SoxProspects News
|
|
|
|
Legal
Forum Ground Rules
The views expressed by the members of this Forum do not necessarily reflect the views of SoxProspects, LLC.
© 2003-2024 SoxProspects, LLC
|
|
|
|
|
Forum Home | Search | My Profile | Messages | Members | Help |
Welcome Guest. Please Login or Register.
Jackie Bradley Jr. - does the glove outweigh the bat?
|
Post by jimed14 on Sept 10, 2015 12:28:24 GMT -5
I think Speier himself would object to attempts to use his quick and dirty comparison to define Bradley's projected future offensive floor. Most of the guys on that list have always been elite hitters who then had a supernova month (as elite hitters are prone to doing). That's not where we're at with Bradley, who had really good minor league numbers, one terrible major league season, and now one supernova month. That was exactly the point of the exercise - to find the worst career OPS+ on the list of people who did what JBJ has done. He said everything except that was his floor. Furthermore, JBJ's 30 day OPS was higher than anyone else on the list. And the list also includes some likely steroid users who were hitting when offense was way higher than it is now. Some mediocre players get hot and a little lucky all the time in baseball, but they don't hit a wRC+ of over 200 for 6 weeks.
|
|
|
Post by jimed14 on Sept 10, 2015 12:39:33 GMT -5
Also, Speier missed Encarnacion last month with a 1.379 OPS. JBJ probably would have won player of the month in any other month but that one since they don't care about fielding for it. And he had 35 ribbies too.
|
|
|
Post by jmei on Sept 10, 2015 13:16:02 GMT -5
I was objecting more to the 92% figure that telson13 raised, since a good chunk of those players aren't really comparable to Bradley. There were also a few methodological issues with Speier's analysis (he looked at Bradley's best 30-day period but compared that to other players' best calendar months; he required a .400+ AVG and used raw OPS rather than an era/park/league-adjusted offensive stat) as well as general sample size concerns that make it hard for me to buy in too much to that conclusion. But yeah, it is suggestive, and I'm excited to see what he'll do next.
|
|
|
Post by thursty on Sept 10, 2015 14:12:11 GMT -5
Those who are claiming that Speier's analysis is the proof that is needed to conclude that JBJ's hot streak is "real" couldn't pass statistical inference 101. And some of you actually know better - those who frequently cite recency bias and how cherry-picking numbers is fraught and that players cannot control sequencing
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 9,016
|
Post by ericmvan on Sept 10, 2015 14:26:37 GMT -5
4.5 WAR floor probably should be 4.0 WAR ... as good as we've seen JBJ play in the OF and as much as scouts love him, he's been more like a +10 defender so far.
Where .265 EqA is an average MLB starter, he has put up age-adjusted numbers of .304, .274, and .311 (not including his last two games) using everything he's done in the minors and majors, excluding last year which has proven to have no predictive power (to put it mildly). On what planet is that not a .265 floor? Speier's study is just a rough confirmation of what anyone should have suspected -- it's really hard to do what JBJ has done without being an average MLB hitter, or better.
A .265 EQa and +10 defense gives you a 4.0 floor.
I'll stand by my assessment that we're probably looking at a 5.5 WAR sort of player, perhaps as early as next year. You don't trade guys like that unless you've got an even better option breathing down his neck.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 9,016
|
Post by ericmvan on Sept 10, 2015 14:38:14 GMT -5
Those who are claiming that Speier's analysis is the proof that is needed to conclude that JBJ's hot streak is "real" couldn't pass statistical inference 101. And some of you actually know better - those who frequently cite recency bias and how cherry-picking numbers is fraught and that players cannot control sequencing That's not the argument at all. This is the same argument that Bill James made when he pointed out Roger Clemens' 15 K, 0 BB game as a rookie is meaningful. That's something that only pitchers above a certain quality are capable of. Speier's analysis reveals that it's extremely hard for a below-average hitter to have a calendar month anywhere near as good as Bradley's last month. Using a rolling 30-days instead of a calendar month would have yielded many more hits, but there's no logical reason to think that the percentage of average-or-above hitters in the sample would have changed. And had he adjusted for era and set a higher bar, it's likely that the percentage of average-or-above hitters would have been even higher. Statistical significance is always a function of both sample size and difference from the mean. Whenever someone is way above the mean for long enough, it establishes a likely floor of some sort. I'm quite confident that if you went to work with the right statistical tools on Bradley's month, your 90% confidence interval would have its floor more or less at MLB average. When you combine that with his ml track record, it pushes your confidence to 95% or so.
|
|
|
Post by mredsox89 on Sept 10, 2015 14:59:42 GMT -5
The whole purpose of the Speier post was just that it's very hard/unlikely (not impossible) for a non average at worst hitter to go on such a tear. So if JBJ can at worst be an average hitter, he could potentially be a 4+ WAR guy at minimum. Yea, he could be the rare case of a guy who puts this stretch together who isn't at least an average hitter, sure. But it's unlikely, which is far better than where things stood 4 months ago, a year ago, etc.
|
|
|
Post by jimed14 on Sept 10, 2015 15:11:25 GMT -5
4.5 WAR floor probably should be 4.0 WAR ... as good as we've seen JBJ play in the OF and as much as scouts love him, he's been more like a +10 defender so far. Where .265 EqA is an average MLB starter, he has put up age-adjusted numbers of .304, .274, and .311 (not including his last two games) using everything he's done in the minors and majors, excluding last year which has proven to have no predictive power (to put it mildly). On what planet is that not a .265 floor? Speier's study is just a rough confirmation of what anyone should have suspected -- it's really hard to do what JBJ has done without being an average MLB hitter, or better. A .265 EQa and +10 defense gives you a 4.0 floor. I'll stand by my assessment that we're probably looking at a 5.5 WAR sort of player, perhaps as early as next year. You don't trade guys like that unless you've got an even better option breathing down his neck. That's what I said when nomar asked what his trade value is. If a team is willing to value him as a 4 WAR player with even more upside, then you can talk about a possible trade. But still, those are the players we should be acquiring, not giving up.
|
|
|
Post by jimed14 on Sept 10, 2015 15:13:21 GMT -5
Those who are claiming that Speier's analysis is the proof that is needed to conclude that JBJ's hot streak is "real" couldn't pass statistical inference 101. And some of you actually know better - those who frequently cite recency bias and how cherry-picking numbers is fraught and that players cannot control sequencing So would you argue that a player who hits 5 HR in his major league debut hasn't proven he can hit more than 5 HR a year?
|
|
|
Post by brianthetaoist on Sept 10, 2015 15:29:15 GMT -5
Those who are claiming that Speier's analysis is the proof that is needed to conclude that JBJ's hot streak is "real" couldn't pass statistical inference 101. And some of you actually know better - those who frequently cite recency bias and how cherry-picking numbers is fraught and that players cannot control sequencing That's not the argument at all. This is the same argument that Bill James made when he pointed out Roger Clemens' 15 K, 0 BB game as a rookie is meaningful. That's something that only pitchers above a certain quality are capable of. Speier's analysis reveals that it's extremely hard for a below-average hitter to have a calendar month anywhere near as good as Bradley's last month. Using a rolling 30-days instead of a calendar month would have yielded many more hits, but there's no logical reason to think that the percentage of average-or-above hitters in the sample would have changed. And had he adjusted for era and set a higher bar, it's likely that the percentage of average-or-above hitters would have been even higher. Statistical significance is always a function of both sample size and difference from the mean. Whenever someone is way above the mean for long enough, it establishes a likely floor of some sort. I'm quite confident that if you went to work with the right statistical tools on Bradley's month, your 90% confidence interval would have its floor more or less at MLB average. When you combine that with his ml track record, it pushes your confidence to 95% or so. Yeah, I think the view of the sample Alex was working on might be skewed ... the relevant point is not "of the 22 players who fit this sample, only X is not a really good hitter," but more, "of the hundreds of below average hitters over the last ten years, none have managed to accomplish this feat." That's a significant statement. It's not necessarily dispositive, but it's certainly interesting.
|
|
|
Post by telson13 on Sept 10, 2015 18:07:18 GMT -5
That 4.5 WAR is based on him posting average offensive numbers. And if you read the article above, you'd see that there is, based on his cohort, a high probability that he does so: 92% of players posting a month like he has had a career OPS of 110 or better. That's why it's a reasonable floor. Sure, he could break a leg or explode in a ball of white light, too. But the whole "Bradley could be crappy and Syndergaard could get better" argument is nonsensical. Sure, it *could* happen, but the odds are against it. At the very least, Syndergaard represents a substantially greater injury risk. If love to see the Sox acquire him, but this comparison was based on numbers and projections, not "the feels." I can certainly envision Syndergaard putting up some 5 or 6 WAR seasons, but right now the data say he's less (and maybe significantly) likely to do it than JBJ is. Personally I don't care what 92% of a small sample of players went on to do after an incredibly hot month. I definitely wouldn't project Bradley's value using that in any fashion. But you are projecting equivalent value, improved performance, and continued health for a 22-y/o pitcher with a half-season under his belt, with no data-driven basis for it. That's where I question your selective use of the label of "absurd."
|
|
jimoh
Veteran
Posts: 4,123
|
Post by jimoh on Sept 10, 2015 18:12:10 GMT -5
This article on Harper suggests that it would not have been hard to find non-calendar month hot-sreaks of roughly a month long, if you go by PAs. "With research assistance from BP’s Rob McQuown, we can look for the best TAv streaks over any given number of plate appearances, regardless of when they occurred on the calendar." Unfortunately he stops at 90 PAs, even though he says he's gpt the list of 100 PAs right in front of him. He's got ten guys, with Harper 10th, with TAv over .538 for 90 PAs, and OPS of, um, about 1460-1700 grantland.com/the-triangle/2015-mlb-bryce-harper-historic-hot-streak/
|
|
|
Post by telson13 on Sept 10, 2015 18:17:26 GMT -5
I was objecting more to the 92% figure that telson13 raised, since a good chunk of those players aren't really comparable to Bradley. There were also a few methodological issues with Speier's analysis (he looked at Bradley's best 30-day period but compared that to other players' best calendar months; he required a .400+ AVG and used raw OPS rather than an era/park/league-adjusted offensive stat) as well as general sample size concerns that make it hard for me to buy in too much to that conclusion. But yeah, it is suggestive, and I'm excited to see what he'll do next. 23/25=92%, is where I got it. I agree with the "methodological issues," in that going by months and not 30-day periods creates an artificial subset that likely has suppressed peak numbers (overlap of "cold" periods). However, there's little reason to believe that this sample is non-representative. Obviously, it should be taken with a grain of salt, but it is interesting and not unsurprising given his minor league history and projections therefrom.
|
|
|
Post by burythehammer on Sept 10, 2015 18:22:39 GMT -5
If there's a baseball team out there that thinks Jackie Bradley is a 5 WAR player then not trading him would be an act of incompetence.
|
|
|
Post by bosox81 on Sept 10, 2015 19:32:16 GMT -5
Yeah, I think the view of the sample Alex was working on might be skewed ... the relevant point is not "of the 22 players who fit this sample, only X is not a really good hitter," but more, "of the hundreds of below average hitters over the last ten years, none have managed to accomplish this feat." That's a significant statement. It's not necessarily dispositive, but it's certainly interesting. Quoted for emphasis. I think this is a crucial point that is escaping many of the posters here.
|
|
nomar
Veteran
Posts: 11,501
|
Post by nomar on Sept 10, 2015 21:20:14 GMT -5
Here's hoping Eric is right. I'd love it. Make it happen JBJ!
|
|
|
Post by telson13 on Sept 11, 2015 0:16:12 GMT -5
Yeah, I think the view of the sample Alex was working on might be skewed ... the relevant point is not "of the 22 players who fit this sample, only X is not a really good hitter," but more, "of the hundreds of below average hitters over the last ten years, none have managed to accomplish this feat." That's a significant statement. It's not necessarily dispositive, but it's certainly interesting. Quoted for emphasis. I think this is a crucial point that is escaping many of the posters here. Exactly, that's an outstanding analysis. And Speier does allude to that in the article, essentially pointing out that this was a feat achieved by exclusively average or better hitters. The mean, median, and mode career performances among that group are all likely substantially better than average, since the 8th percentile of performance in that group was a career OPS+ of 110. And when one considers that not one significantly below-average hitter (OPS+ of 99 being the low-water mark) made the list, that suggests that the subset of hitters that actually do qualify for the list have a requisite skill level that allows them to do so. It doesn't predict what JBJ will be, but it does suggest that it would be highly unusual (1 in hundreds) for a below-average hitter (retrospectively looking at career OPS+) to have a month of that caliber. Therefore, JBJ would represent an extreme outlier were he to make that list and turn out to be a significantly (say, OPS+ of 95 or lower) hitter.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 9,016
|
Post by ericmvan on Sept 11, 2015 0:41:41 GMT -5
If there's a baseball team out there that thinks Jackie Bradley is a 5 WAR player then not trading him would be an act of incompetence. But what if that team is the Red Sox, and they're right? Clay Davenport's not exactly a dummy: he was the stathead among the BP founders. They wouldn't exist without his acumen. I've met Clay and spent a decent amount of time chatting deep sabermetric stuff, and he's as sharp as they come. If you toss out 2014, he's been projecting Bradley as a 5 WAR player. I'm projecting Bradley as a 5 WAR player. So, please, either give us the argument that we're both wrong (which will have to backtrack and undermine Bill James' idea of Major League Equivalancies), or give us the argument that 2014 should matter. And, oh, while you're at it, break down Bradley's current swing and show why his mechanics, bat speed, pitch recognition, and hand/eye coordination are collectively merely MLB average. And then explain why all the scouts who saw him early in his ml career thought otherwise. (Oh, and even if you were right about him, and he's only a 4.0 - 4.5 WAR player, you would keep him rather than trading him for 5 WAR of value, because the downgrade to his replacement would be larger than the upgrade elsewhere. IOW, the upgrade of a 5 WAR pitcher over the best of Kelly, Wright, and Owens is smaller than the downgrade from a 4.0-4.5 WAR Bradley to Brock Holt, plus (crucially) the downgrade from Holt as the first guy off the bench to Deven Marrero or whoemever.)
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 9,016
|
Post by ericmvan on Sept 11, 2015 2:48:58 GMT -5
Rank of Jackie Bradley, Jr. in fWAR/PA, of 142 MLB outfielders with 160 or more PA: 2. (Between Harper and Trout.)
Rank of Hanley Ramirez: 139.
|
|
|
Post by burythehammer on Sept 11, 2015 8:45:31 GMT -5
I'll take the under on 4 WAR too, unless he has a crazy great defensive season, which I guess is more possible with him than most.
I have always liked Bradley, always thought he would hit and still do, which makes him a pretty good player. For pete's sake I was defending him in this very thread last year when he was the worst hitter in baseball, saying that his defense still made him an everyday player. So the fact that I'm being forced into a position where I'm "down on him" shows how ridiculously you are overrating a small sample, which, as we all know, is your MO. That's fine. As usual, I disagree, and so do most others here in this case.
But I can't break down his swing, so what do I know...
|
|
|
Post by GyIantosca on Sept 11, 2015 8:50:23 GMT -5
I can't believe how much JBJ talk on the radio was getting out of hand. I mean Maz and Felger actually suggested PED's! What a slap in the face for a kid that kept working on his craft and was getting circle jerked this season. I admit last season was horrible at the plate but the kid was always a legit prospect. The offense is a bonus most people said if he could just hit a little with that exceptional defense. The Sox are lucky they can absorb those bad deals Ben made because they have six years of organizational control on these kids. It saves them a lot of money on that end. Nobody else gets to steal JBJ from us . That would of been bad seeing him in LA and we have Hanley booting ground balls in the corner in left field. Hanleyis like an old guy with his shoulder. He is probably dying for Papi to retire some could sit on the bench and just hit. I bet you he thought this guy would retire in a year or two and he can DH the rest of the way.
|
|
|
Post by burythehammer on Sept 11, 2015 9:00:33 GMT -5
BTW, know how many position players had 18 WAR from 2011-2014 (four years)? 17. How many had 20? 10. Yet someone how it's far more unlikely that Syndergaard, who has a better major league track record and arguably a better minor league one, does something that 21 pitchers have done over a six year span.
|
|
|
Post by threeifbaerga on Sept 11, 2015 9:06:35 GMT -5
So if Jackie Bradley can put up a defensive season like he did last year in 127 games, where he was a 2.0 war defender, he only needs to be an above average CF with the bat in order to be this mythical 4.5 win player people are adamant he can't be.
What is the baseline for an average hitting CF? .331/.441/.772. Would anyone be surprised if he beat that next year, seeing that player before us right now?
Not to mention the average AL CF has hit worse, .323/.425/.748.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 9,016
|
Post by ericmvan on Sept 11, 2015 9:14:17 GMT -5
I'll take the under on 4 WAR too, unless he has a crazy great defensive season, which I guess is more possible with him than most. I have always liked Bradley, always thought he would hit and still do, which makes him a pretty good player. For pete's sake I was defending him in this very thread last year when he was the worst hitter in baseball, saying that his defense still made him an everyday player. So the fact that I'm being forced into a position where I'm "down on him" shows how ridiculously you are overrating a small sample, which, as we all know, is your MO. That's fine. As usual, I disagree, and so do most others here in this case. But I can't break down his swing, so what do I know... No, this argument suggests that you have no idea how WAR works. An MLB CFer who is an average MLB hitter and average fielder is a 3.0 WAR player. Bradley's going to be a +9 defender at a minimum. He may be +15 once he settles in, but that's his floor. And that's an extra win. So if you're taking the under on 4 WAR, you are simultaneously arguing: "I have always liked Bradley, always thought he would hit and still do" [not qualifying it with "a little bit" or "adequately" or "well enough to start"] and Bradley will be a below-average MLB hitter. Which is it? I asked you to make an argument that he was going to be below-average-hitter, and you responded by implying that you maybe didn't. And the sample size here is 1475 PA. You seem to be ignoring that, too.
|
|
|
Post by burythehammer on Sept 11, 2015 9:15:36 GMT -5
Nobody is adamant that he can't be a 4.5 win player. Somebody is adamant that 4.5 wins is his floor.
|
|
|