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.
Baseball Prospectus Red Sox Top 10
|
Post by moonstone2 on Jan 11, 2014 14:23:12 GMT -5
The reason there isn't a site that has data like that is because it won't tell you much about the major league future of a particular player. In the end isn't that why we really care about prospects at all? Not all players follow a linear development path.
We aren't talking about a couple of examples. We are talking about a large scale failure of SABR based analysis of prospects. Yes there were some Kevin Youkilis', but there were a lot more Jeff Natale's. As I said an entire organization, the Blue Jays, tried this and failed.
|
|
|
Post by jmei on Jan 11, 2014 15:12:36 GMT -5
Are you arguing that sabermetric analysis has essentially no place in prospect evaluation? Because if it has at least some utility, having folks who analyze prospects through that lens adds value, even though scouting-based analysis should predominate. I think the idea that statistical analysis "won't tell you much about the major league future of a particular player" is pretty clearly incorrect-- if prospect A is hitting .280 but with a 25% strikeout rate and a .400 BABIP, you know he's almost certainly not a .280 true-talent hitter, but if prospect B his .280 with a 13% strikeout rate and a .300 BABIP, that's much more sustainable. It isn't the end-all be-all, but to argue that it's basically useless just seems staggeringly false.
|
|
|
Post by brianthetaoist on Jan 11, 2014 15:47:33 GMT -5
That may be but I don't find the logic of assuming a player who hits a double or a triple in nearly 11% of his just over 200 PAs in A ball is going to do that consistently in the majors very convincing either. No qualified player has gotten over 8% over the last three years. Perhaps you could tell me where I said he would do that consistently in the majors? But I suspect from past experience with this kind of argument that this is going to go nowhere, so probably best just to drop it as a difference in viewpoint. I'm not really saying that, though. I'm making a less quantifiable judgement than that, one based on some first-hand looks, reading interviews, etc. Cecchini appears to have a very good baseball mind, one that focuses on details and improvement without getting overloaded or too hung up on failure. This is the art of watching prospects much less than the science, but I think it's an important part and, honestly, part of the fun ... it's where a lot of variance and surprises come from. Fair enough, although slightly off on what I was saying ... I agree that this isn't a huge difference or a case of people massively underrating a prospect. Cecchini is pretty much a lock for the top 100, very good chance at top 50. People like him! My feeling is almost exclusively a negative reaction on my part to what I see as two analytical errors: an over-confidence about a negative assessment of his chances to stick at third, and a view that he hasn't demonstrated power. Every once in a while, I get a slightly-irrational belief that people are making an error about a widely-discussed prospect on the Sox. Right now, that's Cecchini; I just think he's got a much better likelihood of his upside projection than Parks does.
|
|
|
Post by moonstone2 on Jan 11, 2014 18:20:48 GMT -5
Keep in mind that there is bound to be a ton of overlap in scouting and SABR evaluation. Players who scout well tend to perform well and visa versa. For SABR analysis to add value it has to tell you something that scouting does not and by and large the attempts at doing that have failed especially with regards to players in the lower minors. By and large most SABR analysis that I've seen has been something along the lines of evaluator X doesn't like player y who drew a million walks in A ball, therefore evaluator Y is dumb and doesn't understand plate discipline.
As a general rule the more diffuse the talent base the less BABIP is based upon luck. There are players who can square up far more pitches than everyone else and those players have a greater chance of major league success. As to your comparison of the two players, you'd need more information as to the players age, tools, level of experience swing etc. To use an example such as yours to conclude that one player is a prospect and the other is not seems reductionist in my view.
As a final note another reason why SABR analysis has been minimized on these sites is because scouting analysis by industry level evaluators is more accessible than it ever has been before. ESPN, BP, BA, and even this site do a great job of explaining how a prospect projects at the major league level and the range of reasonable outcomes. Most importantly all of them do so in a way that even someone who isn't in the industry such as myself can understand.
You guys are welcome to start a site of your own dedicated to SABR analysis of prospects. But in the end, I think you'll end up touting a lot of Matt Vanderbosh's and Lenny Dinardo and missing a lot of Mark Mulder's.
|
|
|
Post by thelavarnwayguy on Jan 12, 2014 15:24:30 GMT -5
I'll be honest about it and just flat out say I trust sabermetrics more than I do scouts. I've seen scouts be wrong over and over but detailed analysis of saber data over a significant sample size is probably an even better predictor to me. Of course scouts have access to the sabermetric data also, which obviously influences their projections so it's a moot point but if it were a straight either / or situation I would generally take the data.
For example some scouts liked Lin a lot but the data wasn't a good indicator. Scouts thought Nava wouldn't make it but the data was positive. On the opposite spectrum, scouts probably had a better handle on Jed Lowrie than the data indicated in the minors. Or Middlebrooks for example but the jury is still out there.
But the overall point is that scouts use the data in addition to their own eyes unless they are absolute idiots.
|
|
|
Post by godot on Jan 12, 2014 16:22:41 GMT -5
I'll be honest about it and just flat out say I trust sabermetrics more than I do scouts. I've seen scouts be wrong over and over but detailed analysis of saber data over a significant sample size is probably an even better predictor to me. Of course scouts have access to the sabermetric data also, which obviously influences their projections so it's a moot point but if it were a straight either / or situation I would generally take the data. For example some scouts liked Lin a lot but the data wasn't a good indicator. Scouts thought Nava wouldn't make it but the data was positive. On the opposite spectrum, scouts probably had a better handle on Jed Lowrie than the data indicated in the minors. Or Middlebrooks for example but the jury is still out there. But the overall point is that scouts use the data in addition to their own eyes unless they are absolute idiots. Well, you are missing an important point. The issue is the scout's skills and abilities. Good scouts can watch or observe a player, process the info he is seeing and often knows very quickly what they have. They can even get a sense of the player's mental make up, but it gets sketchy here. As for Lin, I hear scouts say he wasn't using his lower body and higher level pitching would cut him up, thus taking away his ability to walk. Of course, scouts would be stupid not to use the "data", but data is after the fact and "hints" at the player's skill sets. Not knocking sabemetrics by any means, but it is not a "science" but rather a (useful) tool. Indeed, disciplines that attempt to be a science such as economics have proven to be 'dismal'.
|
|
|
Post by Chris Hatfield on Jan 12, 2014 21:04:01 GMT -5
I'll be honest about it and just flat out say I trust sabermetrics more than I do scouts. I've seen scouts be wrong over and over but detailed analysis of saber data over a significant sample size is probably an even better predictor to me. Of course scouts have access to the sabermetric data also, which obviously influences their projections so it's a moot point but if it were a straight either / or situation I would generally take the data. For example some scouts liked Lin a lot but the data wasn't a good indicator. Scouts thought Nava wouldn't make it but the data was positive. On the opposite spectrum, scouts probably had a better handle on Jed Lowrie than the data indicated in the minors. Or Middlebrooks for example but the jury is still out there. But the overall point is that scouts use the data in addition to their own eyes unless they are absolute idiots. Unsure whether this is a straw man, a poor memory, or revisionist history. What scouts were wrong on Lin? The dude was barely a top 10 prospect. On the other hand, sabermetric data on Nava, or at least good data, would have taken his age into account and pretty much dismissed him. Blindly loving a guy's slash line without context ? sabermetric analysis. There was NO reason to think anything of Daniel Nava, which is what made him such a great story. Meanwhile, Lowrie's numbers in the minors were perfectly good, and I have no idea what your point is on Middlebrooks. I'm not sure what you're considering a "scout" here, but the job of a "scout" is to report back what they see. It is the job of others to incorporate stats. Other than that, rock on dude.
|
|
|
Post by thelavarnwayguy on Jan 12, 2014 22:09:49 GMT -5
You and I never seem to agree much, but maybe some of that is perspective. I'm not looking to disagree with you. What are you looking for?
Lots of you guys thought a lot of Lin. Lots of you. Not me but lots of you did. Nava's numbers almost always indicated better than the worthless trash heap most of the forum considered him when he was coming up. I remember saying things like "Heh, the guy is hitting .340 right?" and it was always followed up with "forget about it, he's 28 and he's nothing".
The point is that scouts should incorporate every aspect of a prospect if they are truly going to make an eval, but we've had this discussion before. Of course the data should be considered also, for a scout to make an informed opinion. His percieved makeup. His work ethic. His coachability...etc. Everything. That is blatantly obvious to any scout who is striving for excellence.
|
|
|
Post by chavopepe2 on Jan 12, 2014 22:26:12 GMT -5
I hate how any argument about incorporating statistics into prospect analysis devolves into a scouting vs. stats discussion. These two things are not opposites. They do not exist independently and there is absolutely no reason to have to choose one over the other. Saying you'd choose stats over scouting is a false choice. Statistical analysis and scouting are complementary and when used appropriate enhance each other.
Even worse than that is saying that stats, "won't tell you much about the major league future of a particular player." This is completely untrue. If taken in the context of a players age, level, and experience stats tell you a tremendous amount about a players major league future. Are there a range of outcomes? Of course, but that doesn't mean we can't learn a lot from looking at the numbers they put up.
|
|
|
Post by jmei on Jan 12, 2014 23:07:45 GMT -5
By and large most SABR analysis that I've seen has been something along the lines of evaluator X doesn't like player y who drew a million walks in A ball, therefore evaluator Y is dumb and doesn't understand plate discipline. You're either grossly exaggerating or reading the wrong stuff. There is a ton of great sabermetric prospect analysis out there, from research on how minor league BABIPs translate to the majors to translating 20/80 scout grades to fWAR to evaluating how well BA's pre-draft rankings hold up. Even on individual players, there's great stuff like chavo's recent work on Ranaudo or this Alex Speier piece comparing Betts with Cecchini. Keep in mind that there is bound to be a ton of overlap in scouting and SABR evaluation. Players who scout well tend to perform well and visa versa. For SABR analysis to add value it has to tell you something that scouting does not and by and large the attempts at doing that have failed especially with regards to players in the lower minors. I find it impossible to have a reasoned conversation about a prospect without citing at least some of their stats. The first thing almost all of us does when we think about a prospect is pull up his AVG/OBP/SLG or ERA/WHIP/Ks/BBs. If you're going to do that, you should absolutely also adjust for league/park and weigh OBP/SLG properly (i.e., use wRC+) or adjust strikeouts and walks to rate stats (i.e., use K% and BB%). From there, it's a short jump to adjusting for hitter BABIP luck or using your DIPS stat of choice. If you think the above adds zero value to prospect analysis... well, I guess I don't know how to respond. Call you out the next time you cite a stat in one of your posts, I guess? ADD: I also totally agree with chavo's post above. Thinking of scouting and sabermetrics as being zero-sum is pretty dated, I think.
|
|
|
Post by Oregon Norm on Jan 13, 2014 0:11:27 GMT -5
My own take on it is informed by having watched a good Lancaster lineup that included quite a few players who made it to the majors. I was specifically interested in Reddick as he was ripping up the Cal league. Watching him for a weeks time really informed the numbers I'd been reading about.
I remember describing his amazing follow-through and the power it generated, and more or less guessing that it would translate to the majors. His ability to put the bat on the ball was surreal. The problem was he was putting the bat on balls he should have been taking. Now the stats he was putting up didn't give me a hint of that. Though he didn't have a high walk rate, he was making such consistent contact that the average and power were just outstanding.
All I had to do was watch a relatively advanced pitcher, a guy who was too old for the league really, make him look like a fool through six innings to understand where the difficulty would be. And I'm not even a scout. It was that obvious. After three games in San Jose, the team went off to Visalia and we followed them there. He absolutely destroyed the pitching launching balls in the gaps and over the fence, and I mean launching.
Nevertheless, the weakness had been obvious. It's something he has yet to overcome to this day. Seeing him for myself made all the difference in the way I thought about the guy, and as others will tell you, I was a real fan of that bat.
I also saw Nava at the time and all the tools were there. The questions always revolved around his age - he was 25 at the time. He was a very different hitter from Reddick. The approach was fantastic, the same one we see today though greatly refined after all those years in the minros: Wait for a pitch to drive, and if you don't get it get on base via the walk. His pitch selection and discipline were quite apparent. Again, that's not something that would be obvious from just ticking off numbers. Lot's of 25 year-olds will kill in A+ ball. He looked different and it turned out he was.
Given that experience, I'm always reluctant to make blanket statements about players. If I could, I'd watch the Sox' affiliates at least a few times a year. I don't think there's any substitute, and that's from someone who will quote numbers day in and day out.
|
|
|
Post by jimed14 on Jan 13, 2014 9:03:40 GMT -5
You and I never seem to agree much, but maybe some of that is perspective. I'm not looking to disagree with you. What are you looking for? Lots of you guys thought a lot of Lin. Lots of you. Not me but lots of you did. Nava's numbers almost always indicated better than the worthless trash heap most of the forum considered him when he was coming up. I remember saying things like "Heh, the guy is hitting .340 right?" and it was always followed up with "forget about it, he's 28 and he's nothing". The point is that scouts should incorporate every aspect of a prospect if they are truly going to make an eval, but we've had this discussion before. Of course the data should be considered also, for a scout to make an informed opinion. His percieved makeup. His work ethic. His coachability...etc. Everything. That is blatantly obvious to any scout who is striving for excellence. How many 28 year olds put numbers up like Nava did in AAA and never made it to the majors? I'd guess most. That's why age is important.
|
|
alnipper
Veteran
Living the dream
Posts: 619
|
Post by alnipper on Jan 13, 2014 11:07:54 GMT -5
When I first at a prospects projectability I looked at stats first followed by a scouting report. What I saw a stat line like a guy hitting 300 it would catch my eye. Then I would want a scouting report. Later on a good scouting report would catch my eye. While now a good scouting report means more to me you can't count out a guy that continues to hit or pitch with really good numbers.
|
|
|
Post by moonstone2 on Jan 13, 2014 11:43:54 GMT -5
Okay if you hate that then let's redirect the discussion a little bit.
I think we can all agree that the whole idea of these prospects sites is to give us as fans a better understanding into who these players are and what type of player they might be in the majors. For myself, I don't have the time to go and see these players play very often and even if I did, I don't have the training or experience that guys like Chris Mellen have. Sites like BP give me insight into these players and increase my enjoyment of the game which is why I subscribe.
For BP or another site to add the kind of sabermetric analysis that you desire and be valuable it has to add to the fan's understanding of the player and/or come to a different conclusion than scouts might. Let's take the extreme example of Mike Trout. His performance was great in the minors and fortold a great player. But his performance didn't really tell me anything about him that a scouting report couldn't tell me. My personal preference is to hear about the times Trout turned a bobble in the infield into a double than his BABIP.
How could the type of analysis to which you refered to earlier accomplish that? Let's say you knew the erosion of the average walk rate from A ball to AA. What would that tell you about the player's future that BP doesn't already tell you?
Finally we are talking about statistical analysis not just citing the stat line in discussion. I am perfectly capable of looking up a players line on Baseball Reference. What you are talking about is a disucssion of what that line actually means for the players future and if it tells you anything about the players future that isn't already stated.
|
|
|
Post by moonstone2 on Jan 13, 2014 12:03:51 GMT -5
I thank you for those articles as they were interesting.
I think I should have made it more clear that I was refering to statistical analysis adding to my understanding and enjoyment of following prospects. When I say statistics of players don't tell you anything, I mean that they don't tell you anything additional. That doesn't mean don't site a minor league players stat line, to give a frame of reference. It just means you probably aren't going to be able to use statistics to tell me something that Chris and Jason and other sites didn't.
Take Chavo's piece on Ranaudo which you cited and I agree is excellent work. But in the end he concluded that Ranaudo wore down in the second half which is something several scouting sites also cited. It led to his sliding down most year end prospect lists. To add value, at least for me, statistical analysis has to do more than prove or demonstrate what is already publicly known about the player.
Finally I'll note that the piece on BA concludes as follows:
A groundbreaking conclusion!
|
|
|
Post by jmei on Jan 13, 2014 12:19:03 GMT -5
For BP or another site to add the kind of sabermetric analysis that you desire and be valuable it has to add to the fan's understanding of the player and/or come to a different conclusion than scouts might. Let's put it this way: you agree that statistical analysis of MLB players is useful, right? That looking at an MLB veteran's stats provides value above and beyond just reading their scouting reports? If that's the case, why is statistical analysis of a minor league prospect meaningfully different enough to complete negate any possible benefit? I think it provides pretty obvious advantages, ones I can fully expand on later (in a hurry right now).
|
|
|
Post by moonstone2 on Jan 13, 2014 12:20:15 GMT -5
|
|
|
Post by moonstone2 on Jan 13, 2014 12:27:17 GMT -5
Similar reasons as to why college and high school statistical analysis provides little value.
Smaller samples, more diffuse/poorer competition, different purpose in minors than majors. The further away from the majors, the less predictive the performance.
|
|
|
Post by jmei on Jan 13, 2014 12:36:59 GMT -5
Similar reasons as to why college and high school statistical analysis provides little value. Smaller samples, more diffuse/poorer competition, different purpose in minors than majors. The further away from the majors, the less predictive the performance. But you agree that as a player moves up the minor league ladder, their statistics become more meaningful?
|
|
|
Post by cto94 on Jan 13, 2014 12:42:24 GMT -5
Similar reasons as to why college and high school statistical analysis provides little value. Smaller samples, more diffuse/poorer competition, different purpose in minors than majors. The further away from the majors, the less predictive the performance. Except that college and high school stats come from seasons of 30-50 games against a way wider range of competition than you see in the minors, and minor league players play between 2 and 3 times as many games as well as facing much tougher competition on a consistent basis. Obviously minor league stats do have a limited predictive value or scouting would be way easier, but you can't just throw them out as next to useless the way you can with high school, and to a lesser extent college, numbers.
|
|
|
Post by moonstone2 on Jan 13, 2014 12:54:58 GMT -5
Similar reasons as to why college and high school statistical analysis provides little value. Smaller samples, more diffuse/poorer competition, different purpose in minors than majors. The further away from the majors, the less predictive the performance. But you agree that as a player moves up the minor league ladder, their statistics become more meaningful? Yes.......but I don't agree that they will tell you very much about a player that I didn't already know.
|
|
|
Post by moonstone2 on Jan 13, 2014 12:59:26 GMT -5
Similar reasons as to why college and high school statistical analysis provides little value. Smaller samples, more diffuse/poorer competition, different purpose in minors than majors. The further away from the majors, the less predictive the performance. Except that college and high school stats come from seasons of 30-50 games against a way wider range of competition than you see in the minors, and minor league players play between 2 and 3 times as many games as well as facing much tougher competition on a consistent basis. Obviously minor league stats do have a limited predictive value or scouting would be way easier, but you can't just throw them out as next to useless the way you can with high school, and to a lesser extent college, numbers. I don't think you are arguing anything I disagreed with or that refutes the quote above.
|
|
|
Post by jimed14 on Jan 13, 2014 13:02:30 GMT -5
How do you use sabermetric analysis to determine if a 20 year old with great walk rates in low A is great at recognizing pitches or is just standing in the batters box and looking for walks?
There is a lot more to it than walk rates. That's typically why .ISO is important as they move up because if all a guy is going to do is hit a single, why bother walking him? You at the worst have a 7 in 10 chance of having a better result than a walk and a low chance of something worse than that.
That's why players with great walk rates and iffy power are questioned.
And there is no sabermetric analysis of how a player is going to grow into his body and get stronger.
|
|
|
Post by okin15 on Jan 13, 2014 13:06:09 GMT -5
Sabermetrics weren't really developed to improve upon scouting, it was more to improve upon existing statistical analysis which centered on more traditional stats like RBI's and Avg.
Incidentally, traditional stats are great for counting what has happened, but generally fail as player evaluation tools either because they are not predictive (see RBI, Wins), or because they have limited correlation to winning (see Avg).
All that said, I will always lean towards a statistical analysis because I'm not good at scouting, and I am OK at numbers.
|
|
jimoh
Veteran
Posts: 3,989
|
Post by jimoh on Jan 13, 2014 14:21:40 GMT -5
I'm startled that there are people on this sight who don't think that statistical analysis of minor-league players is useful. It's far from infallible, but everyone knows that. Scouting of minor league players is great, but the kind that we have access to usually involves someone seeing a player once or twice, which is extremely fallible. I think it would take a lot of games for a scout to be able to tell, for example, whether a prospect "is great at recognizing pitches or is just standing in the batters box and looking for walks." It's not like the latter player stands up there eating a sandwich.
|
|
|