I just re-read Moneyball this weekend, ten years after I first read it-- it's amazing to look at the people that the Oakland A's were considering (and coveting) as minor league prospects 10 years ago, and to see just how many of these players have panned out as MLB stars-- Youkilis and Swisher for example-- and others became serviceable big leaguers (Joe Blanton, Teahen)-- so my question is: Has the "Moneyball" (or call it the Sabermetrics/Bill James) approach to evaluating and drafting players been validated or debunked?
I know that Billy Beane follower, Theo Epstein, managed to bring two World Series Champs to Boston after 86 years of waiting (but Beane and Moneyballers always contended that post-season success is not something he/they can guarantee-- it's too much about luck, and Boston never ceded the monetary advantage that they had over everyone but the Yankees, so it's possible that Epstein's success is not necessarily any validation of Beane's approach), but J.P. Ricciardi hasn't done anything in Toronto worth noting, and I don't see the Mets doing much under Beane protege, DePodesta, and even Oakland has fallen off from their late '90's- early 2000's run... so is there enough evidence to support the validity of the Sabermetric approach to building a baseball team? Have all teams adopted the approach, to some degree (in other words, victory of the theory, without acknowledgement), or are teams building winners in nearly complete ignorance of Sabermetric principles?
I have not heard anyone say that the Cardinals (who also have 2 championships in the last 10 years, along with another World Series appearance) are a Sabermetric-oriented team-- and I certainly doubt that Tony LaRussa would acknowledge managing that way-- he's about as "old school" as they get, I'd imagine... and the Reds have started to get good finally (after 10 years in the wilderness), by essentially re-applying the Cardinals' master plan (the Reds have a former Cardinal co-owner and former Cardinal GM, who built their championship teams), and the Cardinals/Reds MO seems to be an emphasis on building up scouting strength (of the old-school variety) rather than statistical analysis.
So, to be more specific: Has it been demonstrably proven that On-Base Percentage is the single most important offensive stat in baseball? And if OBP and Slugging Percentage are the two most important stats, is OBP really three times more important than Slugging Percentage? Is pitches seen/drawn really the most important underlying offensive statistic? Is it true that the only thing that matters for pitchers is their ratio of Strikeouts, Walks, and Home Runs Allowed to batters faced?
OBP and SLG were the two offensive stats most highly correlated with runs scored. But at the time people were more interested in AVG and RBI, while guys with high OBP were undervalued. OBP is most definitely a more valuable stat than AVG as AVG ignores the value of a walk in terms of avoiding outs. And SLG is better than RBI because RBI are heavily team-dependent stats (hard to get RBI without men on base). And OPS (the combination of OBP and SLG) is a decent combination of the two. I don't think they said OBP is three times more valuable than SLG. If anything, I can see an adjustment made to put OBP and SLG on the same scale (since SLG can go up to 4.000 while OBP can only reach 1.000).
As for pitching, I don't think Moneyball professed that the only thing that mattered was K, BB, and HR allowed. They just said that these are the only stats that are solely a function of the pitcher. Everything else (hits, runs, earned runs, etc) are team-dependent in that they rely on the abilities of the defense behind the pitcher as well as the pitcher. And this is most certainly true. But I think the more interesting thing about Moneyball with respect to pitching was that they were the first to start the understanding that wins (and to a lesser degree ERA) were being highly overvalued (because wins were dependent on the team's defense, offense, AND bullpen as well as the actual pitcher).
But the simple answer to your question is: yes. Moneyball was dead on when it came to SABRmetrics. And the proof is in the pudding. Nearly every team in baseball has since hired statisticians to their front offices to help make smarter baseball decisions. And teams like the Red Sox and Yankees have used Moneyball principles (combined with huge budgets) to dominate the regular season for the past decade. Similarly, teams like the Rays have survived thanks to a similar approach to baseball thinking. And while the Cardinals may or may not explicitly be following Moneyball principles, their key players (Pujols, Edmonds/Berkman, Holliday, Carpenter) very much embodied what Moneyball was all about.
The irony is that, in agreeing to be the subject of Michael Lewis' book, Beane arguably torpedoed his team's chances to remain competitive. Now that other teams are using similar statistically-minded approaches to baseball decision making the Athletics and their small budgets are finding it hard to compete again.
Take a look here: http://californiawatch.org/dailyrepo...roid-use-13230
Regardless, Beane has not been able to replace the talent on those teams. When Beane traded away Zito, Mulder, Hudson...he was supposed to have restocked the farm...but those trades have by and large been a failure. Also -- his Oakland A teams have never won even a pennant, and have not been to the playoffs since 2006. He has had plenty of time to draft enough talent to field competitive teams...but it hasnt happened.
LaRussa is seen by many to be "old school" but he doesn't discounts stats...they just aren't the end all be all. Duncan, his pitching coach, is the father of the "chart every single pitch and the result against the hitter". LaRussa was HUGE on pitching and hitting matchups....there are many times Tony would use 3 pitchers in a single inning based on statistical matchups. He experimented with batting the pitcher 8th...under the logic that a decent the 9th place batter is statistically more likely to be on base for the top of the order than an poor hitting 8th place batter would fail to drive in runs left on base from the meat of the order.I have not heard anyone say that the Cardinals (who also have 2 championships in the last 10 years, along with another World Series appearance) are a Sabermetric-oriented team-- and I certainly doubt that Tony LaRussa would acknowledge managing that way-- he's about as "old school" as they get, I'd imagine... and the Reds have started to get good finally (after 10 years in the wilderness), by essentially re-applying the Cardinals' master plan (the Reds have a former Cardinal co-owner and former Cardinal GM, who built their championship teams), and the Cardinals/Reds MO seems to be an emphasis on building up scouting strength (of the old-school variety) rather than statistical analysis.
As to your assertion, CDu, that great Cardinal players typify good OBP and OPS....well of course they do! Those stats were invented to describe good hitters...and will describe great hitters throughout the game no matter the situation. Players that formerly were described to "have a good eye" or "have great power" now have the accompanying stats to describe them...that does not mean moneyball changed the game -- managers will always take good hitters with good eyes and great power if given the option...I think where the real difference happens is where you are dealing with a .250 hitter -- some managers have less patience for hitters with low averages than others. Take a batter that hits .243
with a .330 OBP and .400 slg. The other batter hits .273 with a .310 OBP and a .400 slg. Well, the higher OBP will be on base more...but you only get 1 base per walk, whereas the .273 hitter will have more hits which is better for driving in runs...(players on the bases can advance more than base on a hit)...its an interesting question...which is more valuable, which is better.
The pitching philosophy for the STL Cardinals actually is less about Moneyball than the hitting could be described. The LaRussa Duncan philosophy has been about finding quality/overlooked veterans with small flaws in their game and then correcting them in a way that promotes contact/groundballs and limiting walks, usually by the introduction of sinking pitches.
Agreed on all that-- particularly LaRussa departing from "the book", which Sparky Anderson and others often cited, to do some new things-- things which appear to me to have been almost Sabermetrically driven... and yet, LaRussa never seemed like one to just say-- "this is what the stats say to do/play" (which sometimes seemed to be the position of Beane in Moneyball)...
So in some ways, LaRussa seems like sort of a hybrid of the old school/"feel" manager crossed with the Sabermetric tendencies of the James school... perhaps his genius (if he has any), is knowing when to play the straight percentages, and knowing when the numbers lie, and he should go against them... by all rights, the Cardinals never should have won either of those World Series that they got in the last decade, and really (since Beane says post-season success is mostly luck), the Cardinals shouldn't have even been in the playoffs last year, to win the Series (and they weren't much better in the regular season, in the previous year they won it), so you've got to give LaRussa some kind of credit for something over and above simply laying out his best percentage strategy, and letting the chips fall where they may.
It would have been kind of weird to see what would have happened if LaRussa had stuck around Oakland, when Oakland got poor, and Beane took over as GM-- could those two have co-existed, and would they have agreed on what to do/who to play...
But my point wasn't that the Athletics were a success. My point was just that the concepts discussed in Moneyball were successful concepts, and most definitely succeeded in changing the way people think about the game.
In St. Louis there was some tension between the "old school" approach and the stats driven approach in the Stl front office. It is what led to Jockety (more old school) being shown the door and ending up in Cincinnati and now a more statistical driven approach runs the cards front office and draft. Of course I'm not sure anyone would believe that LaRussa was going to listen to some MIT geek about how to coach baseball. With LaRussa and Duncan (pitching coach) retiring I think they believe there will be more symmetry between the major league club and the farm system.
As to last years world series I would argue that the club that played the last 6 weeks of the regular season and playoffs had little in common with the team that played most of the regular season. The Rasmus trade was probably addition by subtraction to start and than the Cards added 4 significant pieces. Rafi Furcal at SS which was a dismal position before his arrival. Edwin Jackson as a starting pitcher. Octavio Dotel who was great for bullpen morale and was lights out against Braun in the playoffs. Plus Marc Rzepczynski who was the left handed reliever that was completely missing from the cards bullpen before his arrival. On top of that Jason Motte came out of the blue to be a great or at least hot closer for the stretch drive. Is what a completely different team that went 23-9 to close the season. They were magical!!!
Kex you are right to point out the dissonance between TLR and management. LaRussa is very intense and doesn't coddle rookies. They have to earn playing time. He usually doesn't "let them learn" on the field which I am sure management feels inhibits development....but Tony is a very good judgement of talent and character. The only ST Louis player I can think of "that got away" and was really successful elsewhere was Dan Haren. The human brain is an amazing piece of pattern recognition machinary...and after 40+ years of baseball, managers like LaRussa are a tremendous asset for separating baseball wheat/chaffe.
CDu, I wasn't clear enough in my previous post. I give the SABR school plenty of credit for providing more stats and more specific ways to numerically define players. The work especially on the defensive end of the field with things like Range Factor has been awesome...
I just can't see Oakland's success in the early aughts as a result of finding overlooked talent using these sabr stats when it was a product of high draft picks from bad seasons and steroid fueled MVP seasons. Every franchise that doesn't have the money to compete with the top end of the league will go through cycles of winning and losing as their farm systems get replenished/depleted/replenished. Losing gets more sure-thing/cant-miss talent (well, as much as there is in baseball). Thats why long periods of success are celebrated (see Braves -- early 1990s-early 2000s) despite not really winning much in the post season.
The team played out of its mind in August and September, make no mistake --- but the whole team got healthy and hot at the same time. Tony was pushing the right buttons -- but the Cardinals would have been a much better team without the 26 blown saves -- by the end of the season they had jettisoned Franklin who was a complete failure as closer to start the season as well as Trevor Millar and Bautista.
Also -- I think you mean with Wainwright out, not Carpenter -- tho I forgive you because Carpenter is out for the unforseeable future....yes pitching was a key question, but mitigated in the offseason because your #1 can always pitch twice in the series.
The recent failures of Billy Beane should be attributed to the fact that now everyone is working off those sabermetrics. In fact, in order to compensate for this, Theo Epstein began going against the plan of drafting proven college players and going for high school players that he could afford to pay out of HS b/c of the Red Sox financial situation and while he knew most would be busts, he could no longer get the value he once enjoyed. So I believe that is why you see some return to more traditional views b/c the value is no longer fully in Sabermetrics and statistics are like gambling, over the long haul you can minimize risk but one outcome might not follow the statistics. It reminds me of the article about Shane, the Rockets, and Kobe, when even when Shane played perfect defense and followed the statistics Kobe still burned him. Does that mean over 1,000 trials it still isn't the right move? So I'd hardly say La Russa winning a championship, when the Cards could have easily missed the playoffs if the Phillies mailed it in versus the Braves, some indication that Moneyball/Sabermetrics failed and the old school came through. At some point, its just about having good players get hot and buck statistics.
I'll post the SI article which more articulates the point that refutes your last point about it all being about OBP, SLG%, OPS, etc.
http://msn.foxsports.com/mlb/story/M...shifted-092011The book was not about how the A’s, a team at the bottom of baseball’s economic food chain, relied on less traditional statistics such as on-base percentage (OBP) to compete with wealthier clubs.
No, it was about how the A’s tried to find players with undervalued skills so they could acquire those players cheaply, exploiting inefficiencies in the market.
At the time the book was written, a player’s ability to get on base was undervalued. No longer is that the case, and other inefficiencies are becoming increasingly difficult to identify.
Advances in technology created such a wealth of new information, teams have become increasingly adept at assessing virtually every aspect of a player’s performance — and assigning a value to it.
“Those who jumped in early got a head start,” St. Louis Cardinals chairman Bill DeWitt Jr. says. “But most clubs now, one way or another . . . they’re on it. No one is saying, ‘What is this? What am I missing?’ They’re all on it. Everyone has their different models for success.”
Last edited by sporthenry; 04-16-2012 at 12:02 PM.
I was reading Bill James books before some people here were born. Wait, I can do better. You know those college essay questions where you have write about one person you'd want to have dinner with? I wrote about James. That was 1988. Got me into Duke at least.
As far as sabermetrics in general goes, the jury's been in for some time. Compare a baseball telecast 30 years ago and now. Compare a column from a sportswriter, or a debate between fans, or the information available to an average fan. James started by saving old newspapers, because that's the only way he could compile the stats he wanted to know about. The leagues wouldn't give it to him. Stuff that's now part of baseball's cosmic background radiation, like batting average going way down after a stolen base, or the average major leaguer peaking at age 27, or a thousand other tidbits we take for granted now, we can thank sabermetrics for teaching us. Fifteen years ago Jack Morris was a lock for the Hall of Fame, as even James wrote, and no one took Bert Cryleven's chances seriously.
On "moneyball" itself, I think a couple of things limit its effectiveness. One is the structural ineffeciencies in the major leagues. Big surprise here, however you crunch the numbers they still say Albert Pujols is the best player in the game. Ten years ago it was Barry Bonds, and ten years before that it was... well, still bonds, and going back another 10 it was Mike Schmidt. (The way James put it: A statistic which is never surprising is boring; a statistic which is always surprising is probably wrong.) There's still no salary cap to give teams a more equal opportunity at the best players, so players end up at teams who can pay for them.
Another factor, alluded to by others, is that shorter sample sizes naturally have variations that make big-picture sabermetrics fail sometimes. I think one way to compensate for this is manage differently and take more risks when the stakes are higher, ie the postseason. You need to go all out and do more in-game and roster-level micromanaging because the cost of one loss is greater the later in the season you get, and because the other guy is doing it, too. Be Tony LaRussa, rather than Bobby Cox.
Last edited by hurleyfor3; 04-16-2012 at 04:57 PM.
You must spread some comments around before flaming the Moderators again.
I still think Pujols raising his average by ~25 points and Furcal playing better defense at short is a lot to hang a championship on, when it was nowhere in sight before that-- but, if the bullpen became that much better, that is big-- that can cost you 5-10 games easy, in the course of 2 months, if those guys suddenly got a lot better, and I guess I wasn't following the Cardinals bullpen that closely.