But like some things that makes total sense to you, there are others who disagree, and it is their right to do so. In particular, prospects who are older than the league seem to get some prospect hounds hearts a-pounding, when all I'm doing is trying to inject some sense of realism into the discussion, else the discussion degenerates into a talk about what we are going to do with our Mega-Millions lottery winnings. So they have the right to think that, but so do I have a right to my opinion. So who is right?
And that's where data comes in, to try to provide an answer. That was my first step to becoming a defender of Sabean over the years, I was like others, complaining about his drafting skills, but unlike others, I pulled the draft data and found that finding a good player, who is a starter and one likely to reach and become a wealthy free agent, is one of the hardest things to do, worse than trying to hit a baseball for a base hit, if you have a draft pick after the Top 5 picks or so, on average. Then, as the saying goes, it gets a lot worse from there.
So in this study, I'm going to look at the South Atlantic League (or SALLY to aficionados) over a period of years (2000-2005) to see what the data says about that, whether much older or much younger than the league. I pulled the Top 100 leaders in OPS and K/BB for each season, then sorted them by age to group them. Baseball-Reference.com kindly bolds the ones that have made the majors, then I separated out the ones that I recognized as a good player.
I was hoping to also show quality by breaking out the good major leaguers by eyeballing them. Unfortunately, my eyeballs aren't up to what they were before, so I'm delaying that portion of my study until I have time to go through all the major leaguers and put them into two categories, good major leaguer or other. This post will just cover the percentage of these top prospects making the majors.
What I wanted to do was put in a qualitative view by separating out the ones who actually were productive, and not just contributors. For example, I don't count Gregor Blanco as a recognized good player. He has been clearly been valuable, but has not really made a name for himself in the league yet. Plus some of these data points might still go up some if anyone breaks out late, but generally, the success rate is so low already in terms of just making the majors, that it won't really won't make that much difference, in my opinion, had I finished up that part of the study. But I decided to be more precise in my categorization.
However, I still think that showing just the percentage of these leaders who made the majors, broken up by age, shows good information for prospect hounds to consider when they are looking at prospects in the minors.
Analysis Says Ol' Folks, Don't Come On Up
Wow, I did not expect the clarity of the results I found. Here is the data:
The average age was usually around just over 21 YO, so I consider 21-22 YO to be basically age average for the league. Roughly a little less than a third of the 21 YO made the majors (32.9%), and roughly the same held true for 22 YO as well. This makes some sense, college players just drafted sometimes end up here and are that age. In any case, pretty bleak odds for prospects who are league average age.
However, any player clearly older (23 YO or older) who are playing in A-ball could pretty much forget about the majors. Only 11.3% even made the majors. I ran the hypothesis test on the sampling from age 23 to 27 against age 23 and it is likely that they come from the same data set, at least in terms of making the majors, so I bundled up that whole set up data above. None of them matched up with the 21-22 population.
And remember, these are only the ones who made the majors, if I differentiate by quality, it would be much lower, but as I noted, the time necessary to do that properly delayed my getting this out during the holiday season, so I decided to just cover MLB status only, as this is still a pretty strong statement: only 11% of all prospects considered older than the league even made the majors. And I suspect that even that is boosted by the possibility that sometimes teams send major league talent down to A-ball for rehab or sometimes as they are falling back down the minors after making the majors. So just by age, if you are a prospect old for the league, the odds of you even making the majors is pretty severely limited.
And as a sneak peak into the results of the qualitative analysis, only 1% of these older prospects ever turned out to be a good player, from my assessment of the major leaguers. So, it is not impossible to make the majors but pretty near impossible to become a good player, and therefore, as I had been saying at various Giants watering holes, being old for the league is a significant factor for any prospect hound to consider when thinking about any prospect's prospects for the majors, even for those with Top 100 level stats.
I also tested the younger ages and they are different from from the age 21-22 grouping at the 99.9% significance testing level, so they are pretty clearly different - and better - probabilities for the prospects who are young for the league. And the odds of being a good player, as I'll show in a later study, is that much higher as well.
But that was usually pretty clear before, prospect hounds knew that prospects who are young for the league are probably better prospects, the clear "ah ha" data point is that even prospects who are age league average among the leaders in OPS or K/BB are not likely to ever make the majors, with over two-thirds of these leaders never even tasting the majors, and even much less for those who are old for the league. And clearly, for the players old for the league, the vast majority of them will never even taste the majors, let alone be a good player.
Of course, a lot of this is self selection by the teams deciding to place the players in this league at a young age. Still, this shows that teams in general know what talent they have on their hands, confirming Matt Swartz's findings at BP and THT that teams generally 1) keep their best prospects and trade their lesser prospects and 2) keep their better free agents and let go of the lesser ones. Which demonstrates that teams are usually good at identifying talent in their own players. Young players that they have identified as talented are pushed to this league, and many succeed in not only making the majors, but becoming good players. But again, a strong slice of self selection involved.
Additional Data: A Study of Hitters in A-Ball at Beyond the Box Score
There is additional corroborating data from a study of prospects in A-ball at Beyond the Box Score (BtBS). Here is the key graphic for our purposes, and they did what I was trying to do with my study, separate out the quality prospects, and they did this for all A-ball leagues, not just Sally, but only BA Top Hitters:
As you can see here, busts galore happens for hitting prospects from 19 YO and over, so not only is being old for the league not that good (though there is a peak at age 22, most likely due to college draftees ending up in this league after signing, but with only 16 members, a bit on the small samples side), but that extends to even the younger members of the prospect pool, swallowing up the 19 and 20 YO's as well. Even 18 YO has a very high bust rate (as do 17 YO, but with only 5 datapoints, hard to say anything really).
As I noted above, this sample is not of all prospects, but of the BA Top 100 hitters, spanning 1990 to 2007. So that is why the rate of success here is much greater than for my sampling (which just covered the Sally and simply the top 100 players, both hitters and pitchers, whereas BBS covered all full season A-ball leagues and only included BA's Top 100 hitters, a much more select and talented group). Still, even with this highly selective and elite list of hitting prospects, the bust rate is incredibly high, I'm sure higher than most prospect hounds probably realizes, for la creme de la creme of baseball prospects. It is undoubtedly much, much worse for non-Top 100 hitters in any team's farm system.
Sidenote: in the bigger chart above the above graphic from the study linked, this data is exploded into low to high walk rate, as well as lot to high strikeout rate. Unsurprisingly, hitters who walk more and/or strikeout less, tend to experience more success in being a productive MLB hitter. But something good to be reminded about.
More of note, there is no sure to go formula to finding out who will succeed in being at least average in the majors, but very clearly, a low strikeout rate is very necessary to keep hope alive that you are not a bust. Even in the Average Strikeout column, most of the Bust% are very high, ranging from 62% to 88%, with pockets of relative success here and there. Even for Low Strikeout, it is not that much better, with Bust% in the 50's mostly. Still, a lot more green in that column for productive and average major league hitters.
ogc thoughts
Surprisingly, not only is being old for the league a big hit on a prospect's chances of making an impact in the majors, but being average age for the league also negatively affects their chances greatly. This corroborates what I've been commenting on the Internet, from my experience in dealing with prospects, which is that I will point out when people get extremely hyped up about any prospect who is old for the league, that the odds (and history) are against them being right that the prospect might provide some value at the major league level. Even most of the best prospects end up as busts, and prospects only on their team's Top 30 (or worse, not even on the Top 30) face even worse odds of doing anything interesting in the majors.
I also understand that each prospect is different. And I know that I'm not qualified to make that determination from a scouting viewpoint on an individual basis (though I think I do OK using available stats).
Still that does not mean that anyone can avoid the issue either by stating this individual difference. The failure rate on prospects in A-ball who are average or old for the league is very high, much higher than that for player young for the league. For any prospect to buck those odds, they need to already be on the radar nationally as a prospect, as are the BA Top 100 hitters, and any prospects not recognized as being the best in the minors, have even greater odds to buck.
So if a prospect is old for the league, while it is not the end for him, his odds of making the majors are that much more harder than that for prospects who are around the average age. And those have that much harder odds than the prospects who are young for the league, getting increasingly better as they get younger.
I'll cover in greater detail the results of my study when I have more time. Obviously, the higher the OPS, the better then odds of making the majors, so I'll give some stats on that. Same for K/BB. There are no absolutes, but things do get better as the prospect shows more talent via these metrics, within their own age groups. And similarly for being good starting players.
And for recent examples of prospect young for the league who did well for the Giants in A-ball, Adalberto Mejia, Kyle Crick, Clayton Blackburn, and Edwin Escobar all excelled in Augusta in 2012 at age 19 and 20 for Escobar. Blackburn was second in K/BB with 7.94, Escobar was 19th with 3.81, Mejia was 22nd with 3.76 and Crick just missed the Top 100 with 1.91 K/BB, #100 was at 1.93. Again, there are no absolutes with age, but clearly youths who can excel in this league have much more talent than any average or older prospects.
However, it is not a sure thing. Jorge Bucardo and Bryan Salsbury were 20 YO who did well in 2010, along with Zack Wheeler (who technically didn't do that well in K/BB, but was 20 YO and did strike out a lot, a metric I also like for pitchers to excel at relative to the league). If I knew how to use spreadsheets better as a stat tool, I would further reduce the Top 100 pitchers to those who made the Top 100 in K/9 as well as K/BB, as an ideal way to differentiate the top pitching prospects.
Adrianza was only 19 in 2009, but didn't hit great overall. However, he did have a nice OBP (.333; league average was .324), as he had a good number of walks (42 in 448 PA) and strikeouts (only 66, for a contact rate of 83%, which is very good in the league, where the average was 77%). He was not among the OPS leaders, not by a long shot, but as the BtBS study showed, prospects who walked a lot but struck out fewer had a much better chance of doing something good in the majors, being productive or at least average.
That he was able to do that while being young for the league speaks well of his latent talent, and hence why I've been so positive about his chances to do something in the majors. He does not have to be a superstar, even if he can become just a productive utility player, a player like Blanco or DeRosa, he can contribute good value to the team. He's battling for a 25-man spot this spring as the team is out of options for him, so he's either making the team or he's getting DFAed. And I suspect that his strong defense (already MLB good when he was in the lower minors) plus strong showing in contact rate and BB/K, plus improvements in ISO, will make him attractive to other teams, and that they will lose him if they DFA him.
I assume you looked at 2000-2005 to give current context but also give some space for current prospects. Interesting stuff. You want to post a list of players for some help with the "made it" MLB'ers?
ReplyDeleteThat is one thing that gets a bit fuzzy - the BA 100 has a big time hit rate with "made the show" but actual bonafides that stick and put up numbers are much more far between.
I agree that the Augusta rotation (that was then the SJ rotation and next year will be the Richmond rotation) has some nice success odds going in its favor!
Yeah, you know the deal, want to be as far back to allow the guys who will make it to make it, but the most recent data as possible, since things do change.
DeleteThe list is way too large for me to post a list of everyone. But I appreciate the offer. I will look back at the data and see what I might do, maybe I'll take up your offer, thanks.
Yeah, I wasn't totally sure how the BA 100 study exactly worked in terms of separating the wheat from the chaff, it's fuzzy, but clearly busts happens a lot all the way down to age 19, then improves significantly for younger prospects, though obviously SSS for age 17.
OGC,
ReplyDeleteYou have just re-proved that the Earth is round after it's been an established fact for hundreds of years. Nobody is saying age vs level does not matter. An 18 year old prospect who succeeds in the Sally League is obviously a much stronger prospect than one who puts up the same numbers at age 22. The problem comes when you start to split hairs between a 21 year old and a 22 year old when the difference is actually quite minimal by your own numbers.
In addition, I'm pretty sure that nobody in their right mind who has a greater then 10% chance of playing in the majors is going to "pretty much forget about the majors" which is a pretty ridiculous overstatement of the case.
As has been said many times in other discussions, age vs level is just one factor out of many in judging a prospect, an important one, but one of many. It is useful in terms of trends in groups of prospects, but becomes much less useful in judging an individual prospect, especially when the difference is just 1 year in age.
I'll try to explain better. In my second paragraph, I state this as what I was trying to prove in this post: In particular, prospects who are older than the league seem to get some prospect hounds hearts a-pounding, when all I'm doing is trying to inject some sense of realism into the discussion, else the discussion degenerates into a talk about what we are going to do with our Mega-Millions lottery winnings.
DeleteAbout ages 21-22, this is what I stated: The average age was usually around just over 21 YO, so I consider 21-22 YO to be basically age average for the league.
I found the percentages to be similar, and so I hypothesis tested them to be very similar, so I consider them to be in the same boat and did not separate them in my study above, please see my first table, the second to last row has the two ages together. In any case, 30% making the majors is not great odds of the player becoming a starter and contributing significantly to the team.
If a prospect gave up just because the odds are against them, they are never going to make it anyway. I've learned that many people give in too much to societal pressures and accept things, making it true in action or inaction as the case may be.
However, what I am trying to do here is assess how likely any prospect might make the majors, and further, who might actually contribute as a starter, and from most people's point of view, 10% odds of simply making it is pretty bad, I think, particularly once you consider the 1% odds I estimated regarding becoming a good starting player. That's lottery ticket odds. You would have to follow 100 of these older prospects in hopes that one of them will turn out to be a starter. I don't have that much time nor would I care to.
From my own personal point of view, just making the show isn't all that exciting for the team. I was not excited for the team by Bocock's cup of coffee or Culberson's or Gillespie's. Excited for them maybe, but from the overall big picture of the Giants becoming the Team of the Decade, not very exciting at all for me.
DeleteI never said that age was the end all or be all. That's something you somehow inferred, but since I do talk about a prospect's other attributes ALL THE TIME, like Adrianza's good control of the strike zone despite being very young for each level he has played at, I think most people who follow me over the years know that already.
But 1% odds of becoming a starter when you are old for the league, irregardless of how well you hit or pitched, is not something you ignore either, if you are following prospects who might become starters in the majors, which is all I'm really interested in.
If a prospect can become a Blanco, useful, that's not interesting to me, and hence why I didn't include him in my splits of the prospects who made the majors and made a name for themselves. I am interested in players who might contribute as a starter (or even a good reliever, plus, to clarify, I include closers as a starting position) to the Giants becoming the Team of the 2010 Decade, and if the odds is 1% of that ever happening, without considering any of his attributes, I'll pass and leave him to you to follow him assiduously. I would rather put my effort and energy elsewhere.
While I enjoy learning about most of our top prospects, I'm just not that excited until they look like they have good stats and other attributes as described by prospects books. But if the odds is around 1% of becoming a starter, I would rather devote my energies to the ones who look more likely to do something significant in the majors.
But that's my preference, and I realize I rain on some people's parades because of it, but the reality is that 1% odds is not that good. I know the truth is painful, but my not repeating it when people get all excited over their 42nd prospect in their prospect list, won't change that reality. But now my post is up, and I'll soon update this with even stronger evidence that older prospects are not all that worth following, and I won't have to repeat it as much.
Or put another way, you certainly don't give up on a prospect just because he's old, as you seem to be advocating, but you can't be that hopeful either and he would be one of those prospects who would need to prove himself over and over again with each new level he reaches, especially if he is not even on any overall prospect list or not even on his own team's Top 10 list. Age and lack of broad recognition of talent means that he has additional significant hurdles to make in order to make the majors as a good starter.
DeleteI know, no big whoop discovery that it is hard or that there are obstacles, but again, my main point as I pointed out is that the numbers suggest that it is foolhardy to be so positive about any over-aged prospect in the A-ball league (and I would add, particularly if he's not even among the Top 10 in his league with his stats), when the odds are roughly 100 to 1 that he ever makes a name for himself in the majors, no matter what he hits. I'm not going on wild goose chases anymore.
I'm not sure what analysts are getting exited about what prospects here. You seem to be erecting a gigantic straw man here. Again, nobody is dismissing age vs level and nobody is projecting future stardom for 24 year olds in A ball.
DeleteI believe in prior discussions I have objected to you making a big deal out of whether a prospect is 22 or 23 in AA or whether they are 20 or 21 in Low A. It looks to me like your own numbers show there is virtually no prognostic significance in those differences.
On the other hand, there are multiple examples of "late bloomers" in MLB. It never hurts to look for interesting stats and "sleeper" in the minor leagues, but if you think I or anyone else are anything but sober about their long term chances of success, you are sadly mistaken.
I would add that your numbers here are not only not a "big whoop" discovery, but they are no discovery at all and these numbers are well known in prospect evaluation circles and have been for a long time.
Well, can you at least admit that you haven't read everything that I've read? I've seen people get excited about prospects who are old for the league. I guess maybe where you go, everybody knows this rule, but I've not seen that to be true in my travels. I've seen guys flip out over guys who are clearly old, for example, many times at MCC.
DeleteNo, you are categorically wrong, I know I've never said those ages. I've never ever discussed prospects ages when they are league average, so you are putting words into my mouth. I've only talked about warning about prospects who are old for the league, that was my stance before this study.
Now my study shows that even average aged leading prospects are not that likely (and I don't know how else to term this when 70% of them fail to even make the majors) to be good prospects either. I was surprised by how bad the odds are for even averaged aged prospects, though now in retrospect, it makes sense, we all know that vast majority of prospects die on the vine as they rise up the farm system, but still, the numbers surprised me.
Maybe 70% didn't surprise you, but it surprised me. I've never said that I'm better than you or know more about prospecting than you, and, in fact, I've stated numerous times that you know a lot more than I do and recommend people to go to your site.
But I've been to sites where people don't know this. I see them get all excited about prospects old for their league and watch them move on to the next guy when he failed. Heck, I got excited by one long ago, I forgot his name, but he had really good stats and he just never made it to the majors, which made me cautious from then on. And I've seen others do the same.
And sure, there are late bloomers. If anyone is skilled in identifying them when they are in A-ball, then that person has hit the jackpot, there are plenty of teams that would pay millions for knowledge like that. I don't have this knowledge and would rather wait until these late bloomers reach AAA and then I might start paying attention to them. I don't have enough time to follow every potentially interesting prospects so I try to have higher standards for who to follow.
You are the one who brought yourself into the conversation. I never once said "DrB" wasn't sober about long term chances of a prospect. There are plenty of people who are not sober, I've seen them, you are the one sadly mistaken, I can't help it if you haven't seen what I've seen.
If these numbers are so well known, I can honestly say that I've never seen them once. In fact, if they are so well known, then the Beyond the Box Score analyst should have known that too, you should go on that website and tell them too, like you do all the time at Fangraphs. The people at Baseball Forecaster are pretty good at prospecting too, and in the 10 years I've been reading them, I've not seen anything referring to this "well known" fact. And I read broadly across the internet, as I'm interested in baseball analytics in general, I regularly read Fangraphs, THT, and I buy various books including Bill James, and I've read many of the publications that comes out of SABR, and I have not seen anything like this. And maybe I've missed it, but prospecting is interesting to me, and I would have been on the lookout for it.
So I'm glad that you already know this information, maybe that's a sign that I'm not advanced enough for your reading level, but I'm pretty sure that there are lots of people who don't know these numbers, and I'm writing for those people.
I'm never writing for people who already know this stuff, I write on things I'm interested in and that I've not seen anywhere else. If I'm not up to your knowledge level, well, I can't help that, but I know that there are people who don't know the things I've written about, I see that while reading at other places.