To begin, lets look at the world around you. While it is big and full of things that seem incomprehensible, modern statistical science has shown that we can measure almost anything. All areas of life have different variables that contribute to our overall understanding of the subject matter. We have methods as simple as measuring amounts of liquid in a container and judging the ground speed of a car to as complicated as figuring out intracontinental weather pattern (stupid German woman) and measuring brain states. We know these systems work simply because they are more efficient than human assumption. Our minds tend to be biased in one direction or another when we make decisions without solid evidence,.
Lets move on to baseball. There have always been the conventional batting average and counting stats, such as Home Runs or RBI's, which we use to judge players. If Player A hits 40 Home Runs and had 140 RBI's, while Player B hits 35 Home Runs with 100 RBI's, Player A is clearly a better player. Or are they?
The statistical revolution in baseball sought to overcome such simplistic notions of performance in what really is a complex game. What if, in the previous example, Player B had walked 200 times while Player A walked 50 times, then a different issue has emerged. Player B may have been unfairly judged due to the lack of opportunities. I know that a traditional baseball mind would tell you the same thing, but statistics give you a chance to put it into quantifiable terms.
The main advantages of using stats are twofold: they provide rates and context.
Rates- when a box score tells you how many hits a player had, it doesn't actually tell you the impact they had on the game. Someone who saw the game without a box score could say "Oh yeah, X's single in the eighth inning really turned the game" but this is just their opinion. These days, we have established rates for WHEN and WHERE the hit, out, or error that really mattered was. Win probability can be calculated during the game because there have been literally thousands, possible millions of baseball games played to this point in history, and we can examine them to show us the most efficient way to win. Rates also give us a better idea of how well a player is performing.
On the offensive side, a player's OPS tells us more about how good a player they are than a raw total like home runs. Yes, chicks dig the long ball, but from examining patterns during gameplay, a hitter whose only goal is to hit home runs can often harm his team. This is why perennially, Ryan Howard's 40-50 Home Runs don't help his team as much as Dustin Pedroia's doubles and high batting average.
Defensively, it has long been a problem with the Error statistic. This has tied into the Baseball Traditionalist's opinion of players for as long as I can tell. In their minds a player who runs the extra 10 yards and then just misses a fly ball is apparently not as good a fielder as the player who lets it drop in the gap and makes no effort to deprive the batter of a hit. This year especially, defensive stats have caught up, telling you a player's true value in terms of range, strength of arm, and general ability.
Context- this is where the thinking element comes in. Applying baseball stats to in-game performance obviously is an indirect manner of operation, because no manager can go up to a player and say "hey we need to improve your On-Base Percentage." However, we can use it to evaluate a player's true talent and value to the team. This year, everyone can tell that Matt Kemp is having a good year from his raw statistics. However, no one will tell you that he is among the top-5 "most irreplaceable players in baseball." His value is currently higher than the vaunted trio of Hanley Ramirez ("best young player in baseball), Joe Mauer (still hitting to the tune of .370), and Prince Fielder (already a yearly MVP candidate at age 25).
Context also tells you if a player is getting lucky or not. Into the month of May, Andrew's one true love David Wright had a batting average near .360. What he didn't know was that Wright's BABIP (Batting Average on Balls In Play) was over .400 at that point. Now, i know that the BABIP stat is a poor indicator of performance when taken in a vacuum, but it can be used to see if a player's performance is so outlandish, so ridiculously out of line with the mean that they truly are getting lucky. We all know that Wright is a good player, and his BABIP should be higher than most players because he has good control of where he wants to hit the ball, but this stat is outstandingly high. The same goes for Edwin Jackson, who is in the middle of a breakout year. Yes, he's talented, and yes he may have improved his pitching style, but the numbers (Left On Base %, BABIP against) show that he is performing well in a lucky manner, rather than a truly dominant manner. And luck is less a sign of talent than domination is
So where did I lead you on this rant? I'm not sure i can say really, but luckily statheads have made life simple. By taking all these seemingly advanced measures and condensing them, they have created more simple evaluations of players: Runs Created/Saved and Wins Created. The point of baseball is to score more runs than you allow the other team to score in order to win the game, and so teams try to find players who fit this philosophy. Statisticians have showed you, Mr. Future GM, a way to do that
No comments:
Post a Comment