Batter's Box Interactive Magazine Batter's Box Interactive Magazine Batter's Box Interactive Magazine
Top prospect lists always catch my attention. Over the past few months I have seen lists from Baseball America, John Sickels, and yesterday Baseball Prospectus. I always check the lists to see how Blue Jay prospects rate and wonder if the Jays, as an organization, have a strong system. However my excitement is tempered by the knowledge that these lists, and organization rankings, have little meaning from a major league perspective. No matter how many "names" your team has on a list there are no guarantees. Many of us are excited with the thought of Alexis Rios, Guillermo Quiroz, Dustin McGowan and Gabe Gross becoming Blue Jay regulars by 2005. But then reality hits me upside the head saying "nothing is sure", and the realization that, on average, one or two of them will crash and burn.


Some organizations have weak farm systems, but they have one or two players who turn into major league stars. Other organizations are strong, but they have a run of bad luck and have their prospects fade when they get to the major leagues. I know that drafted players are a gamble, but how much of a gamble are top prospects lists? Baseball America publishes an annual top 100 prospects list. If the Jays have six players on BA's top 100 prospects list should I be excited? 100 prospects over thirty teams is an average of 3.3 per team. So if the Jays have six players listed, how many of these prospects should make it? I decided to check how sure is it that a top 100 prospect will make it. Before I started I guessed that half of them would make it. What do you think?

I took the BA top 100 list from 2000, and checked the stats for all players in 2003. I divided the players into five categories:

S - All-Stars
R - Regulars
U - Utility major league players
P - Part time, meaning some time in the majors and some in minors
O - Not in the major leagues, either out of baseball or still stuck in the minors.

Here is the split:

 S   R   U   P   O  Total
11 31 15 15 28 100


42% of the prospects are major league all-stars or regulars. Major names among the all-stars are Vernon Wells, Rafael Furcal, Alfonso Soriano, Eric Gagne, Mark Mulder, Barry Zito. Some of the best known busts are the two top pitching prospects, Rick Ankiel (#1) and Ryan Anderson (#9). Others are Josh Hamilton (#13), Drew Henson (#24), Chad Hermanson, signed by the Blue Jay's to a minor league deal (#33), and Wilson Betemit (#99).

I then looked to see if there is a difference between pitchers and hitters. I know drafted pitchers are more risky than hitters but what about top prospects. The numbers:

Type     S   R   U   P   O  Total
Hitters 7 15 11 11 14 58
Pitchers 4 16 4 4 14 42


Pitchers on the top prospect lists are just as likely to be regulars, within the limited scope of my review. Twenty two hitters, or 38%, are stars or regulars. 48% of pitchers are stars or regulars. However more hitters make it to "the show" and hang around as utility guys. More pitchers never make it. One third of the top 100 pitching prospects from 2000 did not appear in the major leagues in 2003.

I then looked at the distribution of stars and regulars by quartile. I would think that the top 25 prospects would be more successful than the prospects from 76 to 100.

Quart   S  R  U  P  O  Total
All-Q1 4 9 5 1 6 25
All-Q2 3 6 3 7 6 25
All-Q3 3 8 2 7 5 25
All-Q4 1 8 5 0 11 25
-----------------------
Quart S R U P O Total
Hit-Q1 3 5 4 0 3 15
Hit-Q2 1 1 2 6 4 14
Hit-Q3 2 3 0 5 1 11
Hit-Q4 1 6 5 0 6 18
-----------------------
Quart S R U P O Total
Pit-Q1 1 4 1 1 3 10
Pit-Q2 2 5 1 1 2 11
Pit-Q3 1 5 2 2 4 14
Pit-Q4 0 2 0 0 5 7


A prospects position on the list is not very significant. The first quartile (#1-25) was the most productive but not by a significant margin. The fourth quartile provided seven all-star or regular hitters. The second quartile produced two. Now I did not look at quality so the bottom quartile hitters include Ramon Santiago, Luis Rivas, Joe Crede, and Hee Seop Choi. Similarly for pitchers, the second quartile was more productive than the first.

Admittedly this is a small sample. However the findings are not unexpected. More than half of the top 100 prospects will not become major league regulars. There is no logic that I can see to determine who will make it and who will not. Some of the most highly rated prospects never were successful in the major leagues, some only marginally successful. If I look at other years I see names such as Ben Grieve who was the #1 prospect in 1998. He has not lived up to that ranking. Bruce Chen, who is in the Jays camp, was #4 in 1999. Matt White and Reuben Rivera made two appearances in the top ten. As a side note, Vernon Wells was #69 in 1999, then jumped to #4 in 2000, but dropped to #12 in 2001.

So what does this mean for the Jays? We should expect that only two of the Jays top 4 prospects (Rios; Quiroz; McGowan; Gross) will make it as regulars. If three make it we should be ecstatic. In 2004 at least one of these prospects will likely struggle at their next level.

Of the Blue Jays top ten prospects, the law of averages suggest four will make it, five would be good, three would be bad. Over the course of 2004 half of the top ten prospects will likely stall or regress. Look at the 2003 BA top ten list for the Jays. Only Rios, McGowan, League and Cash really justified their rankings. Adams could be considered to have moved up slightly or to have stalled, depending on your perspective. Rosario (injured), and Griffin definitely stalled, while Werth, Arnold, and Chulk dropped in the rankings. This is about average.

So remember prospect lists and organizational rankings are interesting. But less than half of the players will make it in the big leagues. But when we are looking for three or four players from the top ten to make it, one or two above or below the expectations can make the GM a hero, or a zero.

How Good is a Top Prospect? | 34 comments | Create New Account
The following comments are owned by whomever posted them. This site is not responsible for what they say.
Craig B - Wednesday, February 25 2004 @ 11:51 AM EST (#77244) #
Can we get that top 100 list, Gerry?

This is very interesting stuff. 42% becoming regulars sounds a touch high to me, but generally this study is about where I would have expected. The fact that the top 25 don't do better than 26-75 I thoguht was very interesting.
Mike Green - Wednesday, February 25 2004 @ 12:07 PM EST (#77245) #
Gerry, 2000 is too soon to do the test. Some of the prospects (Wilson Betemit, for example) might yet make it. I'd say that 50% is probably a fair guess of what the figure might be if you had used 1998.

I know that Ben Grieve has not developed in Tampa since his stellar performance of 1998-2000 with the As, but he's only 27 and it would not shock me if he emerges as a great player sometime soon. He's in Milwaukee now, and this might be a better context for him. Anyways, I'd still count him as a prospect success.
_Dean - Wednesday, February 25 2004 @ 12:18 PM EST (#77246) #
Great article Gerry, it really reinforces the point that there is a big step between being a prospect and having any kind of success at the major league level and that nothing is guaranteed. Injuries and quite often the mental aspect of the game put a halt to many careers. I know BA will sometimes take a look back at a top 100 list or a draft and go over why they had a certain player ranked and how the player has turned out. If the player has not panned out as they predicted they will give examples why. e.g. "still throws 93 but is straight and does not have command of it".
_Ryan Day - Wednesday, February 25 2004 @ 12:41 PM EST (#77247) #
One thing to look at would be the level those Top 100 players were at. A player who's been spectacular at AAA is probably much more likely to succeed than a guy who's shown great stuff at low-A. You might want to differentiate "Injured" from the "Out" and "Part-time" players; it's one thing to fail because you weren't that good in the first place, it's another to do so because of an injury.

Also, I'm not sure you can apply the general rate of success to Jays' prospects. There are so many reasons players fail: Injuries (Ryan Anderson), personal issues (Hamilton), bizarre meltdowns (Rich Ankiel, maybe Ben Grieve). They're not going to be equally distributed.

It's true, of course, that there are no guarantees. But there are no guarantees about players with 10 years of MLB experience, either.
_Kristian - Wednesday, February 25 2004 @ 12:44 PM EST (#77248) #
Awesome article Gerry, it was great data to look at. If you even go back to last years top 100 it already looks like a banner year. Teixeira, Baldelli, Reyes,Mauer,Contreras*,Matsui*,Krod,Cabera,Martinez and Bonderman all project to at least have starting jobs this year and Phillips,Morneau and Cuddyer have had a cup of coffee in the majors. Thats from the top 20 alone. This years top 100 comes out Feb 27th it will be interesting to see how many Jays make the list? Early prediction is Rios, Magowan, Quiroz, Rosario, possibly Hill,Adams, Bush and Perkins.
Gerry - Wednesday, February 25 2004 @ 12:46 PM EST (#77249) #
Can we get that top 100 list, Gerry?

The list is subscriber only on BA but here are some details:

Stars:
4 Vernon Wells, of, Blue Jays
8 Rafael Furcal, ss, Braves
12 Mark Mulder, lhp, Athletics
16 Alfonso Soriano, ss, Yankees
37 Lance Berkman, of, Astros
41 Barry Zito, lhp, Athletics
49 Eric Gagne, rhp, Dodgers
56 Adam Dunn, of, Reds
57 C.C. Sabathia, lhp, Indians
74 Marcus Giles, 2b, Braves
95 Jimmy Rollins, ss, Phillies

Regulars:
2 Pat Burrell, 1b/of, Phillies
3 Corey Patterson, of, Cubs
5 Nick Johnson, 1b, Yankees
7 Sean Burroughs, 3b, Padres
14 Kip Wells, rhp, White Sox
19 Josh Beckett, rhp, Marlins
20 A.J. Burnett, rhp, Marlins
22 Brad Penny, rhp, Marlins
23 Eric Munson, 1b/c, Tigers
27 Tony Armas, rhp, Expos
28 Ramon Ortiz, rhp, Angels
29 Francisco Cordero, rhp, Rangers
32 Jon Garland, rhp, White Sox
36 Milton Bradley, of, Expos
39 Danys Baez, rhp, Indians
53 Jesus Colome, rhp, Athletics
58 Kurt Ainsworth, rhp, Giants
61 Peter Bergeron, of, Expos
64 Adam Eaton, rhp, Padres
65 Ben Sheets, rhp, Brewers
67 Cesar Izturis, ss, Blue Jays
69 Wade Miller, rhp, Astros
73 Carlos Guillen, 3b, Mariners
77 Hee Seop Choi, 1b, Cubs
81 Byung-Hyun Kim, rhp, Diamondbacks
82 Russ Branyan, 3b, Indians
86 Luis Rivas, 2b, Twins
87 Jason Jennings, rhp, Rockies
92 Ramon Santiago, ss, Tigers
96 Joe Crede, 3b, White Sox
98 Aubrey Huff, 3b, Devil Rays
_Roger Davis - Wednesday, February 25 2004 @ 12:49 PM EST (#77250) #
http://www.immune25.tv
BA, and others, have a problem with their top 100 lists; namely, they have to balance those close to "major league ready" with those just drafted but with "percieved" very high ceilings. Some of the last 20 on most top 100 lists are infact the "best" prospects BUT they just don't have the track record yet.

Gross and Adams (Hill too) are "safe" top 100's. Neither of those guys will be more than a journeyman at best, BUT they almost assuradely will at least be a journeyman (this is not meant to insult them, most journeymen have a reasonable career).

IMO Rios,McGowan and Rosario (assuming his arm stays on) are the only "identified" "stars" in our system. Quiroz, Perkins, Peterson, Bush, Arnold, etc. fall into a second tier. Guys like Banks, Isenberg, Marcum, Mastny & the two V's (Vito & Vermilyea) are just too far away to confidently speculate on at this point.
_coliver - Wednesday, February 25 2004 @ 01:23 PM EST (#77251) #
A prospect is a prospect is a prospect...nothing is proven until the player can do the job in the big leagues. This may be unpopular to say in "da box" but there are too many intangibles that cannot be quantified. For every Vernon Wells, there are too many Andy Thompsons, Wilson Betemits, Sil Campusanos, Kevin Witts, Jossephang Bernhardts, etc...

Both Baseball America and Jim Fregosi thought the world of Andy Thompson...
_Jeff - Wednesday, February 25 2004 @ 01:36 PM EST (#77252) #
This an interesting peice. I agree with Ryan Day that a player's level should also be considered.

One question is how did you separate stars from regulars, etc.?

It seems odd that Adam Dunn and Jimmy Rollins are stars whereas Aubrey Huff and Josh Beckett aren't. Also Peter Bergeron, Russell Branyan and Ramon Santiago are hardly regulars.
Coach - Wednesday, February 25 2004 @ 01:42 PM EST (#77253) #
2000 is too soon to do the test

I don't know about that; I think it's quite interesting to see how much the list keeps changing. For example, Marcus Giles, listed now as a star, took a tremendous leap forward last year to justify that ranking. I might have included Aubrey Huff and Josh Beckett as "stars" too; obviously their stock is rapidly rising.

There are many career paths evident already. Someone like Eric Gagne might have been considered a disappointment in 2002, only to win a Cy Young two years later. Others, like Adam Dunn and Pat Burrell, burst upon the scene, looking like perennial all-stars, only to experience some growing pains. Yes, it's too soon to draw definite conclusions on this crop, but there's never certainty about a player's career until it's over. There are several names on the "regulars" list who I think will break out in 2004 -- if I'm right, my fantasy teams will have a good year. It's guesswork; anyone who turns it into a science will become famous, and probably rich.

The 2005 Jays could have a couple of Rookie of the Year candidates and some "established" new stars by 2006. The farm system should continue to pay dividends for a long time after that, due to the sheer quantity of solid prospects. Trying to identify who they will be is a pleasant exercise, involving a lot of luck. To me, the odds on any one player living up to expectations are pretty much 50-50 -- he will, or he won't.
_coliver - Wednesday, February 25 2004 @ 01:53 PM EST (#77254) #
I agree Coach--it is a 50/50 conjecture. Too many nonquantifiable intangibles--we cannot get into the player's mind-set.
Gitz - Wednesday, February 25 2004 @ 01:58 PM EST (#77255) #
I should clarify something I said yesterday: the prospect lists I enjoy are individual team reviews, ones that don't rank the prospects in order or prospectability. Last year, for example, Jordan wrote a rich review about every Jays minor leaguer, which was the best thing I had read, because he didn't put them in order. Ranking them 1-100 is, as I said, a fun exercise, but given that there are literally thousands of minor-leauge baseball player, it's ultimately an arbitrary one.

Oh, and Mike Green: give it up! Grieve is done. Like everyone else, I kept waiting the last few years for him to re-establish himself, but at this point it would be a serious miracle. Which makes me wonder: who has been the most promising young player in recent years to disappear, like Grieve has, for no apparent reason? My memory is dull from years of the Simpsons and from marriage, but I know there are some prospects who came up, had terrific success early on, then faded away in rapid fashion. I trust the Cabal and Da Box regulars will help me out.
Gerry - Wednesday, February 25 2004 @ 02:00 PM EST (#77256) #
One question is how did you separate stars from regulars, etc.?

It seems odd that Adam Dunn and Jimmy Rollins are stars whereas Aubrey Huff and Josh Beckett aren't. Also Peter Bergeron, Russell Branyan and Ramon Santiago are hardly regulars.


If Baseball Reference had an All-Star button beside their name they were an all-star. I assume it means they played in the All-Star game.

Santiago played in 141 games in 2003 so that's pretty regular to me. Branyan I had thought was injured so I gave him the benefit of the doubt. Bergeron I made a mistake, I thought he was injured too, but he was not.
Mike Green - Wednesday, February 25 2004 @ 02:02 PM EST (#77257) #
The list is very interesting. The top 4 position players listed as everyday regulars- Burrell, Patterson, Johnson and Burroughs, are all in the early days of their careers, and all have the reasonable potential to be All-Stars. The position players at the bottom of the regular list- Izturis, Carlos Guillen, Branyan, Rivas, Crede and Huff are, with the exception of Huff, of noticeably lesser quality.

My own guess is that a top 25 position player-a Rios, a Jeremy Reed- has about a 60% chance of being an All-Star or near All-Star, within 3 years, and that the odds are much, much lower (maybe 25%-30%) for the prospects in the bottom quartile.

Roger Davis, "Gabe Gross will not be more than a journeyman at best". I'm not saying that he's going to be great, but it sure looks to me like he could be a solid regular rightfielder for somebody- .280/.360/.470 kind of hitter, with good defence.
_MatO - Wednesday, February 25 2004 @ 02:14 PM EST (#77258) #
Both Baseball America and Jim Fregosi thought the world of Andy Thompson...

Which just shows you how dubious the value of these lists are other than for the fun of debating them. Andy Thompson never did anything in the minors to justify any ranking.
_Young - Wednesday, February 25 2004 @ 02:20 PM EST (#77259) #
Interesting.

If we are trying to quantify whether the top 100 from 2000 have become good or not, why not count up all the win shares they have accumulated between 2000 to date?

That would capture some of the burnouts (Ankiel) who were good at some point during the period while giving more emphasis to players who were both good and healthy (Berkman, Mulder, Zito).

I feel that to rate Ankiel as a bust is wrong. While he is out of baseball completely now, or close to it, he was very useful for the Cardinals for almost an entire season in 2000 (not sure the exact year), wherein he pitched magnificiently until the collapse in the playoffs. While I don't hope that the Jays pitching prospects flame out like Ankiel did, would you really mind if one did come up and pitch amazingly for just one brief season?

I would vote yes. But that is my opinion.
_Max Parkinson - Wednesday, February 25 2004 @ 02:53 PM EST (#77260) #
One thing that struck me while reading the BP roundtables and then list was the constant mention of Dayn Perry's survey of minor league K rates. It seems a little sudden to go from years of "Ks are the only thing that matter", to "Ks don't mean much, just keep the ball in the park". Really, I imagine that the truth lies somewhere in the middle - Robert Dudek might have much more than this. I'm always a little wary when a ship changes course that quickly; even when Voros first started releasing his DIPS findings, people weren't saying that Ks, BBs or HRs didn't matter, and these ideas have been refined and slightly adjusted over the past 3 years.

I would hope that Dayn's findings aren't being clouded by a specific sample group, or some other correlation not currently being looked at. Maybe it would be better if the BP response to his study would be, "Hey, that's interesting - and very opposite of what we currently think when projecting MiL pitchers. Let's expand the study over more pitchers and more years, and if the results are the same, maybe we have some rethinking to do." The message from the authors this week was much more blind acceptance.

My two cents.
_Dean - Wednesday, February 25 2004 @ 03:04 PM EST (#77261) #
Strikeouts - I read that the Jays were trying get Bush to pitch for contact rather than go for the strikout everytime. I think as pitchers try to incorporate something like this you may see various stats make a sudden change and for those of us who only see the boxscores we may make assumptions that have nothing to do with what is actually going on.
Gerry - Wednesday, February 25 2004 @ 03:36 PM EST (#77262) #
I just saw in BA that DJ Hanson has a sore shoulder (impingement) and might miss the start of major league camp. They hope he will be OK for minor league camp. So we may not see him in a major league game until the end of camp, if at all.
robertdudek - Wednesday, February 25 2004 @ 03:48 PM EST (#77263) #
Is Dayn's survey discussed in the non-subscriber area of BProsp? I don't want to comment until I see the study and the specific conclusions alluded to.
_Max Parkinson - Wednesday, February 25 2004 @ 04:04 PM EST (#77264) #
Robert,

It seems that both the original study, and the BP roundtable which refers to the study, and to which I have commented are Premium, and I wouldn't feel right about quoting at any length.

The conclusions are:

1. MiL K/9, BB/9 (and by extension K/BB) rates seem to be higher for bad ML pitchers than for good ones.

2. HR rates seem to show the opposite.

To his credit, Dayn never says in the original study that these findings are at all conclusive, or predictive. That is what bothered me about the comments of the other authors (Sheehan is the one that I remember), which made the study seem much more decisive.
_Chris - Wednesday, February 25 2004 @ 04:36 PM EST (#77265) #
I think you should make a seperation between out of baseball and still in the minors. Although I haven't seen the complete list, I am willing to assume that there were a couple of names on that list who were just beginning their professional careers and were included due to their perceived upside. Perhaps they haven't developed as quickly but I don't think that the results should be skewed who haven't had the chance to play in the majors yet. This is perhaps the result of you using a dataset that is only 4 years old. To get more accurate results of a prospects chances of success, I would think that it would be necessary to use a dataset in which none of the players can still be considered prospects and have had the time to develop into major league players.

If this was done with a dataset from the 1994, I would be curious to see if the results are substantially different. Of course, this is all dependent upon arbitrary characterizations of success as a utility player can be just as important to a ball club as any regular. What we should really be looking at is just the percentage of top 100 prospects that don't have big league careers as any player who becomes a useful major leaguer must be considered to have value.
_MatO - Wednesday, February 25 2004 @ 04:37 PM EST (#77266) #
1. MiL K/9, BB/9 (and by extension K/BB

If both K/9 and BB/9 are higher then K/BB is not necessarily higher.
_R Billie - Wednesday, February 25 2004 @ 05:01 PM EST (#77267) #
Dean, I agree to an extent on that. That's why I think the Jays favour K/BB a bit more than K/9 ratio, especially when it comes to starters. They want efficient pitchers who resolve at bats quickly and get deeper into games.

Bartolo Colon for instance made a tangible change to his approach the year he won 20 games. As a young guy Cleveland fans used to constantly complain about him (not unlike Jays fans with Escobar) being obsessed with strikeouts and not lasting past the 5th or 6th inning. In 2002, his K/9 ratio dipped but so did his BB/9 resulting in an overall improvement in K/BB. He was throwing more strikes and pitching more innings although he still tends to throw a lot more pitches than say Halladay.

That's why with pitching especially, a good scouting report is very important to back up numbers. Not just one that rates the quality of the guy's pitches and quotes the velocity on the radar gun but describes his mindset in a game and whether he has a gameplan or any idea of what he's doing out there. I would say the latter is AT LEAST as important for pitchers as the tools they start with. Those who have both become stars.

David Bush is downplayed at BA because he doesn't have an obvious out pitch (though I always understood he had a solid slider) and yet he throws tons of strikes and has yet to hit a speed bump through AA. That could still mean he becomes a Josh Towers in the majors but it could also mean a Brian Lawrence or Brad Radke. Which wouldn't be bad for a second round pick signed just a year and a half ago.
_John Neary - Wednesday, February 25 2004 @ 09:27 PM EST (#77268) #
Dayn Perry's three-part study is linked below:

Part One
Part Two
Part Three

They're premium articles. I don't want to plagiarize copyrighted material, but here's a brief summary:

Perry compared the minor league records of two groups of pitchers:

Group A: (>1000 IP AND >110 ERA+) OR (>500 IP AND >120 ERA+)
Group B: (>500 IP AND <95 ERA+)

He found that group B pitchers had a lower BB/9 and higher K/9 (ergo, higher K/BB) than group A pitchers. Group A pitchers had a better H/9 and HR/9. The differences in K/9 and HR/9 were very small (on the order of one extra event per 80-100 innings pitched), whereas the differences in BB/9 and H/9 were somewhat larger but still rather small (on the order of one extra event per 35 innings pitched.)

This study is subject to a great deal of confounding, and the comparison groups aren't ideal. On the other hand, I've never seen a better study that supports the use of walk and strikeout ratios to project minor league pitchers.
Gerry - Wednesday, February 25 2004 @ 09:54 PM EST (#77269) #
This study is subject to a great deal of confounding, and the comparison groups aren't ideal. On the other hand, I've never seen a better study that supports the use of walk and strikeout ratios to project minor league pitchers.

I have not seen the study but from your description the statistical difference between the two groups appears to be non significant. In that case we are back to having no statistical system to validate pitching prospects.

I don't understand why you say the study supports the use of walk and strikeout numbers to project pitchers, unless you mean that low BB/9 and K/9 numbers suggest success at the major league level.
_John Neary - Wednesday, February 25 2004 @ 10:11 PM EST (#77270) #
Gerry,

Sorry to be so opaque. The study doesn't at all support the use of walk and strikeout numbers in projecting pitchers; in fact, it argues against it, as you rightly point out. The point I was trying to make was that I've never seen a good study that does argue for the use of those numbers in that manner.

Regarding the actual findings of the study: I'm not sure whether or not the differences in K/9 and HR/9 are statistically significant, but I would imagine that the differences in H/9 and BB/9 are. It's hard to argue that they have much practical consequence, though.

The real problem with this study is that it compares (A) guys with vey good major league careers and (B) guys with lousy to adequate major league careers. It doesn't say anything about guys who didn't make it at all, which would be a much more useful comparison group. I'd much rather be able to answer the question "Will Vince Perkins pitch at least 500 innings in the major leagues" than the question "If Vince Perkins pitches at least 500 innings in the major leagues, will he be above-average or below-average?"

John
robertdudek - Thursday, February 26 2004 @ 12:07 AM EST (#77271) #
I'd love to see the details of this study. I see the following potential problems:

1) Was the age at which the pitcher became a major league regular controlled for?

- Some pitchers develop late and become good pitchers, while others are brought to the majors early. On balance, I suspect that those who come into the bigs at a young age (under 24) have better careers. If the group A pitchers are younger, then they might put up similar K and W numbers in the minors, but, by doing so at a younger age, they ought to be considered better prospects.

2) Are relievers or years where a pitcher pitched in relief excluded from the major league and minor league data?

- We know that its much easier to put up a good ERA as a reliever. Could a large number of the 500+IP/120ERA+ group be contaminated with relievers? Relievers are very often failed starters, which could mean that they didn't have particularly impressive minor league careers (during which a large number of them would have been starters).

3) Is A ball and rookie ball data excluded?

- It's very easy to put up good K/9 and W/9 numbers in the low minors. I'd guess that virtually every successful major league starter has done so. It's not that easy to do that in the high minors. There also might be a built-in bias in that a weaker pitcher might spend an extra half season or even a full year in the high minors, allowing him a chance to improve his numbers (and therefore lift his average), as compared to a bona fide prospect who adapts to the level quickly and moves on.

4) Are the number of inning pitched at the AA and AAA level controlled for?

- To me this is the biggie. One would suspect that the true stud pitchers will breeze through the minors quickly, while the B group pitchers will have to work hard to conquer each level. That might be a much better indication than straight K, W, H and HR numbers.

5) Is there an attempt to adjust for context? Is the data compared to league norms?

- We all know by now that this is always an issue in these types of studies.
Pistol - Thursday, February 26 2004 @ 12:15 AM EST (#77272) #
http://www.baseballprimer.com/clutch/archives/00010832.shtml#comments_32
There was a good discussion at Primer today (once in a while it still happens) on getting balls put in play earlier in the count from the Globe today. COMN.
_John Neary - Thursday, February 26 2004 @ 07:04 AM EST (#77273) #
Robert: Briefly,

1. No
2. I don't think so
3. No
4. No
5. No

That's where all of the confounding comes in.

John
Mike Green - Thursday, February 26 2004 @ 01:42 PM EST (#77274) #
Sorry, John. Let me make sure I understand the study. If I have it correctly, Perry studied the minor league records of 2 groups of major league pitchers: Group A-successful and Group B- playing but below average. He found that Group B as minor leaguers walked fewer, struck out more and gave up marginally more home runs and hits. This is completely counter-intuitive, and I suspect a design problem; retrospective studies are subject to, as you say, a number of problems.

Better to do a cohort study. Take Triple A pitchers, ideally in the same league- Group A, low walk (roughly 2-2.5 per game), medium strikeout (roughly 6-7 per game); Group B, medium walk (roughly 3-3.5 per game), high strikeout (roughly 8-9 per game). Control for K/BB ratio (this should be easy), HR/9IP and age. Compare 5 year and long-term major league performance of each of the cohorts. If one has access to triple A stats from 1980 to 1990, it wouldn't be too difficult.

You could design a study to measure the impact of HR rate by taking cohorts with different HR rate and controlling for both K rate and W rate.

I seriously doubt that H/9IP is signficant, independent of K rate, but you could study it by varying BABIP rates among the cohorts and controlling for HR, K and W rates.
_Wilson - Saturday, February 28 2004 @ 08:35 AM EST (#77275) #
Even thou Kevin Witt hasn't been what you thought a 1st rounder should be, lets do some reviewing in a room that requires history. He was the the Blue Jays 1st rounder in 1994 which most regard as a weak draft for every club that yr and lets remember he was the very LAST pick of that round. There were 8 picks in front of Kevin that yr that never got to the MAJOR LEAGUES! Kevin has hit 13 Home runs in the major leagues including 10 last yr. Here are a list of the 7 that never got a taste of the major leagues.

Antone Williamson 3b Arizona State U Brewers 4th pick

Doug Million Lp Sarasota HS (FL) Rockies 7th pick ***

( passed away due to heart failure)

Mark Farris 3b Angelton HS (TX) Pirates 11th pick @@@

( went on to be QB for Texas A&M)

Jayson Peterson Rp East HS (CO) Cubs 15th pick

Matt Smith Lp Grants Pass HS (OR) Royals 16th pick

Cade Gaspar Rp Pepperdine U Tigers 18th pick

Bret Wagner Lp Wake Forest Cardinals 19th pick

Jacob Shumate RP Hartsville HS (SC) Braves 27th pick

So lets remember not to pass judgement before we have the final FACTS!! These are in the record book and no one can change this history especially for the dearly departed Doug Million!! Everyone be COOL
_John Neary - Saturday, February 28 2004 @ 10:01 AM EST (#77276) #
Mike,

Yes, your understanding of the study is correct. In medicine Perry's study would be called a case-control study; I'm not sure if that term is used in other fields.

Cohort studies are, as you say, more reliable; if I had a good source for historical minor-league stats and a lot more free time, I'd sit down and do one. I don't think that very firm conclusions can be drawn from Perry's study, but his results are nevertheless quite surprising, and I think we should consider the possibility that minor league pitchers aren't as projectable using rate stats as we've come to assume.

John
robertdudek - Sunday, February 29 2004 @ 01:32 PM EST (#77277) #
I think using Homeruns allowed per fair balls allowed (HR/(PA-W-HBP-K)) is a better measure, as it eliminates strikeouts from the equation.
How Good is a Top Prospect? | 34 comments | Create New Account
The following comments are owned by whomever posted them. This site is not responsible for what they say.