Book Discussion: Moneyball by Michael Lewis

What did you think about Billy Beane’s approach with the Oakland As? Are there are any flaws? Why do you think so many people are resistant to the concept (besides the fact that the book’s tone is clearly insulting to the old school guys).

46 Responses to “Book Discussion: Moneyball by Michael Lewis”

  1. Marc

    I think one of the great things about this book and discussion is that we now have four or five years to evaluate some of the ideas, players, and general managers that are discussed in the book.

    I’d summarize the book in this way, Billy Beane runs a team with one of the lowest budgets in baseball. He is routinely trying to compete with teams that can spend much more money so he the edge he tries to exploit is using statistical analysis to determine how to obtain players who have skills that are undervalued and are worth a lot more than the A’s are able to pay. He also tries to identify the strategies that work most often and advises his field manager to utilize these strategies even if they go against common baseball traditions or conventions.

    So although I think that a lot of Beane’s strategies work, the first thing I’d like to look at are some of the other GM’s who are discussed in the book: one of the general managers that I recall being disparaged in the book is Ken Williams of the Chicago White Sox. Somewhat problematically for the author is that the White Sox have since won the World Series and are in serious contention for the playoffs this year using a fairly old school approach. While I have some ideas on this topic, I’ll stop here and let things perculate a little bit.

  2. Mitchell

    The old school approach is popular for two very good reasons. First, baseball is a game whose history influences the present on every play. A pitcher selects a pitch based on what the hitter has done in the past; a hitter is making guesses based on a pitcher’s tendencies. A game whose very gameplay is based so strongly on history is reluctant to change its ways. It happens, every so often, as when Cal Ripken, Jr. broke the tradition of big men not playing the shortstop position, or when Tony LaRussa began hitting his pitcher in the eighth position. However, if you play the averages and you know your personnel, you will MOST LIKELY succeed, whereas if you play a hunch and it fails, you’re considered an idiot. Managers don’t like to look like idiots.

    The other reason the old school ways stick around for as long as they do, for better or worse, is that baseball is run by “baseball people,” most of the time. Baseball people trust baseball people, and award influential positions to baseball people. Guys like Billy Beane and Theo Epstein are generally looked down upon by baseball people because they’re just looking at numbers and not respecting the game’s traditions.

    The argument can (and should) be made that Oakland hasn’t won a World Series since 1989. However, it has been successful (depending on how you define success) for quite some time in the years since. One very interesting thing Mike Wilbon and Bob Ryan discussed on PTI was that if you look at guys Beane trades away, they seem to be on the verge of not being nearly as good as they were. Consider Mark Mulder, Jason Giambi, and Barry Zito.

    The other thing I liked about this book was that Beane made it clear that guys like Giambi and Miguel Tejada are great players, and if he COULD pay them he WOULD. His model, he says, isn’t the ideal, because he’d much rather have held onto these MVPs. Giambi and Tejada (and Eric Chavez, too) were brought up in the Athletics’ farm system, where on-base percentage is preached like it’s gospel; if you look at the way they play the game even today, you can see how that has never left these guys.

    One huge criticism that I hear a lot is from Joe Morgan, who seems personally insulted by Beane’s devaluing of the stolen base. Of course Morgan would disagree with Beane on this, because Morgan was a base-stealer. His point, though, is that if you have speed and if you can turn a single into a double (which is what an SB basically does), why wouldn’t you? It’s an extra weapon you have that someone else might not. Beane’s argument is that the stolen base, when compared to the risk of the out and the benefit of possible runs, more times than not is a losing strategy.

    I’ve often wondered what guys like Earl Weaver would say about this. Weaver, for a period of a few seasons, was the game’s winningest manager and his strategy was always to play for the big inning. Small ball, which guys like Whitey Herzog preached, wasn’t playing the odds well enough.

    One wonders what Theo Epstein’s success rate would be if he wasn’t doing his thing with the payroll of the BoSox. If he were given Minnesota or Oakland, for example, would he be as successful as he is now, or even as successful as Beane?

  3. Mitchell

    Top Ten Baseball Books (subject to change because I’m going to do this off the top of my head without really thinking about it).

    1. The Bronx Zoo by Sparky Lyle.
    2. Number 1 by Billy Martin
    3. Moneyball by Michael Lewis
    4. Nine Innings by Daniel Okrent
    5. Balls by Graig Nettles
    6. Temporary Insanity by Jay Johnstone.
    7. The Umpire Strikes Back by Ron Luciano
    8. Good Enough to Dream by Roger Kahn
    9. Catcher in the Wry by Bob Uecker
    10. Ball Four by Jim Bouton (actually, one of the sequels to this, I’m Glad You Didn’t Take it Personally, is probably better)
  4. Marc

    Like I said, I think Beane’s approach works to a point (after all the A’s won the AL West and went to the ALCS in 2006) but I wanted to point out that old schoolers like Ken Williams have had considerable success so the sabermetric Moneyball approach isn’t the end-all. In particular, Williams is known as someone who’s very involved in trades because he doesn’t sit across a table and insist on the best possible deal, he’s willing to deal value for value as long as his team benefits in the way that he wants them to.

    I agree with your inclusion of Theo Epstein of the Red Sox as a “Moneyball-style” GM. I think that the Red Sox are a great example of what you can do it you apply Beane’s principles AND you have money. The Red Sox have won two world series since *Moneyball* was published.

    I sort of agree with the idea that the A’s players that Beane lets get away generally do worse. Giambi (left as a free agent) is a tough case to analyze because of the steroid issue that hangs over him but he had two excellent years (40+ homers, 400+ OBP) after leaving Oakland before diminishing. Mulder (traded for Danny Haren, great move) has had serious injury problems and Zito (free agent) is a disaster that everyone but the Giants saw coming but Beane traded away Tim Hudson for basically nothing and Hudson has done very well. Miguel Tejada (free agent) has done quite well too and actually never was a typical Oakland player (career OBP 340). I think time will tell regarding the trades that were made this year as Beane traded away Nick Swisher, Danny Haren, and Rich Harden. I think those guys have a lot of good years ahead of them.

    Another prominently featured aspect of the book was the Oakland approach to the draft. Although I don’t have the book in front of me, I don’t know that they got this right despite the celebration that was depicted (“this just doesn’t happen…”). As far as I can tell, none of the players taken in that draft currently play for Oakland five years later (see Nick Swisher, Mark Teahen, Joe Blanton) and some are no longer in baseball at all (see Jeremy Brown). In addition, some high school players that Beane would never have taken are thriving (see Prince Fielder – too fat for Oakland??, Scott Kazmir – won this year’s all-star game). What do you think about that?

  5. Reid

    Check out this article, Billy Beane’s Perfect Draft. The author anaylzes the draft in Moneyball. I didn’t read the entire article, but they basically analyze his draft picks.

    There’s a lot to comment on, but I’ll try to focus a couple of issues. Before I start, let me say that I Beane’s strategy is effective in securing/releasing players for small budget team. The key is to minimize risk here, and the approach seems pretty effective.

    However, I don’t know if his strategy works so well, when applied to the game–especially in the playoffs. Beane’s approach is based on probability, choosing the a course of action that will be favorable over many games. If you do everything to avoid outs, including almost never attempting to steal a base, over a period of games, this will prove effective–i.e. you’ll win more games than you lose. However, in the playoffs every game counts, you can’t afford to lose games and say, “Well, we’d win more games than we’d lose in the long run.” My sense right now is that the better strategy is to look for opportunities to go against the averages and take some chances. For that you have to rely on human judgment–the knowledge and experience of a good manager.

    There’s another difference between playoffs and the regular season, and I want to throw out a theory I’ve been thinking about that can explain this difference. Basically, one of the main differences between the regular season and the playoffs is pressure. The games in the post-season are so important–with the World Series games being the most important–that it creates pressure, unknown in regular season games. My theory is that a successful post-season team will effectively minimize the pressure on their players while increasing the pressure for the other team. In other words handle this pressure better than your opponent, at the same time use that pressure against them.

    Let’s examine this by discussing a situation Lewis describes in the book, the A’s first game loss to the Yankees in the ALCS (don’t remember the year). Here it is:

    Down 5-4 in the eighth inning, Yankees second baseman had gotten himself on base and stolen second. Derek Jeter than walked, and Jason Giambi singled in Soriano. Bernie Williams then hit a three-run homer. A reasonable person, examining that sequence of events, says, “Whew, thank God Soriano didn’t get caught stealing; it was, in retrospect, a stupid risk that could have killed the whole rally. Joe Morgan looked at it and announced that Soriano stealing second, the only bit of “manufacturing” in the production line, was the cause. Amazingingly, Morgan concluded that day’s lesson about baseball strategy by saying, “You sit and wait for a three-run homer, you’re still going to be sitting there.”

    I don’t think Soriano’s steal was a necessarily a “stupid” risk. First of all, the steal puts more pressure on the defense. A wider range of singles could now get Soriano home. At the same time, the steal may plausibly reduce the pressure on the batter. They don’t have to get a big hit to drive in a run. Players who normally wouldn’t go for a big hit might be tempted to, but a steal like this can reduce that temptation. Also, the possibility of stealing adds pressure to the other team. Of course, if the team knows you don’t steal or you don’t have the personal to steal or run the bases well, then this reduces the pressure on the opposing team. Second, I say the steal is not “necessarily” a stupid risk because a whole host of circumstances could make this risk a reasonable one. Maybe Soriano really reads the pitcher’s pick-off move really well; maybe the pitcher has an injury that makes his pick-off move less effective; maybe the catcher has a sore shoulder, etc. These are things that are outside of the percentages, but could determine that the risk is worth it or not. By the way, I don’t think the risk is stupid just because Giambi hit a single later. Maybe he got a hit partly because Soriano stole second. How? By taking off the pressure, by not tempting him to do too much. (I’m wondering if we could get statistics on this sort of thing: OBP of a batter after a steal; pitching ERA after a steal, etc.)

    If my theory about pressure is correct, than this type of risk, which would decrease pressure on your team while increasing pressure on the opponents, would be worth it. Of course, Soriano could get caught stealing, but the question is, is the risk worth it?

    There’s another aspect related to this and that’s Beane’s attitude towards relievers. The book gives the impression that they’re not that critical. He trades a lot of them to get other players, seemingly with the idea that he can develop another. That doesn’t sound right to me. (And I don’t know if this is actually Beane’s approach and attitude. Can anyone confirm his approach?) A really good-to-great reliever is valuable in the playoffs precisely because he reduces pressure for your team and increases pressure for the opponent. Think about Mariano Rivera. Everyone knows (or used to know) that if you have a one run going into the 9th, the game was practically over (anything more than one run and forget about it). That’s huge psychological advantage. Again, this advantage may not be so great during the regular season when the games don’t matter so much, but when the stakes are high having a reliever like Rivera helps your team psychologically while hurting your opponents. Batters don’t feel pressure to score a lot of runs and the other teams feel mounting pressure as the game progresses to score runs. (I have a feeling that a lack of a good to great closer was the big reason the Braves didn’t have a post-season record on par with their regular season. As good as they were in the 90s regular seasons, you gotta say they were a let down in the playoffs.)

  6. Mitchell

    As far as I can tell, none of the players taken in that draft currently play for Oakland five years later (see Nick Swisher, Mark Teahen, Joe Blanton) and some are no longer in baseball at all (see Jeremy Brown). In addition, some high school players that Beane would never have taken are thriving (see Prince Fielder – too fat for Oakland??, Scott Kazmir – won this year’s all-star game). What do you think about that?

    The baseball draft is such a diceroll. The only way really to evaluate this is by comparing the drafts of all the other teams, over time, to see whether or not this works. We do know that Giambi, Chavez, Mulder, Hudson, Zito, Harden, and Street are products of the Oakland farm system (I don’t know if Oakland drafted them or not).

    It’s true that Oakland’s insistence on not drafting high-schoolers means they never had an A-Rod or a Griffey, but how well did Seattle’s drafting of these guys pay off for the Ms, or even Fielder and Kazmir (so far) for their teams? The argument could be made that both teams did okay, but not great.

  7. Marc

    Regarding the draft-

    I agree that the draft is a diceroll and I think Beane would too, but Beane’s (and colleagues, especially Paul DePodesta – who washed out as Dodgers GM several years later) assertion is that statistical analysis make it much less of one. To my eyes, the evidence regarding that draft in particular does not support the presumption that the Oakland draft strategy is superior. Given 7 draft picks in the first three rounds and the elation that was depicted in the book, I think that having only three players in the big leagues six years later, none on your own team, has to be considered disappointing.

    Regarding the M’s, if they had not drafted Griffey or A-Rod, Seattle would not have a team today. Fact. See the 1995 season where they were set to leave if they didn’t get a stadium deal. Do you really believe that Beane would have passed those guys up if he had the chance? Remember that both those players were in the big leagues less than three years after they were drafted and both are first-ballot hall-of-famers. I’ll throw this in too to confound things a little – Beane drafted Eric Chavez out of high school, and the Mariners drafted Rich Harden out of high school well before anyone else had heard of him (Harden is from British Columbia) only to fail to sign him. Oakland drafted Harden two years later and paid him six times what the Mariner’s offered – look it up.

  8. Marc

    Regarding pressure:

    I don’t know if you can statistically analyze this, so it’s a hard thing to quantify. It’s very difficult to compare their regular season stats to their post-season stats because the sample size for the post-season is far too small for most. For a quick example, compare these three rather famous Yankees, two of whom are regarded as “clutch” under pressure, and the other regarded as somewhat of a “choker”:

    Derek Jeter regular season career: Batting 315 OBP 386 Slugging 459
    Derek Jeter playoff career (119 games): Batting 308 OBP 373 Slugging 473

    Alex Rodriguez regular season: Batting 306 OBP 389 Slugging 579
    Alex Rodriguez playoffs (39 games): Batting 279 OBP 373 Slugging 483

    Mariano Rivera regular season: ERA 2.15 saves: 471, blown saves 60
    Mariano Rivera playoffs : ERA .082 saves 31 blown saves 4

    I’m generalizing here based on very few numbers, but I would conclude the following:

    There is no statistical evidence that Derek Jeter is a clutch player. His regular season and post-season stats are virtually identical. Instead I would characterize him as an excellent, probably hall-of-fame caliber player, who continues to play at a high level in the playoffs but without evidence that he raises his level of play. Is that clutch?

    There may be evidence that A-Rod is a choker but we would be drawing these conclusions based on a sample size of 39 games, hardly conclusive. However, consider that while his post-season stats are somewhat lower than his career regular season stats, his OBP and slugging percentage are comparable or better than Jeter’s. If someone plays at an equal or higher level than the clutch captain of the team, how is he a choker?

    I would absolutely conclude that Rivera is a clutch player in the post-season. His ERA goes down significantly. In addition he saves most of the games that he enters and although I didn’t note this, he enters many post-season games in the 7th or 8th inning rather than waiting for the ninth, which is a lot more in the mold of Goose Gossage and the older generation of relievers. Hall of famer for sure. One of the best relievers of all time.

    But how do get these sorts of players? I think the Yankees struck the jackpot and got very lucky with Mariano Rivera. Remember that while he succeeds as a reliever with one pitch he was originally signed and brought up the majors as a starter. He failed miserably as a starter and my guess is because you can’t make it through a lineup three times with only one pitch. When you can use that pitch in short bursts as a reliever, its much more effective. When major league hitter see the same pitch ten to fifteen times in a single game, they start to hammer it.

    I side with Beane on this one. His view is that you build a best team that you can based on statisticaly analysis and your resources. Hopefully that gets you to the playoffs. A short series is usually but not always won by the best team because there’s a fair amount of luck and variation in the short term. And I just don’t know how you can quantify ability under pressure for most players.

  9. Reid

    Re: the Draft

    One of things I think would be wise (and maybe teams have done this) is to look at create a data base of scout’s and executive’s picks. You could then see who were the most successful at picking out players that became good, and you could compare it to Beane’s approach.

    I also wanted to point out that the A’s (at least in the book) relied heavily on trading players during the season, which seemed to help them have a great second half.

    Re: Playoffs

    I don’t know if looking at OBP and batting average are a good way to gauge clutch performance. Being clutch is situational. Do you get the hits in the specific moments you need them. To gauge whether a person was clutch or not, you’d have to look at specific moments when the pressure was on and the team really needed something from you. Determining “when the pressure is on” is relatively subjective, but I think you could get some agreement.

    Also, measuring pressure by statistics is tricky. Some players actually perform better when the pressure is on. But is empirical evidence necessary to prove that there is greater pressure in the playoffs?

    Anyway, what you’re saying doesn’t really address the point I was trying to make (and maybe you weren’t trying to address the point I was trying to make, I don’t know). Do you guys agree on my reasoning for not relying too much on statistics for your strategy in the playoffs?

    If you agree, I would also argue that your regular season approach should become like the one you take in the playoffs.

  10. Mitchell

    The argument could be made that EVERY at-bat, for studs like Jeter and A-Rod, is a clutch situation in the post-season, but I won’t go there. There ARE statistics for clutch-hitting; I’m sure Bill James has invented at least a handful, but OBP with runners in scoring position or with two outs has to be considered.

    Reid’s suggestion that assembling a team (and playing the game) during the regular season in the manner that is most likely to win the post-season is not bad, but I have always been against it. First of all, if your team is up in the standings by five or six games and you’ve got an afternoon game against a weak team, you’ve got to sit some of your studs on the bench, something you’d never do in the post-season. You have to manage your team to the post-season, not play every game as if it were already the post-season.

    There are managers who always coached that way, and Billy Martin was one of them. I’d say that Ozzie Guillen is another. The thing is, those guys couldn’t manage one team over the long haul, because their management style might work for one or two seasons, but it gets old. The only reason Ozzie is still in Chicago is that his team is in first place; I guarantee that if it were in the middle of the pack, he’d be out of there.

    Which brings me to my second reason you can’t manage a team that way in the regular season. It doesn’t work. You might win one world series, which Guillen did, but you’ll finish a lot of your other seasons nowhere near the post-season. Would you rather win two World Series in ten years with eight years of mediocrity to balance that out, or would you rather win one World Series in fifteen years, with twelve or thirteen unsuccessful playoff appearances to balance THAT out? I’m sorry, but I’ll take the one title and the guaranteed post-season. I totally agree with the premise that what it takes to win a season over the long haul (and repeatedly) doesn’t seem to be too successful in the post-season, but give me Bobby Cox or Joe Torre or Mike Scioscia or Chuck Tanner any day over Ozzie Guillen and Bobby Valentine (‘though I do really like Bobby Valentine, because he DOES play his hunches; his hunches are just super crazy sometimes…good if you love baseball, but bad if you’re a Mets fan!).

    An interesting case is Terry Francona, who on the field seems like a Guillen or Valentine to me, but who, off the field, seems to have exactly the right attitude for managing a circus such as a Boston sports team. The guy is perfect at deflecting criticism off his players and aiming it at himself. I loved it when Mike Wilbon asked him how long he was able to enjoy his status as World Series champion before worrying about next year, and Francona said he never got to enjoy it because he was always worried. The guy acts like everything good his players do is exactly what should be expected and that any success he has is all due to his players. I have never liked watching him on the field (oh, except during the all-star game, which I thought was hilarious!), but I love seeing his interviews off it.

  11. Marc

    Regarding pressure: Many people smarter than me have tried to figure out how to measure perfomance in the clutch. Nobody has been successful yet. Trust me, if someone had figured it out, we’d know by now. It’s just really hard to do because of limited sample size and because even batting situations that seem similar can be wildly different depending on who the pitcher is (a hitters stats will change if he’s facing Joe Schmo in a “clutch” situation instead of facing Mariano Rivera). I guess that would be my main point about pressure which wasn’t that clear, unless it’s really obvious you can’t figure out who will perform better in clutch situations. My feeling is that with Rivera, it’s obvious that he is even better in the playoff than he is during the regular season. With A-Rod or Jeter, it’s not. With that in mind, I think that GM’s need to default to finding the best players possible, and there is a wealth of information to help people determine this.

    I’d have to answer Reid’s question about relying on statistics for strategy in the playoffs this way:
    1. Assuming that I’m the general manager, I think the best way to succeed in the playoffs is to put out the best team possible. That includes measuring players and paying them within your budget. So in that way, I believe that you must rely on statistics.

    2. Assuming I’m the field manager, then the single most important strategic decision I can make is who to put on the pitcher’s mound. The pitcher is the most important part of any play in baseball. As a manager, you often can’t control in a single at-bat who will be batting but you can control who will be making the pitch. In deciding who to start a game and when to bring in a reliever, I believe that you have to rely heavily but not exclusively on statistics.

    3. Selecting the batting lineup and deciding when to pinch hit are the next most important decisions and these should also be guided by statistics. But batting decisions are not nearly as important as pitching decisions.

    4. I think that all other decisions pale in comparison to the pitching decision. Yes, you can talk about when to hit-run, when to steal, when to play your infielders in, when to substitute defensively, when to bunt, when to pitch out, etc. There’s an incredible amount of minute decisions that need to be made. But I really believe that at the core, the biggest key to baseball is determining who will win the matchup between the pitcher and the batter. So I think that *Moneyball* overstates the importance that statistics play in the in-game management of baseball that doesn’t involve selecting the pitcher or hitter.

    5. Don’t hear me incorrectly. Pitching and batting aren’t the only things but they are the most important things. Yes, defense, speed, base stealing, and other skills are important, just less important.

    You might have thought that once *Moneyball* came out, everyone would be on to Beane’s act. After the book came out, OBP became the chic statistic and obviously the A’s couldn’t target these players anymore. Beane analyzed things and decided that defense was now being undervalued so he targeted great defensive players because he could get them more cheaply and the other GMs weren’t on to what he was doing. End result, the A’s won the AL West in 2006 (just two years ago), swept the Red Sox in the ADLS (finally getting that playoff series monkey off Beane’s back), and lost to a clearly superior Detroit Tigers team in the ALCS.

  12. Marc

    boy there’s so much good stuff in this thread…

    regarding relief pitchers. I don’t actually think that Beane considers them noncritical, unimportant or expendable. I think that he feels that he is better than other GMs at identifying good relievers and getting them for less dough which is considerably different.

    The Oakland bullpen is not revolutionary in the set up, it’s the same setup as other major league bullpens with a designated closer and designated setup men. What was different about past Oakland closers is that Beane used to find retread relievers or starters to fill the closer role and was mostly successful, although it’s worth noting that when he wasn’t successful he went out and found a closer in a conventional fashion.

    Just to review Oakland closer history:
    late 90s (post Eck): Billy Taylor had moderate success and was traded mid-season but I honestly don’t think anyone thought he was all that great.
    2000: Jason Isringhausen. Always had been thought to have great ability but was a failure as a starter. Beane moved him to the closer role for a cheap salary and he thrived there. Isringhausen pitched great in Oakland and left for St. Louis where he has generally performed great for the Cards, helping them get to 2 world series.
    2001-2002: Billy Koch. Was a setup man in Toronto with a 100 mph fastball. Beane picked him up, turned him into a cheap closer, and Koch thrived. Before Koch could leave for bigger bucks, Beane traded him for…
    2003: Keith Foulke. Foulke was actually a very good closer for the White Sox around 2000 but he did it in unconventional ways with a 89 MPH fastball and a great changeup (kind of what Trevor Hoffman does these days). Beane picked him up for Koch, Foulke did great as a closer and left for Boston as a free agent and big bucks. As the Red Sox closer, Foulke was widely considered the final piece of the puzzle that allowed the 2004 team to “reverse the curse” and win the world series against St. Louis after coming back from 3-0 in the ALCS against the Yankees. After 2004 though, Foulke has basically stunk.
    2004: Arthur Rhodes. this was another case of a retread with a 99 mph fastball with great success as a setup man who Beane converted to closer. Rhodes couldn’t do it and was terrible in the new role, he has since returned to being a very reliable setup man. The 2004 A’s were terrible in part due to their poor bullpen.
    2005-present: Huston Street. After failing with Rhodes, Beane went sort of old school and drafted a college reliever in the first round. In *Moneyball* Beane is pretty disparaging of college relievers but he went ahead and drafted for a need. Street has been good but certainly won’t be mistaken for Mariano Rivera anytime soon.

    Bottom line: It looks to me like Beane considers relievers important but given his budgetary constraints the bullpen is where he tries to stretch his dollar the most. Even in the end though, he went against the *Moneyball* stereotype and drafted a college reliever and has had some good results.

  13. Reid


    I think we’re thinking of approaching regular season games the way you would post-season ones differently. By “approach” I’m referring primarily to tactics and style of play. The specific tactics I’m thinking of are hit-and-run, base stealing–tactics that fall under the “manufacturing runs” umbrella, which Beane seems to scoff at (based on Lewis’ description). I tried to show why using these tactics can be valuable and effective in the playoffs (reducing pressure on your team while increasing pressure on your opponents) even though they cost games in the regular season. All I’m saying is that whatever style or tactics would be especially helpful in the playoffs should be used in the regular season. You don’t want a team using a “big ball” approach switching to a “small ball” approach in the playoffs. That’s kind of a ridiculous example, but hopefully you get my point. I am not saying that you treat a regular season game exactly the same way would a playoff game. Giving key players a day off because your team has a significant lead and playing a weaker team does not violate the approach I’m advocating. (I’m not completely clear on what you mean by managing a regular season game like it’s a post-season one. In what ways did Martin, Guillen, et. al do this?)

  14. Reid


    Here’s the quote from the book that made me think Beane undervalues relievers:

    The central insight that lead him both to turn minor league nobodies into successful big league closers and to refuse to pay them the many millions a year they demanded once they became free agents was that it was more efficient to create a closer than to buy one. Established closers were systematically overpriced, in large part because of the statistic by which closers were judged in the marketplace: “saves.” The very word made the guy who achieved them sound vitally important. But the situation typically described by the save–was clearly far less critical than a lot of other situations pitchers faced. The closer’s statistics did not have the power of language; it was just a number. You could take a slightly above average and drop him into the closer’s role, let him accumulate some gaudy number of saves, and then sell him off. You could, in essence, buy a stock, pump it up with false publicity, and sell it off for much more than you’d paid for it. Billy Beane had already done it twice, and assumed he could do so over and over again.

    Lewis describes losing Isringhausen not as a “loss” but a “happy consequence of a money machine known as ‘Selling the Closer.’ Late Lewis says, “Finding pitchers who could become successful closers wasn’t all that difficult.”

    To me this sounds like undervaluing the closer or at least a belief that they’re not hard to get. I don’t have any data, but my impression is that finding really good-to-great closers is really difficult. Also, the quotes above make it sound as if the primary value of a closer is to trade them to other teams who overvalue them and get supposedly a lot more in return.

    My impression is that a really good-to-great closer is critical to winning the World Series. If we looked at the World Series Champion since the advent of the closer, I wonder if all the champions would have very good-to-great closers. I also wonder how often the ones that made the playoffs, but failed to win had weaker closers. Actually, it’d be interesting to compare the differences between all the teams that made the playoffs but failed to advance with the teams that did won it all. What would be the key differences in terms of characteristics and statistics? Does anyone know?

  15. Marc

    Well, I wonder if we agree but are looking at things differently. Even looking at the quote above, I still believe that Beane considers the closer to be important, he just is convinced that he is able to find one easier and cheaper than other GMs. I think that Beane feels that the closer is overvalued by other GMs. Maybe this is semantics. I don’t think that the A’s lost any of their tight playoff series because of their closer. I think in three cases they lost to a better team (2000 and 2001 to the Yankees, 2003 to the Red Sox). In 2002 losing to the Twins was an upset.

    Let’s look at world series champs back to 1995 and their closers. Although some of the closers were very good-great, this certainly isn’t a requirement. In most cases, I think the team with the better starting rotation won, including the Yankees juggernaut from 98-00 (El Duque, Clemens, Wells, Pettite, Cone in various combinations) which also had the infamous Jeff Nelson righty/Mike Stanton lefty bullpen combination.

    07 Red Sox – Jon Papelbon = great closer but remember the rotation of Beckett (untouchable in the playoffs in 07), Schilling, Dice-K, and Wakefield. No comparison between the Sox and the Rockies.

    06 Cardinals – Isringhausen was hurt, the closer was Adam Wainwright. I don’t think so. The story of the series was probably poor fielding by the Detroit Tigers pitchers who had critical errors in each game that led to big runs that decided three of the four games.

    05 White Sox – Bobby Jenks = became the closer during the playoffs as a lucky find after joining the team in August… the Sox were carried and defined by their starters (Contreras, Buerhle, Garland, Garcia). In the ALCS the White Sox won in five games and each of the four starters pitched a complete game victory which was unheard of. The bullpen was considered an unproven weakness but it didn’t really matter…

    04 Red Sox – Foulke = great closer then although has stunk it up since. Still, the starting rotation clearly outclassed the Cardinals and included two potential hall-of-famers in their prime: Pedro, Schilling, Derek Lowe, and Tim Wakefield

    03 Marlins – Ugueth Urbina = good closer, not great, definitely not as good as he was in the 90’s. the story of the series was the Flordia starters including Brad Penny (won two games) and Josh Beckett (17 innings, one run allowed, shut out the Yankees on three days rest in Yankee Stadium in game 6).

    02 Angels – Troy Percival = good/great closer lost some mph by this time and gave up a monster homer to Barry Bonds. While I don’t want to discount Percival’s greatness, the relief pitcher who really made the biggest difference for the Angels was Francisco Rodriguez. Neither team really had a great pitching staff, strange year.

    01 Diamondbacks – Bum Yung Kim = yeah right…the story of the series was Randy Johnson/Curt Schilling and a game 7 win over the best playoff closer of all time… well… then again the Diamondbacks closer in game 7 was Randy Johnson who was certainly good-great.

    00 Yankees – Rivera

    99 Yankees – Rivera. Beat the Braves and Maddux/Glavine/Smoltz who were clearly on their downside.

    98 Yankees – Rivera = possibly the best team of all time, 125 wins. Overwhelmed the Padres and Trevor Hoffman.

    97 Marlins – Robb Nen = good closer, inconsistent. 100 mph fastball. Best pitchers belonged to the Marlins who had a hard throwing Livan Hernandez (now a junk baller) and Kevin Brown in his prime.

    96 Yankees – John Wetteland (Rivera was the setup man). Probably the one year where the losers (Braves) had the better rotation, the big blow was a homer allowed by Wohlers so a closer who failed was the big story here.

    95 Braves – Mark Wohlers = good closer, inconsistent, 100 mph fastball. I’d have to say that the Braves were dominated by three hall of fame starters in their prime including Greg Maddux, John Smoltz, and Tom Glavine.

  16. Mitchell

    I can’t say this totally for Ozzie Guillen, because I don’t see him manage enough, but I know that Billy Martin hated to lose; hated losing a regular season game as much as a post-season game. If his team lost, he didn’t think anyone should be eating from the clubhouse spread after the game, and sometimes turned the table over to make sure nobody ate. If his team lost, he was angry and miserable and expected his players to be angry and miserable, too. He exhorted his players to push themselves as if every game were Game Seven. He yelled, he howled, he praised, he directed. Most guys liked playing for Martin, but the act tended to wear thin, and guys stopped responding. That’s why Martin’s stints were so short and why he kept getting hired. If you needed a guy to turn your club around for one or two seasons, he could do it. But he simply could not maintain a consistent level of excellent performance from hs players.

    I suspect Guillen is the same way. His antics work, but his players get tired of him, I think.

    This is in contrast to those calm, steady guys like Torre and Cox, who simply had faith that their guys would come through, because the averages said they eventually would.

  17. Reid


    Yes, Beane (according to Lewis) didn’t want to pay millions for a closer because he felt that creating one was more efficient. That makes it sound like creating a truly good-to-great closer is relatively doable and easy. Lewis says, “Finding pitchers who could become successful closers wasn’t all that difficult.” I disagree with that. Have the A’s had reallly good-to-great closers? There are parts of the quotes that indicate he doesn’t place a high value on closers: calling the loss of Isringhausen not a loss but a “money machine;” saying that saves make a closer sound vitally important (suggesting they’re really not.)

    Why do you say that having a good closer was not a requirement? In your list, all but one of the winners (’01 Diamondbacks, and I don’t know how good or bad Kim was that year) had a good-to-great closer, at least for that year.

    You mention that the starters, and they’re important. Based on my memory, my sense from these series is that both teams had good-to-great starters. I don’t recall too many times when one team’s starters were way better than the others, where the starters were the deciding factor.

    There’s another thing. I mentioned earlier about the psychological effects on a team from a good-to-great closer. I don’t have any statistics, but I think a closer could help the performance of starters (and hitters). If you have a Rivera (not to mention a bullpen like Nelson/Stanton), you know you don’t need a big lead. Heck, a one run lead is virtually a win. That has to be a huge psychological advantage for your starters. The closer can give that to your team and at the same time create tremendous pressure for the other team. There’s pressure for them to get a lead before the last inning or else their chances of winning will be slim to none.

  18. Reid


    I would bet most managers really hate to lose, and I don’t think becoming surly after a loss and taking it out on the team is totally out of the ordinary. More to the point, I wouldn’t characterize that behavior as managing regular season games like a post-season one.

    Also, I wouldn’t identify Martin’s “yelling, praising, howling, directing, etc.” as a successful post-season approach. It’s just his personality, and it works for some managers and not for others. I can’t see some managers switching to a Martin’s raving in the playoffs when they were basically calm during the regular season.

    As for Torre and Cox, are you sure they adopted the more “calm” approach because they were confident that their guys would come through? I can think of other plausible reasons for this approach: it was not part of their personality; they didn’t feel like ranting could work over a long period of time. (The latter would be one of my reasons for resorting to that approach). Related to the ranting is the idea that the idea of using emotion. Emotion can be double-edged. If you get your players really amped, there’s usually a period when they have to come down from that. John Wooden claimed he didn’t want his players to be too emotionally charged for this reason.

  19. Marc

    I reviewed the Oakland closers in an earlier post. I would say that with the exception of 2004, the Oakland closer has at least been good, some of them great.

    I guess I’m not being real clear. I have said several times that Beane thinks that the closer is important but I never said that I do as well (I do). I firmly believe that the closer is essential, I just happen to believe that it’s harder to find four good starters than it is to find one good closer. In fact, I think it’s harder to find an one ace starter than one great reliever. I even think it’s harder to find great position players than it is to find one closer. So my take: find a great closer but prioritize starting pitching and certain position players first. I suspect that Beane thinks this way too.

    Often, the team with the best closer isn’t the team that wins the world series (see Mariano Rivera, hasn’t won since 2000, has been on the losing world series team twice when he was clearly the better closer).

    Who was the dominant closer in the early 2000’s? Eric Gagne. How many playoff series have the Dodgers won in the last 10 years? Zero.

    Who was the dominant closer last year? Probably JJ Putz of the Mariners (fireman of the year). The Mariners missed the playoffs last year and are dreadful this year.

    How many innings does a good starter pitch in a season? – the standard is 200 innings.
    How many innings does a good closer pitch in a season? – probably 75 (Rivera’s career high as a closer is 80.2 innings).

    Consider this year’s AL East and the three contending teams. I don’t think I’m going out on a limb to say this:
    The Yankees starting rotation is a shambles but they have Rivera. Will most likely miss the playoffs.
    The Red Sox have a great rotation with an ace (Josh Beckett) and Papelbon although their setup crew is a litte shaky. Will probably make the playoffs and are a serious threat to win the World Series.
    The Rays have a very good rotation with serious question marks about their closer. Much better chance of making the playoffs than the Yankees, their closer situation may keep them from winning the world series.
    -My take: The Rays are in a much better situation than the Yankees. They may get lucky and find out that their setup stud – Grant Balfour (which really is a bad name for a pitcher) is the closer that they need. The Yankees are stuck TEN GAMES BEHIND THE RAYS with a rotation of Mussina, Pettite, and the three fake stooges and it’s unlikely that they will find the real Larry, Curly, or Moe any time soon. Rivera is still awesome but his impact is minimal.

    To close, let me restate clearly what I think about closers: The closer is a very important part of the team but it’s harder to find great starters (and great batters).

  20. Marc

    And regarding World Series starting rotations and one side being clearly better, this is arguable but real quick:

    The Red Sox rotation was clearly better than the Rockies rotation in 2007 and the Cardinals rotation in 2004. You don’t have to take my word for it…

    The White Sox rotation was clearly better than the Astros rotation in 2005. Basically, the White Sox had the best staff in baseball that year and both Clemens and Pettite well past their primes for the Astros.

    The Diamondback twosome of Johnson and Schilling was basically unbeatable in 2001. They were the two best starters in baseball that year. Didn’t matter who the other team was, the
    D-backs clearly had the better rotation that year…

  21. Marc

    Oh, and one last note… sorry….

    The Boston Red Sox have the advantage of spending huge bucks and they apply *Moneyball* principles to building their team. Billy Beane is probably really envious of this:

    Red Sox salary 2008: $140 million.
    Closer (one of the best today): Jonathan Papelbon.
    Papelbon’s 2008 salary: $750,000.

  22. Reid

    Again, based on the quotes I’ve mentioned I don’t know how you draw the conclusion that Beane really values a closer. But whatever. I want to make clear that I’m not saying that having a good-to-great closer, by itself, guarantees success. You can have the best closer of all time, but if the team is shaky in starting pitching, hitting or defense, you probably won’t be world series champs. Based on Lewis’ characterization I think Beane under-values the reliever.

    What a team should go after first–starting pictures, good hitters, solid fielders, etc.–I’m not sure. I was just trying to suggest that a good-to-great closer is almost essential to winning it all. They may not be as essential to be successful in the regular season, but I sense that they are crucial in the post-season.

    As for starting pitchers, if we were to look at the last 20 world series or longer, you feel pretty confident that the teams weren’t closely matched in terms of starters? Because my impression is that if we were to look at them the starters, you could make a case that many of them were pretty close. I’m too lazy to look at all the different teams and their starters, but I checked real quick on the Yankee lineup when they played the D-backs. OK, I’ll give you Schilling and Johnson for the D-Backs, but do those two alone make them entire D-Back starting rotation better than Clemens, Mussina, El Duque and Pettite? I think you could argue that the Yankees’ starting pitching as a whole was pretty equal to the D-Backs. And that series was super close. The hit by Gonzo off of Rivera was a freak hit.

  23. Marc

    In 2001, since the D-Backs had the ability to pitch Schilling and Johnson FIVE TIMES in seven games, I believe that those two pitchers by themselves gave them a rotation advantage over any team that they would have faced. Consider the 2001 stats below for the pitchers between the two teams, I don’t know how you can come to any different conclusion other than the one that Johnson and Schilling were the two best pitchers in the series. Yes, the series was close because the Yankees were better in the bullpen and had a better lineup, but in the end, those two pitchers won the series for the D-Backs:

    Record Innings Hits SO Walks ERA
    Johnson: 21-6 249 180 372 71 2.64 Won NL Cy Young
    Schilling 22-6 256 237 293 39 2.98 2nd Place NL Cy Young

    Clemens 20-3 220 205 213 72 3.51 Won AL Cy Young
    Mussina 17-11 216 208 182 48 3.15
    Pettite 15-10 201 224 164 41 3.99
    Pettite 4-7 95 90 77 42 4.85

    Recall also that yes, Gonzo blooped a hit to drive in the winning run of a tie game AFTER three of the previous four batters reached base and tied the game against Rivera (blown save), AFTER Schilling (three days rest) outpitched Clemens (four days rest), and AFTER Johnson came in to finish the game on NO DAYS REST (started the night before and pitched 7 innings). The Yankees probably had the better team, The D-backs had Johnson and Schilling. The D-backs won.

    And for my hopefully last word on Beane and closers, what I’m trying to say is that I believe he wants to spend money on other positions first because a quality closer can almost always be found cheaper. I say that he assigns them a lower salary value, you say that he undervalues. Not quite the same but close.

  24. Mitchell

    I’m am soooooooo soooooooooo soooooooooo unconvinced about the need for a dominant closer in order to win a championship, which I was sorta trying to communicate the night we had sushi together.

    Mmmm. Sushi.

    Marc actually posts a list of closers of Series-winners, and Reid’s response is

    Why do you say that having a good closer was not a requirement? In your list, all but one of the winners (’01 Diamondbacks, and I don’t know how good or bad Kim was that year) had a good-to-great closer, at least for that year.

    I would like to point out that Reid’s first argument was that you needed a “very good to great” closer in order to win the series. I think there is a huge difference between good and very good, don’t you? ‘Cause if we’re talking good, then yeah: I think Byung-Hyun Kim and Robb Nen qualify, but those guys were quite inconsistent in the years their teams won. We’re talking Rex Grossman inconsistent. If they had it when they went out there, they could be lights-out. But the problem was that you didn’t know until they started throwing. You didn’t have that problem with Rivera or Hoffman. Ugueth Urbina was the same way, only more unpredictable. I played a lot of fantasy baseball in those years, and saves were tough to come by; there was ALWAYS an early run on closers in our drafts, because the difference between “great” and “very good” was huge, and the dropoff to “good” usually meant you hated closers all year.

    Billy Taylor was a great case. He inspired none of that Goose Gossage fear in batters, but he went out there and got three outs. If we’re talking about how a dominant closer gives a team a psychological edge, Billy Taylor wouldn’t qualify. All he did was trot out to the mound and close the door, but opponents were never intimidated by him. Bob Wickman, for a few years, was the same way.

    The evidence that Beane is right about closers is that he kept finding closers where people didn’t think closers existed. He got bargain-basement saves, and because the league over-values closers, he was able to deal those guys for players he considered more valuable. You might place a super-duper-duper high value on closers, but if someone wants to offer you what might amount to three times the closer’s worth, you have to take it. And that’s what Beane does.

    Foulke, Isringhausen, and Taylor were basically nobodies by the time the Athletics made them closers, and I don’t know the numbers, but I’ll bet they got more saves per buck than any other closers in their time. That’s getting value for your players, which Beane does. His success with overlooked players like these means that he doesn’t undervalue closers at all: He values them exactly right.

    Marc said,

    Regarding the M’s, if they had not drafted Griffey or A-Rod, Seattle would not have a team today. Fact. See the 1995 season where they were set to leave if they didn’t get a stadium deal.

    Okay, perhaps that’s good for the city of Seattle (and I wouldn’t even say that: Now they have a sucky team AND an expensive stadium), but if we measure success by post-season trips (or by World Series victories), how good for Seattle were these draftees? I’m not saying they didn’t get two awesome players, because they did, but did drafting high-school studs give the Ms any advantage over the As, who drafted college players? There’s really no way to tell. In fact, I’d say that the real heroes of that franchise for the past decade and a half were trade acquisitions: Jay Buhner and Randy Johnson. I’m not sure how Omar Vizquel found himself on the Ms, but I’d put him third.

    Man, that was a fun team to watch.

  25. Mitchell

    Oops. I forgot about Ichiro. I guess you have to put him first, then Buhner, then Johnson.

  26. Marc

    Well, obviously the M’s haven’t won any world series games and neither A-Rod nor Griffey has ever been to one, but that isn’t for their lack of talent or effort. But I’m very familiar with the M’s history. And the fact is that the M’s were run like a small budget team through the 1990’s because they had a tightwad owner. They relied on the draft and trades to build their 1995 team.

    Omar Vizquel was signed by the M’s (foreign player) and then traded to the Indians in 1994, before the M’s ever made the playoffs. Don’t forget about Edgar Martinez, also signed by the M’s (foreign player) who was the premier DH for about 10 years and may have been the best right handed hitter in the 1990’s along with Frank Thomas and Juan Gonzales. Also, don’t forget about Tino Martinez (drafted by the M’s) or Jeff Nelson (drafted by the M’s) who starred for the M’s in 1995 before stupidly being traded to the Yankees and being an essential part of all four Yankees teams that won four out of five world series. Trust me, the Yankee fans can talk all they want about the Jay Buhner trade but they made up the difference in the Martinez/Nelson trade. Without the players that the M’s drafted, the Yankees don’t win the WS.

    The M’s made it to the playoffs in 1995, 1997, 2000, and 2001. They made it to the ALCS in 1995, 2000, and 2001 (which incidentally, were the years in which they had a quality closer, maybe some points for Reid there). However, after Johnson, the 1995 and 1997 teams didn’t have a good rotation and the 1997 team didn’t have a bullpen. They were all offense.

    Griffey was the best player in the AL in the 1990s when you combined offense with defense. A-Rod is probably the best player in the AL in the 2000’s although his defensive value slipped when he moved from SS to 3B. The question I asked was whether Billy Beane would have passed these guys up if he had the chance to draft them. Nobody can know for sure, but I suspect he would have drafted them because they really started making big bucks AFTER they became stars on the likely hall-of-fame level, not before.

    The M’s acquired Randy Johnson in a trade for Mark Langston, a classic Beane style trade of a draft pick who became a star for a young, unproven pitcher (sell high). Johnson led the starters through 1998 when they traded him for several pitchers, among them Freddy Garcia who was their best pitcher in 2000 and 2001 (Garcia beat the Yankees twice in the 2000 ALCS, the last year the Yankees won the world series and then had the best ERA in the AL in 2001). Even if you wish to say that they acquired these guys in trades (which Beane does all the time), this all came about because they drafted Mark Langston out of high school in 1985 and those fruits were borne out for the next 17 years.

    The M’s traded Griffey for Mike Cameron in 2000. Cameron is not at Griffey’s level offensively but he was Griffey’s equal or superior defensively after the trade. They used the salary difference between Griffey and Cameron to sign John Olerud and Kasuhiro Sasaki.

    The M’s lost A-Rod for the stunning amount of $250,000,000 in 2001. They used the money they didn’t pay A-Rod to sign Ichiro and Bret Boone.

    The 2001 M’s won 116 games. I think you would have to look at them as well managed through the late 1990’s to the early 2000’s.

    Just to make sure I hear you correctly Mitchell, are you seriously saying that the M’s made a mistake in drafting Griffey and A-Rod and then trading Griffey and using A-Rod’s salary in the way that they did?

  27. Marc

    Oh, and by the way, since 1995:

    M’s four playoff appearances, 3 ALCS (2 since 2000), no World Series.
    A’s five playoff appearance, 1 ALCS (1 since 2000), no World Series.

    I think Beane is brilliant in navigating his salary constraints. I think the Red Sox show what can happen if you use *Moneyball* priniciples and have money. I think the M’s had possibly the worst GM in the majors for the last five years (Bill Bavasi) and hopefully things will turn around soon. But if you use Mitchell’s bottom line of playoff appearances and world series wins, Beane and the A’s have not been any more successful than the M’s in the last 15 years and in the 2000’s have been to one fewer ALCS than the M’s. I would in fact argue that the M’s have been more successful over that period of time.

    What have you done for me lately? Neither the M’s nor the A’s will make the playoffs this year.

  28. Reid


    I don’t have much of a problem saying that Schilling and Johnson were the two best pitchers that year, but I don’t agree that the D-Backs starting pitching was decisively better than the Yankees just because they had those two guys. Schilling also didn’t outpitch Clemens in that game. The stats are similar, but Schilling gave up 2 ER to Clemens’ 1. You talk about the importance of starting pitchers (and I’m not saying they’re not important), but the series came down to Rivera. If Rivera pitches like his usual unstoppable self, the Yanks are champions. (Rivera was key in the other three Yankees wins, too; getting the win in one game, a save in ther other, and preserving a tie and later being pulled in extra innings.) But this discussion centers on one series, and we’d have to look at bunch of them to see the impact of the closers.


    I’m sure there is a difference between good and very good closers, but I don’t know if I can make a distinguish them in a meaningful way. My point is that having a solid (whatever that means) is critical, partly because it gives your team a psychological advantage. (Closers that are very inconsistent probably wouldn’t qualify as a very good or solid.) Let’s hear your case for why you don’t think closers are crucial.

    Oh, I wanted to add my two cents about the impact of Griffey and A-Rod. I’m not sure what you’re asking. Are you asking if Griffey/A-Rod helped the A’s in the post-season than the college picks helped the A’s in the post-season? If so, why were you asking that question, again?

  29. Marc

    Ok, I’ll agree to disagree on decisively better although I guess it depends on how you look at things. My bad on Clemens who pitched better than I remembered, but here are the composite stats for Johnson/Schilling in their five games versus the seven games for the Yankees starters.

    J/S (4-0): 5 gms started/1 relieved, 40 innings pitched, 45k, 6 earned runs, ERA 1.15
    Yankees (1-3): 7 games started, 41 innings pitched, 43k, 18 earned runs, ERA 3.95

    The D-back duo started 5 games and gave up 6 runs. Amazing. Dominant. To my mind decisively better than the Yankees starters. Yes, this is a small sample but that’s the point with a short series, you have a small sample. We talk about how a dominant bullpen gives a team a big advantage in the late innings? I think the theory is similar here: if you have two starters who are dominant then your team would expect to have a big advantage in a short seven-game series.

    I’m done on that topic. We can agree to disagree. But I think this opens up another question about pitching on short rest. Schilling’s bloody sock games with the Red Sox are now legendary but I think you need to give him even more credit for the job he did in this series. He pitched on three days rest twice, in game 4 and in game 7. And he was dominant. Yes, the guy is obnoxious but man did he bring it in big spots.

  30. Marc

    I now understand my deep seated need to have the last word on this blog topic. See below, someone help me!!

  31. Mitchell

    Why would I be asking if Griffey and A-Rod helped the Athletics in the post-season? I’m baffled by your question. The original point, made by Marc, was that avoiding high-schoolers in the draft as a matter of policy may not be a good idea, because it keeps you away from guys like Rodriguez and Griffey. My response was that the team that did draft those guys didn’t do any better with them, and they probably had to wait longer for those players’ development. Marc countered with the argument that those guys saved baseball in Seattle, but I said that saving baseball in Seattle doesn’t mean success for the team; it means success for Seattle. Marc then offered post-season numbers that (in my view) prove my point: drafting high-schoolers didn’t make the Ms any more successful than the Athletics.

    As for closers, “having a solid,” whatever that means, might be a psychological advantage and it might not. Billy Taylor, when he was the Athletics closer, saved 33 games and had an ERA of 3.58. He walked 22 and struck out 58 in 73 innings. Those are hardly the kind of numbers that give anyone a psychological advantage, but they were good, solid numbers and Taylor was a reliable closer.

    He was traded to the Mets the next season for Greg McMichael and Jason Isringhausen. This is evidence of Beane’s approach to closers. Taylor didn’t make his debut in the majors until 14 years after he was drafted (drafted in 1980 when we were eleven years old and debuted in 1994 when we were in our mid-twenties!). Those 90 saves he earned the As cost the team almost nothing, and then the As traded him for two players nobody wanted and turned one of them into the team’s next closer. Brilliant.

    I’ll get into closers in general later. This is a crazy week for me and I can’t access VI from work.

  32. Reid


    Did you really think I meant Griffey and A-Rod helped the “A’s” and not the M’s? It was a typo, man. As for Griffey and A-Rod not helping the team do any better with them, what’s your basis for that? I think determining if two players–by themselves–made the team better and to what extent is very difficult. I think looking at their production would be a fairer way to gauge the picks. They’re both hall-of-fame players, so if you’re trying to determine whether avoiding high school players is a good policy just by looking at those two picks, you gotta say it’s not a good policy.

    As for closers, I’m not clear on precisely what would make a closer good enough to have the impact I’m talking about. But I do believe that getting a closer that can almost guarentee a lead in the last inning is very valuable in the playoffs. I’ll wait until you have time to lay out your argument before I comment further.

  33. Marc

    Ok, I’ll concede the point that the M’s and the A’s have had roughly the same success in terms of post-season appearances since 1995.

    I realize that long posts get tiresome and hard to read. These are the two points that I thought I got from Mitchell’s original post that I disagree with and only because he specifically mentioned the M’s and Griffey/A-Rod.

    1. The Mariners only did “ok” by drafting Griffey and A-Rod. I disagree but it seems we have differing views on what it means to be successful. That’s fine. However, there is also the unsaid implication that the M’s should have drafted differently. Maybe taken college players who would have been cheaper? If Mitchell is seriously saying that the M’s should have passed on Griffey and A-Rod because they would have been better off drafting college players those years than I think he’s off his rocker.

    2. The next implication is that Beane and the A’s would passed on Griffey or A-Rod if given the chance. This is unknowable but I also think is wrong for these reason: Beane has in fact drafted high school players (see Eric Chavez) if they can get to the majors quickly and then be traded if their salary demands get too high. He views players as commodities with specific values. Can you imagine what types of deals he may have spun with hall of fame players like Griffey and A-Rod?

  34. Mitchell

    Man, I’m having a hard time posting this comment for some reason, and I’m logged in as admin.

    I’m not saying that the Ms shouldn’t have drafted Griffey and A-Rod; I’m saying that NOT drafting them didn’t result in their being less successful than the teams that DID draft them. A-Rod and Miguel Tejada actually began on the Ms and As in the same season (A-Rod’s first full season, anyway) and while that’s a huge drop-off in talent, I don’t think it’s much of a drop-off in production, at least not during those years. A million things make a team successful; it seems to me that the idea of taking college players, especially pitchers, is sound.

    Also, sorry about the As/Ms thing. I honestly didn’t know what you meant, and it didn’t occur to me that it was a typo. I really thought you didn’t get what Marc and I were discussing, which is why I outlined the entire thing again.

    As for closers, let me just say that if you have a Mariano Rivera or a Dennis Eckersley, of course your team is going to rock. But for some reason, while amazing starters with great longevity make their debuts every year, the same cannot be said about closers. Closers seem to shoot to stardom quickly (often out of nowhere) and then quickly burn out and become nobodies. John Rocker, anyone? And that’s just for starters.

    With this in mind, Beane’s idea of grabbing an undervalued guy to do the job and then trading him away before he flames out is brilliant, and I can’t believe more teams haven’t figured that out.

    I’m with guys like Bill james, who say that the save statistic is overrated and that the closer is an overrated position. Rather that save your relief ace for the ninth inning, you should be putting him in for a critical inning when the other team is most likely to do damage. If Magglio Ordonez, Miguel Cabrera, Carlos Guillen, and Gary Sheffield are due up in the eighth, give the freaking ball to Rivera THEN, and let someone else throw to Matt Joyce, Edgar Renteria, and Brandon Inge in the ninth. In fact, depending on how your middle relievers have been doing, I’d even bring Rivera in to face those guys in the seventh, then someone else to pitch the eighth, and hope someone else can face the top of the Tigers’ lineup in the ninth. In the ninth inning, I think there’s more pressure on the hitter than on the pitcher, since even mediocre pitchers have a 70% chance of getting the guys out in the ninth.

    Additionally, in this era when stats can be broken down into specific kinds of matchups, I think it makes more sense to pitch guys on the mound in situations where they are most likely to do their best. I’m not saying that if you have a Rivera in the pen, you shouldn’t give him every opportunity to shut the door in close games, but if you don’t have a Rivera in the pen, you should still be able to win games when you already have a three-game lead by throwing other guys in the ninth inning based on matchups.

    This is not a glamorous way of shutting the door, but I think it can be done. First, establish a bullpen ace and throw him in the most important of situations; don’t save him for the ninth inning. Second, get guys who can mess with a hitter’s rhythm, and before the hitters get too familiar, put someone else in there who can change things up. Chad Bradford, a side-armer, and Tim Wakefield, a knuckleballer, have been great this way. If you can get ground-ball specialists to pitch to the sluggers and fly-ball guys to pitch to the speedsters, that’s cool, too, but most bullpens can’t really get that deep.

    I just remain unconvinced that a dominating, reliable closer is necessary, especially since they are so very rare in this game.

  35. Marc

    Ok Mitchell, I’m done with the Jr/A-Rod draft debate and would have been done with those players entirely until you somehow decided to make an A-Rod and Tejada comparison. I can only presume that you are making assumptions without looking at the stats because there is no statistical comparison between the two players and never has been. A-Rod has always clearly been the player with more EVERYTHING. Especially if you limit the comparison to 1996-2000 when A-Rod was with the M’s and Tejada was breaking in.

    This is a good time to bring in a new concept created by Bill James, whom Mitchell (and I) clearly admire. The concept is called “win share” where James took a whole bunch of metrics, including offense and defense, and figured out how to measure a player’s production and contribution to the teams record that year. Obviously I can’t explain the whole thing in this space but there’s a pretty good explanation on Wikipedia. I bring this up because Mitchell mentioned production and A-Rod blows Tejada out of the water. According to James, each win share is worth 1/3 of a win.

    Arod averages 12 win shares a year more than Tejada, which means that statistically he is worth four more wins a year than Tejada over their careers. That’s huge.

    Is Arod worth his $28 million dollar salary? Probably not. But he provides more bang for the buck than Tejada, believe it or not. Tejada’s salary is $14 million, about half of Arod’s. in 2007, Arod had 39 win shares, Tejada 15. So far in 2008 Arod’s contributed 19 win shares, Tejada 8.

    And Tejada’s career year of 2002 when he won the AL MVP? Arod had 23 more homers, 11 more RBI’s, scored 11 more runs, and won the gold glove. Arod was merely a better hitter and defender, Tejada was merely the MVP. Go figure.

    If you’re going to rely on stats, you cannot deny Arod’s clear superiority. And stats are at the heart of *Moneyball*.

  36. Mitchell

    …But Tejada was on my fantasy team, and I trounced the team that had A-Rod.

  37. Marc

    Mitchell, arrgh, sorry, there is still one last thing about the draft that I’m waiting for an answer to. Then I’m done. Promise. Sort of.

    Do you really think that Beane would have bypassed Jr or Arod in the draft those two years?

    And I bet you scored a lot of fantasy points with your CLOSERS!!

  38. Mitchell

    I honestly have no idea if Beane would have drafted those guys. I have to say, I wasn’t impressed with A-Rod at all when he was a nineteen-year-old. That changed, of course. And of course I scored a ton of points with my closers: They are as overvalued (if not more overvalued) in fantasy baseball as in real baseball. 🙂

  39. Marc

    Fair enough Mitchell, let’s try this: if you were a GM of any team and you had the first pick of the draft in 1987 and 1993, would YOU have bypassed Jr or Arod in favor of a college player?

    Just to save you a little time and even give you the advantage of hindsight, there was one other probable hall of famer in the 87 draft besides Griffey, a college catcher named Craig Biggio. There were of course other great players that year but those two were the cream of the crop. As for 1993, there were some great players (Billy Wagner – college, Chris Carpenter – high school, Jason Varitek – college, Derrick Lee – high school, Torii Hunter – high school) but only one probable hall of famer.

  40. Mitchell

    With hindsight, of course I would have. But let me as YOU: How reliable are scouting reports, especially of high-school athletes? You know, Colt Brennan has put up some decent numbers in his pre-season games with the Redskins, but he’s playing against third- and fourth-stringers, guys who probably won’t be on NFL rosters come September. Watching high-schoolers tear up high-school pitching and deciding they are future hall-of-famers is ridiculous. Of two best high-school baseball players I’ve seen (and granted; I only really watched much high school baseball for one year while we were in high school), one pitched fairly well for USC but his career ended there, while the other didn’t play more than one season at Santa Clara. Of course, I’m no Major League scout.

    By the way: I think Biggio belongs in the Hall, but it’s pretty tough for a catcher OR a second-baseman to get into Cooperstown. He played for a pretty low-profile team and did it in a steady, non-attention-getting way. I hope the writers were paying attention, because I have a feeling he doesn’t get in.

  41. Marc

    You would have what? Bypassed or drafted those two?

    I think that scouting high school players works pretty well for position players but not pitchers. Here are the other top high school position players who were drafted #1 overall since 1987: Justin Upton, Matthew Bush, Delmon Young, Joe Mauer, Adrian Gonzalez, Josh Hamilton, Chipper Jones, Phil Nevin. The only one I’ve never heard of is Matthew Bush, most of the others play at the all-star level and Chipper is a probable hall-of-famer. The results are much worse for high school pitchers, which I think is one of the big points in *Moneyball*.

    Biggio has 3000 hits as a catcher/2B. Every player who has reached 3000 hits regardless of position is in the hall of fame (except Pete Rose, and so far, Rafael Palmeiro) and catcher/2B are generally weaker in terms of career offensive numbers. I think he gets there. The question for Astros fans will be if Bagwell gets in since his career numbers don’t stand out as much when you compare him to other 1B.

  42. Reid

    Mitchell said,

    I’m saying that NOT drafting them didn’t result in their being less successful than the teams that DID draft them.

    OK, but that doesn’t mean passing over them was a good decision. Maybe the A’s would have been a lot more successful if they had drafted them. To me, the question is sort of strange and seems inappropriate. We’re trying to judge the policy to forgo high school players or not and using the specific example of two guarentee hall-of-famers is not a good way to make case for the A’s.

    On to closers…let me address your points one by one:

    Whether closers burn out quickly or not is not really relevant to their value in the playoffs. (It is relevant to how you manage your payroll and the future of your team, but that is a separate issue.) A team just needs a dominant and reliable closer for a particular year. Staying with one closer like this or getting different ones like this every so often is not so crucial.

    Your next point seems to say that the pitcher normally designated for the closer role, shouldn’t be used at the end of game but near the end whenever the opposing line-up is the toughest or in the most critical moment. So if I understand your argument, you’re saying using this pitcher at the end of the game is overrated or inappropriate, hence the closer is overrated. That’s an interesting argument, and there might be merit to that idea. Let me give you an explanation as to why I think this may not be a wise move–and thus show the value of closer.

    First, I think you can psychologically hurt your dominant reliever by using him in the way you suggest–if your revolving group of relievers are inconsistent. If using a pitcher like Rivera as a middle reliever doesn’t consistenly lead to wins (I would also add that the team should feel confident victory will be all but secured when the relievers come in after him), you can hurt him psychologically and ultimately weaken his performance. Here’s the kind of the thing that can be going on in his mind: “Man! What does it matter if I preserve the lead since the guys after me have a good chance of blowing it?!” The only other example that comes to mind is when you have a great quarterback throwing to receivers who frequently drop the ball. The QB has to be psychologically resilient to not let a sense of helplessness creep in and affect his performance. Now, you might say that is what starters and middle-relievers deal with. Right, which is the reason having a your dominant closer is so critical. The dominant closer can make the starters and relievers have a more powerful incentive to preserve the lead because they know if they do, they’re all but assured of a victory. In this case the middle-relievers would be thinking something like, “I just need to preserve the lead for these two or three innings and we win.” The starters–if they have great closer and middle relief are thinking, “I can keep the lead after the 7th and we win.” This is a huge boost to the morale and psychology. Not having a dominant (which includes consistency) can have the opposite effect.

    Now in the playoffs–where the critical difference between the regular season is the pressure–this kind of psychological difference is super significant. You frequently hear that the difference between major league players (or professional players in general) is mainly mental, because at that level all the players have ability. I would say that is even more true in the post-season. The diference is often mental the teams have to deal with pressure.

    Second, I think there is good reason to believe that the rotating relievers will not be consistent. You need someone with the right pychological make-up to close out a game, which I’m assuming is not easy to find. He has to be able to take pressure for one thing. Furthermore, if you have a designated closer–versus a rotating one–that person can psychologically prepare for that particular role; the fact that he is regularly in that position will give him a chance to become accustomed to that situation giving him a greater chance of success. I believe the reliever pitching in the last inning(s) needs a different psychological make-up than the middle relievers (or starters). With rotating closers, you’re expecting several pitchers to have the right psychological make-up and you’re not allowing them to prepare or become used to that role. The fact that these pitchers will never be certain if they’re going to close out a game or not (since the manager will decide based on the line-up and moves of the opponent) is a huge psychological disadvantage and lessens the chance of success, imo. Another way of looking at this is by understanding the importance of players knowing their roles. This is true in any sport, I think. Very talented players can perform poorly if they don’t understand their roles or if they’re not able to get comfortable in a particular role (which can happen if the role is not clearly defined or constantly changing). An important part of any coach’s job is to clearly define these roles and make sure players are in roles where they can excel and help the team. A closer by committee doesn’t allow players to get comfortable in their role.

    There’s another variable that can lead to uncertainty in the pitchers and overall team as well and that is the judgment of the manager. Part of the success of the approach you’re suggesting depends on finding the right match-ups, making the right decisions of when to pull players, etc. If the manager is not really reliable or if the team starts feeling like many of the loses are due to the manager, not the pitchers, than that can add another level of uncertainty about the outcome of the game. When you have a closer, the manager’s decision is taken out o the picture: in the 8th or 9th Rivera is coming in and pitching, no matter who’s up. This a kind of stability and certainty that I believe greatly helps a team.

    A lot of my position is based on the idea that making the victory certain in the last innings (if you have the lead) has a significant psychological cascading effect on the rest of the team.

  43. Marc

    I have the advantage of living in a Major League city (although it can be argued that the M’s don’t really qualify as an MLB team) that is very literate (Seattle reportedly has one of the highest per capita number of bookstores in the country – which is only one metric). I think our debate here is pretty cool but I would point out that there is a huge amount of really interesting baseball writing all over. I’d point interested folks to the blog below by the Seattle Times Mariners beat writer, Geoff Baker. Yes, this is a local blog written about a terrible team but there is really interesting stuff here and the topic for 8/29/09 deals with the analysis of so-called “sabermetric teams” this year. Of course, being Seattle, there are THREE different blogs devoted to the team. This one is the easiest for me to read, the others deal with more true statistical analyses.

    Of course, maybe nobody is interested since this involves baseball and specifically the M’s. But this can’t hurt right?

  44. Reid

    Since this thread seems to be dying out, I wanted to throw one last question, of a more philosophical nature, out there. As I read the book, the thought of two types of styles of management/coaching occured to me: one relies on instincts and human judgment while the other relies more on statistics and quantifiable data. The book seems to be about the battle between these two styles.

    Of course, the wise coaches uses the combination of the two, but finding the right balance is the trick. Each side has its limitations. With human judgment, we can put more importance on certain aspects of performances than they merit. We have biases from tradition (the way things have always been done), fears (doing something new that will be perceived as foolish), and emotions (protecting a player out of loyalty) that cloud our judgment. Quantifiable data doesn’t have these weaknesses (for the most part–afterall, people with the limitations above have to identify and retrieve the data), but they have blindspots as well. In other words, quantifiable data, by itself, can’t give you a complete picture of a player or team.

    My question is do you think it is possible that one day quantifiable data (like statistics) will be able to create a complete picture–to the extent that a coach could rely almost exclusively on this data rather than human judgment?

  45. Marc

    I think that there are too many variables to quantify and too many unknowns in baseball to ever rely completely on data without judgement. If you have a tightly controlled set of players who act exactly the way that you intend them to each time, you know exacty what the playing conditions will be, you know exactly what the capabilites are of your opponent’s team and individual players, then I think you can rely on data almost exclusively, My best argument for this would be the game of chess where it was shown convincingly that a computer will often beat the grandmaster by analyzing data and making the best move. But since atheletes are much different than chess pieces I think that data canonly take you so far.

  46. Reid

    Yeah, I agree. I guess this is not as interesting a question as I thought. But I was thinking that while you build a powerful enough computer that could factor in all of the variables, but the real obstacle would be gathering a large enough sample size for that specific situation (which is can be very specific). Basing decisions on statistics means that you’re relying on past events to decide on situations in the present, so if you don’t have a large enough sample size the statistics are not very helpful.

You can add images to your comment by clicking here.