At the time, I bet it seemed outlandish that David Nied was left unprotected in the 1992 expansion draft. Nied had put up a 2.84 ERA in Triple-A the previous season, his third straight year with a sub-3.00 mark in the minors, and in a late-season promotion to Atlanta he had a 1.17 ERA in 23 innings. Both expansion teams wanted him badly, and the Rockies snatched him up. After the draft, Sports Illustrated quoted a scouting director who called Nied "the one slam-dunk guy in the draft." Later, SI said “he's a can't-miss prospect who will probably be Colorado’s starter on Opening Day. More than one (scout) has wondered what the Braves were thinking when they left him unprotected.” He was the 23rd-best prospect in baseball, according to Baseball America.

That was November 1992. Five years later Nied was the subject of a postmortem by Jerry Crasnick. Nied pitched 218 not-very-good innings as a big leaguer after that expansion draft, hurt his elbow, and retired young. When Crasnick found him, he was “working 10-hour days as a sales representative for Cylinder Heads International, the Grand Prairie, Texas, business owned and operated by his father, Glen.” Five years!

Shocking, but also not, because, you know. TINSTAAPP.


Rob Neyer noticed something about the Oakland A’s the other day.

According to Baseball Prospectus, they've got an 83-percent chance of winning a Wild Card. There is a considerable chance that, one month from now, the A's will be playing in the World Series. And I have to think, however far they advance in the tournament, they'll be the first team in postseason history with an all-rookie starting rotation.

I don't really have any analysis for you. It just seems like something everyone ought to know.

I've been wondering what the A’s rotation says about TINSTAAPP, or if it says anything about TINSTAAPP. Oakland’s staff is not exactly a group of top prospects who all cashed. Jarrod Parker is. Dan Straily was on Kevin Goldstein’s midseason top 50 this year, so maybe he is, but maybe he was too old to fall under the TINSTAAPP warning. A.J. Griffin isn’t; he’s a 13th-round pick who started this year, at age 24, in Double-A. Tommy Milone isn’t; he didn’t make the A’s top 20 before this season. And Travis Blackley is the guy they had in mind when they came up with TINSTAAPP: a former top prospect who missed a year with injuries, was bad, took a long route back, and is still a rookie at 29. Furthermore, the A’s need those rookies because of an injury to Brett Anderson, a top prospect who has started 38 games over the past three seasons; and an injury to Brandon McCarthy, a former top prospect who has found success only with his third organization.

What about the Rays? The Rays have graduated at least one pitching prospect into the rotation each of the past four seasons, with great success: David Price is a true no. 1; Jeremy Hellickson isn’t, but he has almost the same career ERA+ (122) as Price (124); and Matt Moore remains all sorts of good things after his first full year. Other Rays prospects, of course, have missed, but even they have contributed: Wade Davis and Jake McGee are now shutdown relievers, and Jeff Niemann produced a bunch of wins as a league-average starter getting paid the minimum. Do the misses validate TINSTAAPP?

Or the Giants, who will try to win a second World Series behind a core of pitching prospects made good: Matt Cain, Tim Lincecum and Madison Bumgarner, all first-round picks taken almost consecutively (David Aardsma broke up the streak) in the middle of the last decade. Or are the Giants just the franchise that had Foppert, Ainsworth, and Williams and reaped nothing from them? It’s hard to say.

But all of these are just anecdotes. Anecdotes that suggested (to me) that young pitching is getting safer than it used to be, but just anecdotes all the same. So I went and got some numbers.


First though, back to 1993, the year of David Nied. David Nied was, relatively speaking, one of the big successes of that year’s pitching prospects. There were 19 pitchers in Baseball America’s top 50 prospects before the 1993 season. Brien Taylor was considered the best of them, then Todd Van Poppel. Great start! Jason Bere and Allen Watson (eighth and ninth overall) had relatively long, replacement-level careers. Tyrone Hill (10th) and Kurt Miller (11th) did not. Tavo Alvarez was 17th. Brad Pennington was 18th. If we look at the six years after this list came out, the most productive pitcher in the top 50 that year was Bobby Jones, who made one All-Star team. The next two best pitchers were relievers. Nied, the guy who was selling cylinder heads for his dad within five years, produced the sixth-most WARP in the six years after these rankings came out.

TINSTAAPP is credited to the great Gary Huckabay, and, according to a 2003 piece by Joe Sheehan, it was born sometime in the latter half of that decade. In the context of the time, TINSTAAPP made all sorts of sense. Just look at those prospects! Look at how bad they all turned out! Nineteen names, and we get Bobby Jones, for goodness' sake. Wrote Sheehan, a few years later:

Then again, I'm dogmatic about TINSTAAPP. I pretty much believe that you can throw all pitchers into a bin until they're 21 years old or in Triple-A. (If a pitcher is at Triple-A at 20, that's a warning sign as well.) It's not a performance analysis thing, because even great numbers from teenagers aren't going to sway me. It's just a concession to what we know about physics and physiology, and how the two intersect at the corner of Jobe and Andrews.

In the latter half of the 1990s, there was good reason to doubt pitching prospects. Pitching prospects turned out terrible.


Okay, finally, to the numbers. I looked at the pitchers in BA’s top 50 each odd-numbered year since 1993. I limited it to odd-numbered years because there’s so much overlap from year to year, and for some reason that I couldn’t pinpoint I didn’t want to deal with the implications of that. I looked at how each group did in each season from the year they were ranked until the fifth year after they were ranked—six years in total, long enough that all would have a chance to be impact pitchers in the majors if their abilities and ligaments held, but not so long that we’re going to capture the successes who flopped about, switched teams, had surgery, switched teams, learned a cutter and finally contributed.

The 1993 class was hysterically bad. Over the six years following the BA rankings, the 19 pitchers produced just 31.8 WARP, 10 of them by relievers Troy Percival and Todd Jones. That’s 1.67 WARP per prospect, spread out over six years*. As for the rest of the prospect classes:

Prospect Class Average WARP Average IP
1993 1.67 290
1995 3.99 484
1997 4.34 392
1999 4.15 428
2001 4.57 362
2003 5.42 459
2005 5.05 528
2007 5.31 552

There’s a bit of stagger there, but for the most part it’s a line moving ever upward. Most of the gain is in quantity; on a per inning basis, the pitching prospects aren’t doing any better now than they had been (excluding the terrible 1993 class, at least). It's about one win per 100 innings. But there are more innings, which has always been most of the point.

The question of why prospects are prospectier is fertile ground for a follow-up. It could be modern medicine. It could be pitcher protection. Maybe pitchers are contributing earlier, before their arms explode. It could be (my unresearched hypothesis) that pitching in the high-offense era of the 1990s created all sorts of physical stresses, the way that growing up in poverty makes people more likely to get heart disease 50 years later. Heck, it could just be that Baseball America is getting better at picking prospects. Or it could be noise. The 2009 class, for instance, lags behind its predecessors, though the story of the 2009 class is still well incomplete.

But an extra win per prospect, if we’re talking about a real phenomenon, and especially if we're talking about a phenomenon that continues going up, sure isn’t nothing.


On the other hand, batting prospects still vastly outperform pitching prospects, at least by WARP. The position players on BA's list in 1993 produced an average of 6.7 wins per player. The 2003 class might have been the best year for pitchers, at 5.4 wins per prospect, but the hitters still outperformed them, at 7.2 wins per player. And the most recent class to matriculate through all six years, the 2007 class, has produced a whopping 8.9 WARP per player in six years. Ryan Braun alone has produced nearly as much value in six years as the entire class of 1993 pitching prospects did.

So either pitcher wins are much more valuable than hitter wins, or pitching prospects remain the worse bet.


“I really hate TINSTAAPP,” a friend told me Thursday. “It’s just lazy analysis.” I loved TINSTAAPP, and I’m not sure I don’t still love it, because sometimes it’s harder to be pithy than it is to be literally correct. But I also never knew what to do with it, and I don’t know now what to do with it. Does it mean that every pitching prospect should be traded just as quickly as a pitching-prospect sucker can be found? Does it mean that the 20th-best prospect on Baseball America’s prospect list (or our’s) is always better than the 19th-best, if the 19th-best is a pitcher? Is it, as in the dogmatic suggestion offered by Sheehan, that nearly all pitchers are equal until they’re old enough to play poker at the Bellagio? Or is it just a clever way of describing the world around us?

Probably mostly the last one. And, promisingly, the world around us seems like it might be changing.

*The 1993 class is somewhat disadvantaged by the strike, but not enough to make a noticeable mark on these numbers. The 2007 classes are, of course, a week short of completing their six-year cycle.

Thanks to Bradley Ankrom and Ben Lindbergh for research assistance.

Thank you for reading

This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.

Subscribe now
You need to be logged in to comment. Login or Subscribe
One of my favorite BP quotes of all time, courtesy of Gary Huckabay, circa 2007:

"When I first wrote that "There's No Such Thing as a Pitching Prospect," it meant two things, one of which has kind of become lost over time. Yes, it means that pitchers get hurt at approximately the same rate that methheads swipe identities and lose teeth. That's what all pitchers do, not just prospects. But it also had another meaning-that guys who are totally blowing people away in the minors like they're so many hot dog pretenders before Kobayashi are absolutely not pitching prospects-they're already pitchers, and more time in the minors only means time off the living, pulsating clocks that are their labrums, rotator cuffs, and elbows."

Perhaps the real issue is that prospect prognosticators are just poor at evaluating pitching prospects. Maybe sabermetrics has helped to make for better analysis over time, which may explain the improvement in WARP since 1993. Or perhaps what is really going on is that intangibles carry much more importance than the prospect gurus care to admit or are able to estimate. Also, there are more opportunities for production from hitters (8-9 spots) vs pitchers (5 rotation slots and a closer), so maybe its easier for hitters to outproduce pitchers with more opportunities.
Intangibles carry much more importance than most prospect gurus care to admit or are able to estimate. Bingo. It's vital, but its hard for outside eyes to recognize and measure. Physical talent can only take you so far in this game. It's east to watch a 20-year-old arm throw in the mid-90s with sharp secs and project him as a future ML arm. But that's just the raw physical stuff. In order to fully utilize that stuff at the highest level, overcoming failure and making adjustments to find sustainable success, you have to possess certain characteristics that might be impossible to identify until the player actually reaches the point of failure or mandatory adjustment. Players at the highest level play at the highest level, and its not just about tools.
Please enlighten me if I am mistaken about this, but is one wrong to assume that there must be pitcher value (call it WARP if you will), since the game is not all about offense?

Doesn't WARP, by definition, suggest that there must be a whole lot of pitchers above the replacement level?

And if there is, but not much of that value is coming from "prospects," where is it coming from?

You're close. Run suppression is half the game. But the pitcher is only one (large) part of run suppression. You could think of it as the batters "stealing" a share of the pitchers' WARP by playing and getting credit for defense.
When Andy MacPhail was the O's GM, he had a development philosophy summarized as: "Grow the arms, buy the bats". His theory was that it is close to impossible to obtain a pitcher in a trade or in free agency once they have established themselves. It is easier to obtain a hitter.

Well, let's look at the O's development path. The hyped prospects the last few years for the Orioles included the following arms: Adam Loewen, Matt Riley, Jake Arrieta, Zack Britten, Brian Matusz, and Chris Tillma. Well, Loewen and Riley got injured and essentially disappeared (I think Loewen is in the Blue Jays system as an outfielder). Arrietta and Matusz started the year in the O's rotation, got sent down because of poor performance, and are now back up as relievers (for now). Britten and Tillman have had injury issues, but are both now in the O's rotation, having joined it mid-season this year.

The hyped batting prospects for the O's the last few years have included Nick Markakis, Adam Jones[acquired in a trade], Matt Wieters, and Manny Machado. All of them are now regular starters for the O's, Machado having joined in August.

Some of the pitchers may establish themselves as rotation regulars, it's too early to rule that out. From this small sample, however, it appears that batters are the surer prospect path.

I din't include high draft picks who never even established themselves in the minors (Rowell, Hobgood), and new draft picks who are too young or new to be evaluated (Bundy, Gausman)
You know, I'm gonna take anything Andy McPhail does or says with a big rock of salt. But I think the premise is true. In fact I'm coming around to the use the 1st round pick on a big bat and then draft pitchers and more pitchers. Best player is a good rule of thumb but best bat seems more effective over time.
It seems to me as if there are two things to think about here: first, whether we're asking the right questions, and second, whether there has been a real change in prospect maturation when measured against those right questions, whatever they are. The first of these intrigues me, because I have long suspected that at least for starting pitchers, there may be something bogus about the concept of "replacement level" as applied to starting pitchers. "Replacement level," if I understand it correctly, is that level at which talent is more or less freely available. Replacement-level players are fungible: if yours goes down, you can find and acquire another who'll do approximately as well, and never know the difference (ignoring salary, "clubhouse leadership," etc.).

So consider the starting pitchers who are clustered around replacement level according to today's stats pages: Philip Humber, Chris Volstad, Jeremy Guthrie, Nathan Eovaldi. I fully believe that, at any given time for one inning, you can find other "freely available" pitchers who will do as well as these guys. But how many "freely available" (as in, not already in a major-league starting rotation) guys are there who can do that for five innings at a stretch, 30 times a year? Given the large number of pitchers in major-league rotations with negative VORPs, I suspect the answer is "not many."

So my point, relevant to this article, is: Maybe the bar for judging whether a pitching prospect has really "succeeded" in having a big-league career is too high, and/or fails to give sufficient credit to those who shoulder a starter's workload. Guthrie, to pick one, would drag down the 2002 prospect class average if you reported it. That doesn't mean there are many guys who could do better.
Great post
I think you found a much better way of raising the question than I did: Perhaps the standard used to determine a "replacement level" pitcher is way off?

If one hitter can equal the WAR of an entire class of pitching prospects, it would be foolish to draft any pitchers at all.

But am I wrong to think that would be an absurd idea?

Maybe just not in the first round. Let other teams harvest the Kershaws, Strasburgs, Lincecums, Verlanders, and Halladays.

If WAR assigns very much of the value of run suppression to fielders instead of pitchers, for anything other than exceptional plays, it would seem that the WAR for pitchers is being artificially reduced.

The longer that WAR is with us, the more struck I am out how subjective of a statistic it is, especially concerning run suppression. While it contributes to our understanding, it remains an incredibly blunt tool, albiet one that masquerades as a precise surgical instrument for many observers.

Finally I'm not clear on the reason for dismissing those who have to get past injuries, learn a new pitch, change organizations, or become relievers before they contribute. Why is their ultimate value diminished?
Just to answer your last paragraph, in my mind it is because this exercise, and the acronym, are about PREDICTING pitcher success, and that it is quite difficult to predict who will get injured, recover, change organizations, learn a new pitch, change sequencing, and THEN succeed. If someone can predict that when a kid's just graduating high school, then hats off to him/her, but I think this is trying to get at predictability, and not ultimate value (which is a very different thing that I wouldn't want to see diminished, I absolutely agree with you there).
Obvious, I speak only for myself, and not for Sam.
I also spent some of the time I was working on this wondering about the replacement level pitcher. I have nothing to contribute to the question. But, like you, I wondered.

To provide the other side of it, though, Roger Clemens produced something like twice as much value as the rest of the 1983 first round combined. So the Braun/1993 comparison may not be about replacement level definitions at all; it may just be about the relative quality of Ryan Braun and the 1993 pitching prospects.
Has anybody ever studied whether, on balance, highly-drafted pitchers or highly-drafted position players produce more WARP in their careers? The argument that pitchers are inherently unpredictable would seem to create the possibility of a market inefficiency.
Fantasy enthusiasts are already well-aware of this "pitchers are much less predictable" notion- many strategists recommend investing a disproportionately higher percentage of $ towards batters than pitchers b/c of this.

So: if fantasy players are aware, then without a doubt GM's are already aware to a much greater extent.
Here's one I found, though not exactly what you asked about:

And another:

As I searched for these I was recalling that I had read something about value across all rounds over time, but maybe that was for football and not baseball...
One thing that has changed dramatically over the past 20 years, even the past 10 years, is the 'babying' of young pitchers. I would think that should sharply reduce the # of injury driven blowups, making 'pitching prospect' less of an oxymoron. The Giants experience, or Dodgers with Kershaw, or Rays with their young rotation -- all examples of what you can do if you manage innings and injury risk
I don't know how BA's process of ranking prospects works or how it's changed over time, but there are so many things that go into rankings that depend on the prognosticator.

I'm able to communicate with scouts from all over the world at any time. And scouts are getting better at using communication tools and responding. This means that prospect journalists can have eyes on every player at all times. So I can pull open my spreadsheet of scout quotes from specific games and base my rankings on that. 20 years ago I might talk to a scout at a game for 9 innings about everything he saw that season. It's very clearly easier to give readers an idea of what a prospect is expected to become.

I also think there are things that we've always undervalued. Tom Milone isn't supposed to be a big leaguer, but he is and he's a pretty good one. Tyler Cloyd could do the same, even though he doesn't have a plus pitch.

This is just my long-winded way of saying that this is a growing process. We learn things and apply them to future evaluations. It isn't all that different from sabermetrics, really.