free NBA basketball picks
Home | NBA PLAYOFFS | Regular Season | Commentary | Contact Us

How Shot Selection Correlates to Wins

At a time when the US Olympic Men's Basketball team is under attack for its poor outside shooting, this is a good moment to investigate what wins in the NBA in terms of shot selection by distance.

We just published a quick article that looked at some breakouts of NBA shooting performance by distance numbers, and as we refine these numbers more they will play a larger role no doubt in our coverage going forward. What matters now is to understand better how these stats relate to a team's won-lost record during the year.

Here is how we split up the shots:

  • Non-Measured - a certain percentage (around 8%) of shots did not get measured.

  • Close Shots - the smorgasboard of close-in action, including dunks, layups, tips, etc.

  • 6 to 11 feet - what we are referring to as Short Two's

  • 12 to 17 feet - perhaps the true mid-range shot distance

  • 18 to the three point line - aka Long Two's

  • 3 Point Shots - the shots from behind the line

  • Now with this data aligned in our full databases we can extract all kinds of situational numbers (such as long two's with the shot clock running down, or close shots immediately following an offensive rebound), but for this first study we'll keep it simple with just the full regular season tallies.

    Simple Correlations
    To begin with let's examine how factors correlate with the regular season wins, focusing on eFG% ('effective field goal percentage' which adjusts for three-points shots made, and in the absence of three's is identical to the normal FG%) and what percent of a team's shots were taken or allowed at each distance.

    effective Field Goal Percentage (eFG%)
    Shot Distance
    95% CI
    -.02 to +.64
    Close Shot
    .34 to .81
    6 to 11 ft
    -.26 to +.46
    12 to 17 ft
    -.07 to +.61
    18 to line
    -.15 to +.55
    3-Point Shots
    .17 to .74

    With just 29 observations (one season) to go by, we are facing some huge standard errors, which makes the 95% CI stats a little scary in terms of coming to strong conclusions. Nevertheless, the key thing on offense is your Close Shot FG%, which brings a .58 correlation to the party. On defense however, it's your ability to limit your opponent's three-point eFG% that counts the most, at a -.49 correlation.

    Not surprisingly then, the two most significant net eFG% distances are Close Shots and Three-Point Shots with .63 and .51 correlations. Of course, the 95% CI's suggest that our presumed order of importance could get completely shuffled around if another sample is viewed. Still, "layups and threes" are the order (and winning business) of the day in the NBA, at least for the 2003-04 season.

    Let's turn then to the respective ratios of how often a team is shooting (or allowing) a shot from the different distance grades.

    percentage of total shots at each distance
    Shot Distance
    95% CI
    -.41 to +.32
    Close Shot
    -.28 to +.45
    6 to 11 ft
    -.33 to +.40
    12 to 17 ft
    -.48 to +.24
    18 to line
    -.43 to +.30
    3-Point Shots
    -.31 to +.42

    The correlation on strict attempts by distance relative to the total attempts, proves to be small, regardless of distance. Restricting three-point attempts by your opponent has the strongest value at -.31, but the nets are all very marginal, which suggests that there are many ways to win in the NBA in terms of your balance between distances, what matters is the percentage of shots you make and allow (which would seem kind of obvious...)

    All right though, it's the moment you've all been waiting for -- REGRESSIONS!

    Regression on eFG by Distance to Wins
    When you run a regression using 29 observations with twelve stats per team, you're going to get some fairly suspect numbers...albeit with an R^2 of .73 (and an adjusted R^2 of .53)

     Intercept 30.0   83.8  
     Non-Measured Shots 43.8   59.6  
     Close Shots 274.3   77.5  
     6 to 11 Feet Shots -62.7   73.4  
     12 to 17 Feet Shots 145.4   100.6  
     18 to line Shots 61.6   68.3  
     Three-Point Shots -62.3   61.8  
     Non-Measured Shots (Defense) 19.5   77.9  
     Close Shots (Defense) -115.5   65.4  
     6 to 11 Feet (Defense) -0.6   94.9  
     12 to 17 Feet (Defense) -1.5   101.3  
     18 to line (Defense) -96.9   128.9  
     Three-Point Shots (Defense) -162.3   56.4  

    From this regression we would state that the most important performance areas are:

  • Close Shot eFG% on Offense

  • Three-Point Shot eFG% on Defense

  • 12 to 17 Foot eFG% on Offense

  • Close Shot eFG% on Defense
  • Believe it if you will, but of course the standard errors are so vast that for the most part it could easily go topsy-turvy in viewing a different season. The strongest indicator by far though is the close shot eFG% on offense, which suggests that yes indeed, having an inside presence or finding ways to get easy buckets (whether in transition, off dazzling passes, or through having perimeter players that can take it 'to the hole') is a critical component of a successful team.

    Of course, the Pistons only ranked #21 in the league in close shot FG% during the regular season, which perhaps suggests that what it takes to win in the regular season and what is optimal for peak playoff performance could differ. Detroit was solid in defense from all zones, and basically top of the charts when it came to defending the three-point shot.

    In our analysis of Using the Shot Clock Wisely we had concluded:

      Look for the easy/quick basket but don't get tough defense especially in the early parts of the shot clock...and you may be more successful than you are now!
    For our wrapup here we will state:
      Make a high percentage of your close shots...and limit your opponent's three point shooting percentage...and you may be on your way to a great season!

    Rate this Feature
    Poor   Fair   Good   Excellent

    Enter your comments in the box

    Email (optional)

    We want your feedback! Tell us your thoughts

    Copyright ?2003 by, All Rights Reserved