Felix Hernandez won the AL Cy Young Award today despite have -3 wins for a terrible Mariners team. Stat geeks crowed with pride about how the most useless of stats, wins, was finally overlooked in favor of their geek ball stats, such as xwins. They claimed that such a thing would never have happened in the 80's as the voters focused on wins and meaningful games pitched.
Before I trash sabermetrics in general, let me take on the idea of King Felix's win. He absolutely deserved it and everyone, everyone, knows it. The infuriating part is that you don't need sabermetrics to know it. All of the conventional stats show he was dominant and his run support explained the wins. And, should that fail, you could trust your own lying eyes, which made it plain as day that Felix dominated every time out while CC Sabathia was lucky to get his fat ass through 6 innings most nights and that David Price was extremely hittable. If the writers back in the 80's wouldn't have noticed that then that's a condemnation of them and their stupidity and laziness. Don't try and use the shortcoming of your journalistic forefathers to justify your love of useless statistics.
This is the basic problem with these geeks though, they need a number to tell them what anybody who knows the sport can see with their own eyes. Sometimes the numbers can also be used to try and convince themselves that what they see isn't what they see. Take this years Yankees. Once again, the stat heads like to use tons of pie charts and Venn diagrams to prove that Jeter isn't a very good defensive shortstop. Anybody who saw him moving around like a chilled corpse could've told you that. There are a number of stats that show how badly Jorge Posada is defensively and is now falling apart offensively as well. We all know that anyway since when Francisco Cervelli was behind the plate, teams magically didn't steal every base in sight and AJ Burnett was even able to locate the plate from time to time. What is A-Rod's VORP (Value Over Replacement Player)? Since the replacement player is Roberto Pena, I'm gonna say "a shitload". Want examples of advanced stats deluding the mind? Look no further than Nick Johnson and Javy Vassquez. Cashman was enamored with the idea of putting Johnsons ridiculously high OBP. Which would've been fine except everybody on earth knew that his one year of relative health was a fluke and that he'd spend 100 games on the DL. And he did. Javy Vasquez had himself a comeback of sorts by pitching in the pitcher friendly NL in the super crappy NL East. All the numbers looked like he could be a decent #3 or 4 and at least chew up innings. Luckily, the Yanks already had Javy for a few years and saw first hand that he was a mental midget and couldn't handle the pressure of playing in NY. Surely this first hand experience would deter them from such a stupid move. Nope, they loved the numbers and signed him anyway and history repeated itself. He couldn't even get any playoff innings. I guess there isn't a stat yet for being a mental case.
Basically, only two kinds of people need sabermetrics. The first one is people who don't really follow the sport, haven't played it, but are for some reason (fantasy league, journalist) need to find a way to develop an opinion. Fine, let the babies have their bottle. The second type of person is the worst though. These are the people, mostly in the media, who have been watching the sport for years, know the history, played it at some point in their life and are very familiar with every aspect of the game. These people use sabermetrics for one of two reasons. The first reason is that they are out of things to write/talk about and they use these goofy stats to try and stay relevant. Bush league. The second reason is that they're cowards. For whatever reason, they are so insecure about developing, expressing and defending an opinion about a player/team based solely on what their very capable eyes and traditional stats tell them, that have to try and invent all these convoluted stats to lend an aura of legitimacy to their opinions. They can bank on the fact that if they are in some kind of debate and start referencing VORP, PECOTA, PW/BFW or UZR, that whoever they are talking to will probably concede the point since the don't have their formula sheet and abacus handy.
The whole thing is a cop out. I know Sergio Mitre sucks because he sucks every time I turn the game on. I know A-Rod is better than Roberto Pena because he's clearly better every time I turn the game on. They only way these stats might be helpful is if you're trying to differentiate between two guys that suck. Comparing John Buck with Ronny Paulino or something like that. Is all this math really worth it for that? if that decision is vital to a team, they aren't making the playoffs anyway. It's time for all you stat geeks to grow up, grow a pair and stop trying to take the easy way out when it comes to forming an opinion. Do the hard work yourself and watch the games or don't get involved in the process at all. If people like you get any more sway, we won't even need to play the games anymore, we can just simulate them through a supercomputer in Bristol. That'll be exciting.
Posts
I'd like to pose you a question. If baseball statistics didn't exist--the only thing that anyone kept track of was win-loss record--and you had to tell me the best player on each team solely based on what you saw with your own eyes, how many do you think you would get right? Not many, I'd wager. The difference between hitting .257 and .285 is one hit a week. Between .285 and .314... another hit a week. That's two hits a week to differentiate you from a player with a mediocre average from a player with one of the best averages in baseball. Now, batting average isn't a great stat, but I think you'd have to agree that unless you watch every single game a team plays, you're probably not going to pick up on that difference, especially since the hits don't all come uniformly. What if he slumps hard in the last month? If at the end of the season someone told you he was still hitting .314 you'd be rather surprised, wouldn't you? And I'd give you TREMENDOUS odds that in this hypothetical stat-less world you would not even come close to guessing the best player on most teams.
ReplyDeleteMy point there is that baseball, uniquely among sports, really thrives on statistics. It's a slow game with a high chance of failure played over a long season. We rely on statistics to tell us who's good, when we should get excited when someone's batting and when we should cringe, to feel confident when our ace is on the mound and upset when he's not. There are obviously exceptions--Pedro Martinez and Barry Bonds, in their primes, made it obvious in every single game they played how good they were. And anyone who watches NL baseball pretty quickly learns to be disappointed when the pitcher is at bat with two outs, or the garbage-time pitcher is on the mound. But there's a large swathe of territory between "ridiculously good" and "ridiculously bad," and we need statistics to differentiate between them.
Statistics are important. I think you'd agree with me there. So why do you hate sabermetrics? Sabermetrics is nothing but a movement to find better, more telling statistics than the ones people have been using since baseball's inception. The only fight between traditional and advanced baseball stats should be one of actual efficacy (which the advanced stats win hands-down)--it's not some sort of ideological battle. I can't stress this enough: [b]the sabermetrics geeks and the traditional guys want the same thing[/b].
Do you know how we know that back in the day, they would not have given the Cy Young to Felix solely based on how many wins he had? Because they did that multiple times. Hell, you don't have to go back to the '80s, in '05 Cliff Lee was a thoroughly mediocre pitcher who came in second in Cy Young consideration solely because he had 19 wins. We have tons of examples of GMs making poor choices simply because they didn't have access to (or didn't agree with the interpretation of) advanced stats. We also have tons of examples of great players being ignored for awards that they deserved because they didn't have good enough traditional stats (Bret Saberhagen, anyone?) So there's no reason to be upset about the geeks "taking over" with their fancy stats or whatever. They're on your side.
I never said I wanted a statless world. I agree stats are important, I just think all these sabermetric stats are overly convoluted and not necessary. The fact that Felix may not have won the Cy Young in the 80's (he probably wouldn't have) or that Cliff Lee didn't deserve his award in '05 just means that the sports writers voting were ignoring the traditional stat categories, not that the current stat set was lacking. It's an indictment of the voters more than any sign that new quantitative measures were needed. I do also think that the average fan can fairly judge baseball players since the action is slow enough for the eye to catch and make the appropriate comparisons.
ReplyDeleteThe regular baseball stats , you know,the tried and true ones we all fell in love with as kids? Those stats by and large haven't changed in their meaning in over 100 years ! Different eras can sku them a little I know , but that's the beauty of baseball.
ReplyDeleteWhen I look back in history and see that Babe Ruth went 94-46 off the mound over 10 seasons with a 2.28 ERA. I know exactly what that means. The dude was a badass.
More often than not when I see sabermetrics , I see someone who has already lost the argument once but just can't let it go. Wether it's for a player or against him.
If you don't know baseball well enough to understand how huge the difference between .285 and .314 is then maybe you aren't the best person to be talking stats anyway.
An extra hit a week??? Over a season that's the difference between "good year kid" and "All-Star" Over a career it's the difference between "good hitter" and "Hall of Famer"
In a game where failing "less" than 70% of the time is considered excellent one extra hit a week is gigantic! You don't need sabermetrics to tell you that.