That's ridiculous. Every time a player has a streak, you're supposed to start them regardless of a more talented player in front of them? Think of what this standard implies. Hypothetically, if Bagwell gets cold, and Orlando Merced hits better over 30 at-bats, should we bench Bagwell? Of course not. The difference, of course, being that Bagwell has a 10 year history of success, while Ensberg has never played in the major leagues before. Play the more talented player, who has the capability to do better right now. The problem here is that we don't know who's more talented. You can talk about projections all you want, but minor league top prospects fail in the majors all the time, or in many cases take a year or two to become reasonably good. Ensberg's not even a "top prospect", really. This team needs to win <B>now</B>, not 2 years from now, and they'll play the players who give them the best shot of doing that now. Thus far, Ensberg has not been that player by any stretch of the imagination. We have a better option, so we're using him. Ensberg will get his chance. None of our rookies played "full-time" their first years and they seem to be OK (Ward, Berkman, Hidalgo), so permanently damaging his development is not a real issue. How does one define "right now?" yesterday? the last 20 at-bats? The last 50? The last 100? You don't define it. The manager watches the games and bases it on what he sees, instead of purely looking at statistics. What kind of swings were they each getting? How comfortable are they each at the plate? Is one of them pushing too much? Fortunately, baseball isn't purely statistical analysis. Otherwise, we wouldn't bother playing the games. Right now, we can see that playing Ensberg for 500 at-bats will probably help the team more than playing Blum for 500. That's all we need to know. No, we really cannot see this because MLE's are not, by any stretch, reliable for any given player. A thousand Ensbergs may be more likely to benefit a team than a thousand Blums, but that doesn't mean anything for this particular Ensberg. He may theoretically help the team more, but the bottom line is that in terms of reality, Blum has helped the team far more than Ensberg thus far, so he's earned more playing time. one player has upside. The other has none. Didn't everyone say that about Luis Gonzales too? Every reasonably young player has potential upside.
Ah, so now you drag it out! This contradicts your previous assertion that prior statistical performance (career history) does not matter. Now it's only the unproven rookie that doesn't get the benefit of this history. Problem is, there's no brightline here. Why does past history prove effective for evaluating Bagwell, but not Ensberg, when MLE's generally work out? And of course it matters that we're only 13 games into the season. Otherwise, you have to bite the bullet of the argument that any player should be benched after 13 poor games relative to any other. You keep on regurgitating the same flawed thinking. Now for the harder job... Ah, but we do know that MLE's generally work out. In fact, an MLE is compiled from past history. We know how hitters generally react from shifting from AAA or AA to the majors. Over an enormously large statistical sample, you can calibrate the averages. Now, if the standard deviation was very high, this would be useless. But it isn't. Generally, the MLE works out. Hence, if one believes past performance is indicative of future success, the same logic applies to both Bagwell and Ensberg. Once again, Ensberg's MLE calculation is superior to Blum's prior major league performance. I'll trust the MLE calculation of Ensberg's entire last year over 13 games. Statistics don't indicate what happens, they simply make predictions based on probabilities. And the problem with believing that Williams is capable of making that judgement, is that the teams that are moving more and more to a sabermetric interpretation of baseball are generally more successful. Think the Mets pitching coaches in the mid-90's knew what they were talking about? Or how about Dierker when he stuck in Mike Jackson over Dotel? The bottom line is, you either believe that stats can represent what goes on in baseball... or you don't. You can't accept that they do, but don't in certain situations. That's accepting the premise, then denying it once logic takes it to a certain point. The flaw in your interpretatin of my argument, is that your argument rests on the idea that I believe that stats are what is actually going on. That's not true. Stats are a model for what occurs on the field. Of course you have to play the games. Otherwise stats would not exist. Stats don't determine plays - plays determine stats. But stats are a wonderful analystic tool. And I'll trust them more than subjective analysis anyday. Vicious regress time again! Blum has helped the team more? Ok, then apply that universally. Next time Bagwell slumps, and Merced is streaking, bench Bagwell. And you have to go with probability. Of course, it's entirely possible that Ensberg will play worse than Blum. But as you admit, over 1000 hypothetical scenarios... that won't happen. Which is why you have to go with that... it's foolish to take the (random #) 30% odds over the 70% odds, just because of a ridiculously small sample size. That's like saying that because the Giants have won the only game they played without Bonds, and have a winning % of 1 without him... that they're better off without him. My apologies. I made an over-generalization. Let's put it this way: the odds that Ensberg has significant upside are much better than the odds that Blum does.
It doesn't contradict anything. I never asserted that prior performance doesn't matter. I asserted that Ensberg has very little prior performance and shouldn't get the nod based on upside alone. Right now, Blum is playing far better than Ensberg and he gives a struggling team a better chance to win, especially since there's no guarantee that Ensberg will succeed.
The bottom line is, you either believe that stats can represent what goes on in baseball... or you don't. You can't accept that they do, but don't in certain situations. That's ridiculous. Stats have their place, but they are not the be-all, end-all in baseball. Otherwise, we would be able to predict the winner of every division every year without playing the games. Stats are useful, but they are certainly not the only thing to look at. Simple question: Let's say that Blum was batting 25-25 this season and Ensberg was 0-25 thus far. Do you think Blum would deserve to keep playing over Ensberg? If yes, then I'd argue that your logic above is flawed, because now you're accepting that this small sample size *does* have relevance and overrides the MLE's and Blum's previous history. In that case, it's just a matter of degree -- you don't feel Blum is doing "enough" more than Ensberg to earn playing time right now. If no, well then we'll just have to disagree. There's no question in my mind that I would play Blum without thinking twice in that scenario.
I don't really want to jump into any argument here...but, oh well. Seems to me RM95 just threw out a statement not expecting someone to word for word attack him. His position is that Ensberg is a rookie who, despite what his MLE or stats suggest he will do, we only know what he has done so far this season versus what Blum has done. We also know that the Stros offense was struggling and the team seems to be perfroming more successfully (for whatever reasons) when Blum starts. IF it was Bagwell, it WOULD be different. It doesn't have to be some theory that works for every player or only one player. Who cares. It is just an opinion in this circumstance. I disagree. I can believe that stats represent what goes on in baseball, but not in all situations. Why? Cause I do. Take last year's WS. I haven't checked, but I'm sure that stats would say that the Diamondbacks closer couldn't blow two consecutive WS games in the same way. I doubt he had ever done that, I question how many times in his professional life he had ever blown two saves in a row and I wonder how often two saves had been blown by the same player in consecutive games by teh same closer in World Series history. Stats said to pitch Kim in that second game cause he couldn't possibly blow it again. As a manager, though, I would have never thrown him in. Just my subjective opinion (and a retrospective one, but I seriously did cringe when he came in again), one not backed at all by stats. But I still believe that the statistics of baseball can represent what happens.
This one's for you, Dylan From the Chron: "I'll be happy to get past the 100 at-bat point," [Ensberg] said, adding that through his minor league career he was notoriously a ".150 hitter in April" Here's the opinion of a guy who's seen Ensberg play a heck of a lot more than we have: http://tz.astrosconnection.com/msg_view.cfm?messageid=86843 the important points: "Ensberg is a verrah smart player, he isn't necessarily a talent-ladden kid like Hank Blalock that is going to step right out of bed in the morning and get you a couple of knocks that night. What is Morgan's forte is to get used to the new level of play he's in, whether it be a jump from A-Ball to AA, AA to AAA, AAA to the Dominican, AA to Venezuela and ultimately AAA to the majors. Once he gets comfortable and makes the proper adjustments, then he's a very good offensive force for the team he's playing on." and "So one can make all sorts of pronouncements of a players value based on small sample sizes and all, but the reality is you have to understand the player as well as the stats for context. ...With Morgan Ensberg, he has the "makeup" (not a term oft used in context of stats) to make adjustments. He's not a talent only, he's a ballplayer who is smart and uses that to his advantage. He rightly says that 100 ABs become a benchmark to see if improvement has happened and he now has an appropriate sample size to understand what he needs to do to improve and help this club, which in my opinion he will definitely do. So does this mean that Geoff Blum is not a starter for now? No, that means that if a platoon situation helps Ensberg improve, so be it. Sitting on the bench won't help him in any case and having only one option at third is stupid (like last year). Unless that guy is Scott Rolen or Troy Glaus, you need good reserves/bench players, even platoon guys if you want. Heck, what was so wrong with 1998 and a platoon of Berry/Spiers? How about Garner/Walling? " Thanks, Noe.
I explain my position, and you ignore it. You're way too smart for that. Either you're being disingenuous, or you're not reading carefully. I specifically stated that stats do not dictate performance. Performance defines stats. Stats model nearly perfectly what has happened in the past. They sure as hell model it more perfectly than what a manager can subjectively remember accurately. Statistics can make predictions based on probabilities concerning what will happen. Judging by what goes on in the league, they at least seem do so better than those who judge subjectively. The organizations that seem to emphasize probabilities and statistics seem to have done better in recent times. The Padres and Athletics have fared well (with low payrolls), while the Royals and Brewers have faired poorly. Yes, yes, I know those are the extremes... but you typically hear the most statistically oriented comments coming from Towers and Beane... while the Brewers GM sometimes makes comments that have an actual contradiction in statistical knowledge. Anybody batting 25/25 is an inconceivable statistical fluke. And yes, I'd bench him. Why? Because over 500 at-bats, he's still probably not going to be as productive. On to Jay-Z Actually, it does. Or rather, at least, there has to be a reason for thinking something is true, n'est pas? Said reason does not change from instance to instance. If it does not, then it's universal and can be worded abstractly. If it does change, then it's arbitrary and not worth telling anybody about. "Cause I do." Riiiiight. So, we should accept how you feel, simply because you say so? Then please keep your opinion to yourself. If it can't be translated into a sensible, understandable comment, then you have no reason to inform anybody else of it. You believe something. I believe something else. If that's all there is to it, there's no way to ever reach true communication between us. So why bother? As for Kim... well, that's silly. It's almost impossible for statistics to represent what will happen in a very short span of time. Looking at single box score, or two in a row, tells you very little about the value of the players involved. And it's entirely possible that Kim was pitching scared the second time out. I'll readily admit there's a psychological element to baseball. But the psychological element is most important in a microcosm... and the larger sample size you take, the less important it becomes. (incidentally, I also think benching Ensberg was bad for psychological reasons, since I think lack of confidence and managerial support can really hurt a rookie)
The Padres and Athletics have fared well (with low payrolls), while the Royals and Brewers have faired poorly. Yes, yes, I know those are the extremes... but you typically hear the most statistically oriented comments coming from Towers and Beane... while the Brewers GM sometimes makes comments that have an actual contradiction in statistical knowledge. This could be just because the Brewers and Royals GM's are just dumb in general. I agree that stats will predict long-term better than subjective analysis. However, I believe subjective analysis is a better short-term predictor. As someone else mentioned in the Byung-Byung Kim thing, stats aren't good for considering things like confidence and slumps. In that respect, I trust a manager's judgement more in the near-term than using stats. Blum's been performing better recently, and if Williams sees something that makes him want to keep him starting for a few more games, that's fine with me. We're not talking about a season-long benching here. Anybody batting 25/25 is an inconceivable statistical fluke. And yes, I'd bench him. Why? Because over 500 at-bats, he's still probably not going to be as productive. That's where we disagree. While you say it's just a fluke, I'd say he's seeing the ball well, hitting the ball well, as well as getting lucky. I'd continue to start him until he tapers back off to sucking (if he doesn't, even better). At that point, I'd give Ensberg his shot again. But as long as he continues to go 3-3 or 4-4 every game, I'll keep him in there. (incidentally, I also think benching Ensberg was bad for psychological reasons, since I think lack of confidence and managerial support can really hurt a rookie) I disagree here. Players go through hell to get to the majors -- especially a player like Ensberg who wasn't an instant phenom. If they can't deal with a week off, they don't have the mental makeup of a long-term player, in my opinion.
Thank god this discussion is winding to a close. I hate arguing with 3 people I consider smart all at once. Could be, but I'd like to return to a point I mentioned earlier. How do you know when someone's streaking and slumping? Is it over 10 bats? 50 at-bats? 100? 20 is too few to even prove a streak. That's well within the province of luck. I have trouble trusting managers, simply because we're learning that so much of what managers have clung to is simply not true, if you trust statistics at all. My guess is, you can even explain why pretty simply. Managers learn through feedback mechanisms. Feedback mechanisms can work, but they're also notoriouis for encouraging superstitions. He remembers what's dramatic. That's just human nature. For instance, you're more likely to remember a 9th inning homer than a 1st, even though the difference in importance is probably non-existent, and at best, marginal. Managers, like any other human being, remember stuff like that. Stats don't. Stats are as close as you can get to objective. They cut through all the crap. Incidentally, did you read Joe Morgan's column today on ESPN.com? It was the biggest piece of drivel I've ever read. He actually argued that OBP is less important than speed for a lead-off hitter. Now, let's look at this. Jeremy Giambi is a good example of a lead-off hitter with high OBP and terrible speed. Jimmy Rollins is a good example of a lead-off hitter with terrible OBP and good speed. If both played SS, who would you rather have lead-off on your team? My guess is, Giambi. The problems with Morgan's thinking are myriad. First, a lead-off hitter is only guaranteed 1 inning in which he bats first. Secondly, OBP is much more important for a hitter, and you want the better hitter to have more chances, which is likely if he bats ahead in the order. Then, of course, you have the fact that if someone gets on base a good 10% more often... the occasional stolen base is going to be much less important. Especially in an era of the long ball (less true in less homer-happy eras). Conserving outs is the #1 task of any hitter. And guys with high OBPs do that better than fast guys. Morgan, in some ways, typifies managerial thinking. And that's really fricking scary. Most managers were also more afraid of Sosa than McGwire in 98. That's pretty scary as well. (hundreds of other examples, as well)
haven...what is your fascination with MLE? I bet if you asked Jimy Williams whose MLE was better, he would answer "how should I know what kind of car those guys drive?"...if managing was based on computer projections, we could hire a Pentium processor to print out the lineup card everyday. As Cassel stated earlier, Williams manages on hunches, and he constantly adjusts. Maybe he thought a few days off would clear the rookie's head. Maybe he had a hunch Blum was on a tear and deserved the at bats. Whatever his reason, I'm sure it didn't come from anywhere but a gut feeling, which is where all great managers find the answers.
haven= why the Jimy hatin? He is a good manager. Managing real baseball isnt like a computer game. There are certain circumstances which go beyond stats and talent. Ensberg isnt going to suffer from not starting a few games. Lineup tinkering can be useful...
Oh. My. God. Why do you even bother trying to argue? Let's do this slowly. First of all, Williams was not popular among players in Boston. How do I know this? Because I ****ing live in Boston. Boston players hated him. They constantly b****ed and moaned (anot not just everett) about the constantly changing line-ups. Several players insisted that the inconsistency negatively affected their performance. So, I don't see any reason to believe that he somehow "judges" correctly. Secondly, very, very, very few managers do many things on "hunches." Managers may look qualitatively at things. But I can absolutely, 100% guarantee you that a manager rarely says, "well, gee... X seems to be swinging better than Y, but I guess I'll send Y out there, because I have a hunch that he'll hit better." That doesn't happen. Of course, perhaps you mean by "hunch," that a manager intuitively digests a great deal of information at once, without consciously evaluating it, and comes to a decision that way. This is possible. But that's not really the classical meaning of a hunch. That's just a decision that hasn't been thought out analytically. As for managing with computers... of course you don't. How obvious can you get? Where did I urge this? I didn't of course. That wouldn't work for very obviously. A computer can't teach plate discipline. A computer can't tell that one hitter has a whole in his swing that certain pitchers can exploit. Or that a player is a little scuffed up with a pulled ham string, and shouldn't start. Or that a player has been playing his heart out lately, and perhaps deserves a reward on effort alone. There are all sorts of things that a computer can't determine. But we can know, based on past performance, how players are likely to perform in the future. This is simple. You do the same damned thing in every day life, without realizing it. In tennis, when you hit to your opponent's backhand, it's usually because you've experienced that he's weak there. When you slow down on a certain stretch of highway, it's often because you know it's a speed trap because of past experience. Baseball is no different. Except, past experience is even more useful, because you have thousands of events of practically the same type, spread out over a relatively close period of time. Over the course of several hundred at-bats, patterns develop. And those patterns will tend to hold true in the future, regardless of whatever your "hunch" happens to be.
Great One-M story from last night. From the Chron: Two days after being called up from Class AAA New Orleans, the sidearming righthander was brought in with the five-run lead. He struck out Todd Walker to lead off the inning, and he could hear his wife, Jennifer, shriek in excitement from the seats down the third base line, where she sat with their boys, Darrin and Cole, and his parents. Puffer lost his control soon after, hitting Jason LaRue with a pitch and walking Wilton Guerrero. That drew a visit from manager Jimy Williams, who usually visits only to remove his pitcher. "He came out and asked me what I was going to have for dinner," Puffer said. "He also asked me who I was going to dinner with and who I was going to call tonight. I kind of assumed he was going to take me out. He was very calming, and I was excited that he left me in there." Puffer rewarded his manager's confidence, collecting a fielder's choice grounder from Barry Larkin and striking out Juan Encarnacion to seal the victory. Afterward, Puffer considered the postgame spread before leaving to eat with the family members who made the trip from Round Rock and Orange County, Calif. "His first time out on the major-league mound, right?" Williams said. "I don't know what kind of advice I can give him. Just relax him a little bit if I can. Just put the confidence in (catcher Brad) Ausmus behind the plate. Good thing for us he got the next two hitters."
Wow...that is actually pretty cool that Jimy did that. You know, despite the stories, I have been very pleased with the job Jimy has done so far. He only let Carlos throw 90 pitches last night, and I think some of the lineup tampering is useful. Sure, it is unlikely that Blum is going to hit better than Ensberg all season, but there is still that off chance possibility that Blum has learned something or changed his swing or who knows what that allows him to have a breakout season. The guy is still young, too, so it is certainly a possibility. If Jimy didn't at least allow him to ride out the hot streak, we might never know.
haven...what "past performances" does Ensberg have against major league pitchers. Regardless of what he has done in the minors the past two years, most of the guys he did it against are still in the minors. How does what he did against some 28 year old AA pitcher tell a manager anything about how he will do against Morris, Schilling, or Glavine? Anyway, my point was, managing is a lot more than analyzing numbers, which you obviously agree with. You were saying Ensberg deserved playing time over Blum because his projected stats from past performances dictate he should. I was saying that stats and projections be damned, if a manager sees a guy playing well, while another is struggling, he's going to play the one he believes deserves to be out there.
The "Genius" Billy Beane surely made Joe Morgan's day today. Phils pick up Jeremy Giambi from Athletics http://msn.espn.go.com/mlb/news/2002/0522/1385606.html OAKLAND, Calif. -- The Oakland Athletics traded outfielder Jeremy Giambi to the Philadelphia Phillies on Wednesday for utility man John Mabry. Giambi, the younger brother of New York Yankees first baseman Jason Giambi, was batting .274 with eight home runs and 17 RBI in 42 games this season. ``Jeremy was off to a good start but we were concerned he was too one-dimensional,'' A's general manager Billy Beane said. Mabry, used mostly as a pinch-hitter this season, was hitting .286 with three RBI in 21 games for the Phillies. Going into Wednesday's game against the Baltimore Orioles, the A's had dropped four straight, including their three-game weekend series against the Toronto Blue Jays. The A's have won just three of their last 17 games. ``This isn't a drastic thing,'' Beane said. ``We're not looking to do anything drastic. We just have players we need to take a look at.'' However, ESPN's Peter Gammons reports that the Athletics have been so frustrated by the early-season collapse that they has been shopping players -- such as Giambi -- in an attempt to shake up its roster and clear payroll. The A's have seven of the first 39 picks in the June draft. The A's had made a number of roster moves on Tuesday before the trade. They optioned rookie first baseman Carlos Pena, second baseman Frank Menechino and right-handed reliever Jeff Tam to Triple-A Sacramento. To replace them, Oakland recalled infielder Esteban German and outfielder Adam Piatt from Sacramento and purchased the contract of infielder Larry Sutton. Beane said the A's plan to use Mabry off the bench. ``It was apparent we needed to improve our defense, and we wanted to give Piatt some at-bats,'' he said.
