No sport inundates fans with more statistics that baseball. Even before the stats revolution of the last decade or so we had way more statistics than any other sport. Batting average, home run totals, runs scored, runs batted in stolen bases…wins, losses, saves, earned run averages, strikeout totals, innings pitched…you get the idea. And those are all the “old stats!”
Down the line we starting paying attention to on base percentage, slugging percentage, hitting (and pitching) with runners in scoring position. Then came the numbers for how hitters and pitchers did on the various counts during an at bat.
Defensively we noted assists, put outs, errors, fielding percentage.
All of these stats were simple facts. They didn’t involve anything unique other than putting percentages on some areas based on simple formulas.
Now things are way different. Statistics based on sometime convoluted formulas have been devised that to some eyes only serve to confuse and make the game far more complicated to follow that it need be.
Adjusted earned run average for example is a stat that is to take into account far more than just a number based on earned runs. A whole lot goes into it than just the simple facts. Or how about WAR? That stands for wins above (or against) replacement. This is actually more than one interpretation of WAR. Simply it means that players value is compared to others who play the same position. How many more wins he would be responsible for versus the average replacement.
This is where to my veteran eyes things start getting ridiculous. There may be a formula for it based on past numbers, but there is absolutely no way it can really mean much unless you just want a way to rank players based on it. Why? The games are played by human beings. The variables in any random game don’t always fit formulas. You can’t control the outcome of any game played by humans with a computer print out. All you can do is try to increase your chances of success, based on PAST outcomes. That is where the analytics for defensive alignments come in and THOSE are very important in the modern game.
This doesn’t involve using WAR or any other stat directly, but if you have any old copies of The Bill James Handbook look in the back for the projections. These predict what a player will do the next season based on his past. Bill used all sorts of formulas to come up with the numbers. Anyone who wagered on most of them would have been in the poor house by the next fall. Injuries cannot be predicted, of course, but those players who played a full season there were some huge misses. Others not so much, but all really just educated guesses even if statistical formulas were part of the process. Carlos Beltran according to James was going to hit .280 with 25HR and 96RBIs. What he did in 2007 was hit .275 (close enough) but with 41 HR and 116RBIs. While I am in the “B’s” how about Craig Biggio? James and company weren’t far off. They said .256 19-63 for the 40 year old. He actually hit .246 21-63.
Where they projections were weakest were usually with older players who had reasonably good seasons the year before. The physical decline doesn’t always show up in the numbers in which the projections are largely based, but are seen by scouts, coaches and players.
In Biggio’s case he had a poor (for him) year the season before. His average had started to decline as he had to “cheat” more at the plate….meaning he had to start his swing earlier to be able to catch up to fast balls. And he was now 40 years old. Beltran had hit only 16 home runs the previous season, but was healthy and the projection called for an improvement to 25. Instead he hit 41.
My point is less to totally discount projections but to emphasize what players actually do on the field is most important and being too involved in numbers is missing the point of why the game is played.
The other thing that statistics don’t routinely show is WHEN a hitter, pitcher or fielder does something is far more important to winning games that the simple fact they did it itself.
For instance a team may be a poor fielding club. They make more errors than anyone else. They may have defenders with poor range. But if they only make most of those errors at times when they don’t result in hurting the team they don’t matter as much. If the clubs’s pitching staff can pitch around errors there is no harm. If defenders with poor range rarely have to have any because most balls are hit in the right spot…that lack of range won’t show.
The same can apply to teams that strike out excessively. If they are doing it in circumstances where hitting the ball into an out would have no advantage they don’t matter.
If a pitcher gives up line drive after line drive but all of them are caught no one knows from stats that he was really hammered. (No doubt someone keeps records of such, but few will make a final decision on a player from a piece of paper. They have to see.) He got outs that is what we know.
Stats, no matter how detailed or how creative simply cannot show the whole story. They can be fun for some folks to play around with and try to prove this guy is better than that guy using the numbers. But let’s actually see both on the field and playing the game. That is always the best test. The best guy is actually going to have the hits that wins games and not put up all his numbers in a 14-2 laugher. The best reliever is the man who is nearly perfect nine of ten times even if that one time he gives up six runs to blow his earned run average sky-high. The best hitters are the ones that can be counted on to be .300 or better in clutch or run driving in positions…even if his overall average is barely .250.
The player who goes 1-5 is only a .200 hitter. But if that one hit drove in runs he is more important than the 3-5 hitter who drives in or scores nothing.
The beauty of baseball is that over time good hitters will wind up with good averages. The same for home run hitters and good pitchers. Poor fielding teams will usually be exposed over a long season. But WHEN those players do what they do to get there is what makes winning or losing teams… all statistics (old or new) be damned.