By Jon Nichols

When Moneyball came out a few years ago, one thing that became evident was the “war” between baseball stat-heads and traditional scouts.  Which was more accurate: objectively using numbers but ignoring a feel for the game, or using opinions based on the naked eye and discouraging the use of stats? As it turns out, the correct answer is somewhere in the middle, especially for basketball (as I’m sure most general managers have realized).  Numbers are great because they give you an objective opinion, but they also only paint a small picture and rarely can explain all of the events leading up to a certain outcome (for example, think of a solid screen that leads to an easy layup).  Along the same lines, numbers only indicate what has happened in the past and cannot predict the future. 

Basketball statistics certainly have come a long way.  Teams no longer only have simple box score totals such as points and rebounds.  Instead, they can use pace-adjusted statistics, plus-minus data, assist rates, etc.  In addition, they have complete player metrics such as Composite Score, PER, Wins Produced, Offensive and Defensive Ratings, WARP, and a multitude of others.  Even defensive stats, considered the most unreliable data out there, have come a long way (check out the dMULT and dQUAL stats at the new player pages at Basketball Prospectus for an example of this). 

So where do teams go from here?  Obviously the current metrics can be improved, but in my opinion, that’s not where new ground will be broken.  You can only do so many things based on the available information out there, and there are some very smart numbers people doing great things with that information.  If anything is left to be discovered, those guys will figure it out. 

Instead, new discoveries will be made when it comes to connecting the two sides of the “war.”  How can the things a scout sees be translated into a form that can be recorded and used to objectively compare multiple players?  Stats will one day be able to measure the contributions of the players that do the little things that help teams win games.

Similarly, stats will be able to measure intricate levels of detail used to make coaching decisions.  Perhaps a team is considering using a Princeton offense that requires their big men to make accurate passes to players cutting from the perimeter.  Some day advanced stats will exist that can measure a player’s accuracy on those passes, helping coaches understand how best to utilize the athletes on their roster.  Or maybe there will be numbers that reflect the quality of screens a certain player sets.  This will help coaches and general managers alike when they run their teams. 

Finally, stats will improve in the area of measuring interactions between different styles of players.  It should be pretty obvious to most basketball fans by now that you can’t just try to accumulate as much talent as possible without any regard to how players fit alongside each other.  A team full of slashers can’t operate without players drawing the defense outside, and vice versa.  Teams will be able to figure out what skills they’re lacking, and advanced stats will be able to tell them where to look for those skills.

Copyright © 2009 Basketball-Statistics.com