Hoo boy, sabermetrics.
Recently, I’ve found myself in Twitter wars over my apparent inability to fathom and accept this new form of statistics that has been integrated into the game of baseball and appears to be out to completely replace those that we have used for more than a century. Not only am I constantly bombarded with claims that Michael Young has the worst rWAR in the history of the Rangers and that Ron Washington HAS to know that and shut MY down, but now I am apparently a moron and bad baseball fan for not accepting Keith Law’s claim that RBI is the most worthless stat in all of sports. Yeah. I’ll get to that.
The main point is that there appears to be a developing rift among baseball observers between those who have completely adopted the way of thinking of Bill James, Billy Beane and others who believe these new “sabers” are the only CORRECT way of thinking and those of us who believe the standard system of batting average, home runs, RBI and runs scored still works.
Let me say this. I have no problem with thinking outside the box and considering new ideas. Innovation is key to advancing as a society. And to their credit, there are some things from this new wave of statisticians that can make sense.
For instance, there’s James’ claim that sacrifice bunting is counterproductive because you should never give up your outs. I’ve seen too many times with my own eyes teams sacrificing a runner to second base and falling to bring him home, so that’s valid. It also fits into the style of play Earl Weaver’s Baltimore Orioles played to great success in the 60s-70s.
But remember, this is also the guy who claimed a true “closer” in baseball is not needed and that a committee bullpen will work fine. The Red Sox tried that in the early 2000s and failed miserably, not figuring just how much coming in late and getting the last three outs requires a certain human mindset that not every pitcher has.
Because I’ve discovered I can’t clear make my argument within Twitter’s 140 character-at-a-time system, I am composing this to make my case of sabers vs. stats. Of course, that phrase in of itself reflects the complicated issue I’m dealing with, as I must sort through the major “new stats/sabers” out there and compare them to the old guard. From the multiple claims among different people, it does appear that I still struggle to get exactly what is a “sabermetric” and what is just a new statistic. There appear to actually be new numbers that some would put as sabermetrics but I wouldn’t – at least not among those I shake my head at.
For example, there’s “WHIP,” or “Walks and Hits per Innings Pitched.” This is something that makes sense to me because it actually represents an on-field result to show how effective a pitcher can be. If there’s one stat that I do feel is becoming outdated, it’s ERA. But that’s because of the fundamental change in the way the game is played now, since pitchers rarely go a full nine innings anymore. If I were a broadcaster, I would probably mention something like straight averages of runs allowed per game and innings per game. So maybe I’m a little guilty of what I’m accusing others of doing in that sense. But again, that’s because of how the use of pitchers has so drastically changed as opposed to other players.
That brings me to the other side of the field and one of the new stats out there – OPS, or “On-base Plus Slugging.” It’s obviously easy to see how this is calculated: Add up a player’s on-base average and slugging percentage. But what does the resulting number mean? I get what they were TRYING for: It’s supposed to show a player’s combined skill at both getting on base and getting extra base hits. But the percentage number itself has nothing specific to represent. If someone has an OPS of .724, does that mean he has a 72 percent chance of getting on with an extra base hit? It doesn’t seem like an accurate way of determining that. And it’s entirely possible that the combined numbers can equal over 1.000, which is makes the number even more nonsensical because it is not possible for someone to succeed at over 100 percent. (This also reflects why I’ve always been skeptical of slugging percentage also, because its calculation method – total bases divided by at-bats – can have the same result).
And then there’s the big one: WAR, or supposedly “Wins Above Replacement.” I may be wrong, but it seems like this is meant to be baseball’s version of the quarterback rating (and if I was a football fan, I’d likely have a beef with that as well). More specifically, it seems to be some way of calculating a player’s total value through hitting, baserunning and fielding. There also appear to be different versions of it, from fWAR to rWAR, and it uses such a complicated method of calculating so many factors that I still haven’t found a source that completely shows exactly how it’s added up. And even if I ultimately do and figure it out, there’s still that big question: What does the number itself represent besides a vague answer of “a player’s worth?” It’s a rating and not a statistic, and the biggest problem with a rating is that it can be subjective. The mere fact that there are multiple forms of WAR currently fighting for supremacy prove this, because some use factors (like different baserunning numbers) that others don’t. It’s a coefficient, and a coefficient’s worth is much more open to debate (college basketball’s RPI and football’s BCS system come to mind) than a stat that is simply tallied like runs and goals.
There are many more “sabers” out there that I can’t write about without making this article long enough for Random House to consider publication, but those appear to be among the biggest. And honestly, I can’t give a reason why these numbers shouldn’t exist. Everyone’s going to look for something new to try and find an edge in calculating ability, and baseball has always created more and more stats that can border on the ridiculous (watch the running gag from the announcer in “Little Big League” for a good example) but still have worth.
If I have come off as snarky toward sabermetrics on Twitter and other forums (and please note that sarcasm often comes second nature to me in those venues), it’s not really because I have disdain for their existence. Moreover, it’s the attitude many of their proponents exude. The two main types of people whose opinions on the game I respect the most are those that played at some high competitive level and those who have watched it much much longer than me.
But without singling anyone out – okay, maybe ONE person – it does seem like many fans of sabermetrics, especially in the media, didn’t play the game or see Carlton Fisk’s 1975 home run when it happened. Yet they are smarmy and elitist in claiming that their way needs to completely replace the old way, especially when some so-called “experts” claim THEIR numbers that show potential should just completely replace OUR numbers that mostly represent results. When Keith Law goes onto his ESPN radio show and gloat about how he’s out to eliminate the RBI as a stat altogether, he’s not going to endear himself to me.
More than anything, though, is just how much sabermetrics remove the human, real world situational aspect. People like Law say that when a player gets a hit with two men on, the two RBI are worthless in gauging value because the chances were just as likely he’d get that hit with no one on. In saying that, Law assumes players have the exact same chance of doing anything regardless of the situation, and that’s not the case. Players need to change their approach and focus on what exactly they need to do depending on the situation. If Josh Hamilton comes to the plate with first and third and one out, he has to think “drive the ball to the outfield to bring in at least one run, keep it off the ground to avoid the double play.” If there’s one out and a runner on third only, he can just put the ball in play nearly anywhere. If there are two out, though he has to go all out for a hit; sure, he was trying for a hit in the other scenarios, but here he knows it’s absolutely needed. Good players know what their options are to succeed in the situation they’re in, and those that do the job in those areas have it shown in their results like RBI. It’s very possible that a player can easily hit doubles and home runs with no one on base because they’re more relaxed, but they struggle to do so with runners on base because they overthink and try to do too much. “Potential” stats often don’t factor these human, mental elements. And while “result” stats don’t really do this either, they at least show how often said player actually got the job done. And yes, you can make the same argument against batting average; that’s why I agree that BA can be misleading and say runs scored and RBI are better gauges.
And yes, there are other factors like how few opportunities a good RBI man will get on a poor team like Seattle. You always have to factor in things like games played and how good the players around him are. And really, that’s an underlying example of how individual stats can’t always determine team performance and vice versa. It’s why even Michael Young can have such a poor OPS and WAR (or few runs scored and RBI, in my case) and the Rangers can still lead the American League.
Heck, now that I think about it: I’ve probably had my own sabermetric for years, if that’s what you would call it! For at least a decade now, I’ve combined a players runs scored and RBI to measure worth by see in how many runs he’s directly produced (you’d have to subtract his home run number to be totally accurate). Basically, a good player would have at least as many combined runs as games played or hits; a great player would have a season total of around 200. So yeah, I’m probably just as guilty, you can judge me, but again, I’ve simply chosen to calculate on numbers that represent results over potential.
Ultimately, I wouldn’t say I have a problem with the creation and use of these new sabermetrics. It’s more the apparent intent of their biggest supporters to make them the new gospel of baseball, as if Eric Nadel should be stating a player’s rWAR rather than his home runs when he steps to the plate. Most of those that worship them take them to be the absolute ultimate way to judge a player, but the truth is: There is no ultimate way and never will be. And there’s no guarantee that, over the long haul, the new way will be significantly better than the old method.
Billy Beane and Co. with the A’s supposedly use these hard new numbers to draft players, and it worked in the early 2000s to get a small-market team to the playoffs – but little beyond that. Theo Epstien may have used similar tactics building the Red Sox, but I doubt the Moneyball method showed value in a guy like Dave Roberts, who’s stolen base the Sox wouldn’t have won in 2004 without. And I’m sure Jon Daniels takes these things into account as well. I’m also willing to bet he listens to guys like Nolan Ryan, Ron Washington and his 30 scouts, and when they say “this kid can play,” he takes their judgement over raw numbers.
That’s why, even as I still cling to the old-school method of believing Josh Hamilton’s RBI and Elvis Andrus’ steals as a measure of success, I remember that there will always be factors you simply can’t quantify and that numbers themselves can never perfectly show success in real game scenarios. The eyeball test will always come first and foremost, which is why I didn’t need numbers to believe Rusty Greer could deliver in the bottom of the ninth; I saw him do it enough times.
And that’s why no statistic or sabermetric can replace the coefficient of Mister Randy Galloway himself: WtDG.
Watch the Damn Games.