Manufacturing and traditional business settings have understood the importance of objective, meaningful metrics for many, many years.
Metrics aren’t necessarily used as quotas or objectives (as Dr. Deming fights against in his 14 points) but as reasonable indicators of “Are we doing a good job?” or “Are we on our way to accomplishing our planned outcomes?” or “Are we properly meeting the needs of our current customers with the plan to serve even more customers down the line?”
For a sport that is so addicted to statistics, it’s surprising that there has been such a large disconnect in baseball between on-field success and proper metrics. Only now are baseball teams more open to considering more meaningful data that are better indicators of top team performance, but it took books written by Bill James, SABR, and Michael Lewis (Moneyball – more here and here) to generate the momentum for the metrics push.
Baseball still largely focuses on metrics that are not directly linked to winning – batting average (hits per time at-bat), runs batted in, home runs – but at least there are some that are identifying the proper direct metrics and letting them be the primary guide for improvements. From Rob Neyer of Baseball Nation:
As I’m sure you know, there’s a strong relationship between runs scored, runs allowed, and winning percentage. We’ve known this for quite some time and … well, I suppose it’s always been obvious that if you score more runs than your opponents you’ll probably win more games than you lose, and that if you score a lot more runs than you allow, you’ll probably win a lot more games than you lose. But Bill James actually quantified this relationship with something he called “the Pythagorean method,” and of course that method’s been refined a bit over the years.
A baseball team’s primary objective: win the baseball game.
– What dictates the difference between winning and losing? Run differential – you score more runs than the other team, you win.
– How do you control run differential? Score more runs, allow fewer runs.
Identifying the direct metrics (such as run differential) is not rocket science. Getting to the next level, the indirect metrics, adds complexity. Let’s take a gander.
– How do you score more runs? Produce outs at a lower rate and get on base at a higher rate, and don’t run into outs when you’re on the bases (caught stealing or baserunning blunders – TOOTBLAN).
This is why on-base percentage is such a critical statistic to Moneyball. It doesn’t matter how you reach base, whether it’s a walk or hit or being hit by a pitch, because the only way to score runs is by getting players on base. (Yes, hitting home runs counts as getting on base.)
– How do you allow fewer runs? Minimize how often opposing players reach base.
– How do you minimize how often opposing players reach base? Employ pitchers who effectively generate more outs and defensive players who generate more outs.
Relative to the current level of analysis on offensive production of baseball players and teams, proper pitching and defensive metrics are tricky and statistical experts still don’t have the metrics perfected. There are a LOT of variables between pitching and defense that require deeper analysis. However, the experts are getting better and the metrics are becoming more reliable.