disrupt the other team’s offense and still make the playoffs?

So I played it out in my head – if you have poor ball movement, can’t defend the other team, and don’t shoot the ball well, how on earth can you win and end up third in the conference? Well if you’re not getting assists and missing shots, the only opportunity to score is on offensive rebounds.

Turns out the Monarchs were the best offensive rebounding team in the league last season…and in the history of the WNBA (by a wide margin). Three Monarchs -- Rebekkah Brunson, Adrian Williams, and Yolanda Griffith – were top 10 in the league in offensive rebounding percentage.

Then I thought about the fact that I had not factored in turnovers yet. Obviously, turnovers are harmful and if you have good ball movement but toss the ball away, you can’t possibly reap the benefits. In the Monarchs-Sparks game from Thursday night, turnovers figured prominently – the Sparks had 26 turnovers to the Monarchs’ 14.

Those of you more familiar with basketball statistics might recognize this line of thinking – it’s very similar to Dean Oliver’s Four Factors, which ranks shooting percentage, turnovers per possession, offensive rebounding percentage, and free throws as the four most important aspects of basketball.

Could synergy differential be used along with some of aspect of the Four Factors to create a meaningful score for the WNBA? I think so…and I’ll post that thinking tomorrow.

In the meantime, I think synergy does strengthen an argument consistent with common sense: in general, the most successful teams have better synergy than their opponents.

What I found is pretty interesting – the ratio of assists to field goals made, which is roughly assisted field goal percentage is huge. But star power is bigger. If you have both ball movement and a top 10 scorer, there’s almost no way for a team to stop you.

Relevant Links:

NBA Game Theory - -a post about A/FGA ratio in the NBA from D Sparks at the Arbitrarian blog.
http://arbitrarian.wordpress.com/2008/01/28/nba-game-theory/

An A/FG ratio thread at APBRmetrics
http://sonicscentral.com/apbrmetrics/viewtopic.php?t=1670&highlight=assisted+field+goals

Transition Points:

An interesting post from Eric Musselman’s blog that I enjoyed:
Can a good coach help an average (or below average) team to be more competitive and win a few more games over the course of a season? Of course.
Effective coaches who know not only the strategic and technical aspects of the game (i.e., the X's and O's), but also the "soft" side of the game (e.g., how to motivate, communicate, etc.), are likely to get the most out of their players.