Don’t do it. Not in the first round. Unless every one is shouting “this player is can’t miss”, then maybe, but not without some trepidation.
There’s been this sentiment lately around here, and it happens every year, that “why can’t we draft a high school pitcher? The Cardinals never draft high school pitchers in the first round”. It’s almost as if people want them to draft a high school pitcher just because it would be different. There is a method to the Cardinals’ madness, trust me.
Myth: High school pitchers have higher upside. No, they don’t. I researched all the 1st round picks made between 1990-1999. The Top 10 1st round pitchers were Mike Mussina, Barry Zito, Kerry Wood, Roy Halladay, Mark Mulder, CC Sabathia, Matt Morris, Josh Beckett, Ben Sheets and Joey Hamilton, in that order. 6 out of those 10 were drafted out of college. 3% of all first round high schoolers became stars while under team control.
Adding in those who became regulars, those who produced around 2 wins above replacement for their clubs on average, while under team control, the figure goes up to 8%. 70% of high school pitchers completely bust. Why should the Cardinals gamble $1.5-$3M or more on a grouping with such a sketchy history?
As a whole, I found that all HS pitchers between ’90-’99 were worth .438 WAR per season. That’s essentially about what Brad Thompson contributes to the Cardinals every year.
To be fair, college pitchers didn’t do a whole lot better. 68% of them busted. 11% of them were regulars or better. On average they were worth .561 WAR per season. Interestingly, southpaws performed better than right-handed pitchers. On average, a college lefty was worth .721, making them a much safer pick than their right-handed counterparts. Contrast that with high school lefties (and there are a lot of them in this draft) were worth .285 WAR per season.
Hitters are much, much safer. College hitters on average were worth .932 per season. High school hitters were worth .804 WAR per season.
I know many of you will say “screw it” and keep banging the drum for a high school pitcher. Go right ahead. But the research says you’re wrong.
Draft a hitter. If there’s a good college lefty to be had, and you really want a pitcher, then history is on your side. There are several of those in this draft. But don’t draft a high school pitcher. Not unless you and everyone else is relatively certain that he’s another Josh Beckett or Roy Halladay.
Entries (RSS)
I was thinking that the likelihood of a HS pitcher in the bottom half of the first round to become a great pitcher would be even more dramatically skewed. Turns out that both Halladay and Wood were taken after the fifteenth pick, so half of the cited high schoolers. Doesn’t really add much to your point: that a college lefty or the best hitter available is a wiser choice.
here here. I remember about a year or so ago I was interested in the same question (HS vs college pitcher) and did some homework resulting in a fanpost (diary I think it was called) over at VEB. Same conclusion as Erik.
apparently some other actual sabermetricians did the same work previously too. Seems like there is lots of evidence against drafting expensive HS pitching. Now if you can get em cheap and late….
beautiful post. long wanted to see something along these lines.
fyi, port is left, starboard is right.
Hm, I think things are probably a little different now than back in 1990. I wonder if we would look at a more complete view of this. What if we specifically looked at the players drafted from around pick 15 to the end of the sandwich round from 2000 onward? How many of the high school pitchers went on to become good starting pitchers and how many of the college pitchers went on to become good starting pitchers? Or for the most recent draftees, which ones are considered better prospects?
We know that in general college players are safer bets than high school players and hitters are safer bets than pitchers. You can see this with the average WAR values. Maybe that’s because there are more busts with high schoolers, but also more studs?
My problem with this drafting strategy (which is basically the one that the Cardinals are deploying right now), is that it will give you a lot of average pitchers but not many studs. Drafting high school pitchers is risky because a lot of them are bound to bust, but maybe if you do it enough, you might get a stud pitcher. When’s the last time our system has produced a high caliber starting pitcher? Jaime Garcia seemed to be the closest to that before he got injured and hopefully he’ll bounce back from that, but we still are lacking tremendously in this part of player development.
Maybe we should focus more on how do we develop good starting pitchers rather than should we be drafting high school pitchers. I think both sides of the argument can agree at least on the fact that we haven’t done a great job of developing starting pitchers.
We know that in general college players are safer bets than high school players and hitters are safer bets than pitchers. You can see this with the average WAR values. Maybe that’s because there are more busts with high schoolers, but also more studs?
no, their were actually more college “studs” than high schoolers.
The HS pitcher riskiness line is often thrown out with regards to Porcello yet we give a free pass (on riskiness) to the Cards drafting Kozma. Want to know who the last significant HS shortstop drafted in the 1st round outside of the top 10 picks was – Royce Clayton in 1987. The Kozma pick was the ultimate in risk yet all anyone wants to talk about is how the Cards don’t draft HS pitchers because they are too risky. If risk was the issue they never would have drafted Kozma.
not so. HS shortstops are some of the least risky picks. As an average they produce .9 WAR per season.
I’d rather have Porcello b/c I’m not an idiot. But I can see why the Cardinals and 24 other teams passed on Porcello. $7M+ is a lot of $$ to give to any draft pick, let alone a HS P.
@ fpslackers – the data seems to indicate that you aren’t any more likely to get a stud by drafting a high school pitcher. In fact, the opposite seems to be true.
fpslackers, of course it’s true that “you might get a stud pitcher” if you draft a high school guy in the first round. But the point of erik’s post (and a lot of other research) is that it’s at least equally true that you’ll get a stud pitcher if you draft from the college ranks in the high-money rounds… and on average more value from that college pitcher.
And plenty of “stud” high school pitchers are drafted in later rounds, too, where they don’t waste a team hundreds of thousands of dollars when they (likely) don’t pan out. See Derek Lowe, Jake Peavey, Zach Duke, Kevin Millwood, James Shields, and so on. Note that the Cardinal’s own best high-upside starting pitcher prospect was a high school pitcher drafted in the 22nd round.
cardio,
I don’t think too many people here are giving the FO a free pass on Kozma. In fact, I think that has been one of the least popular moves, as judged by the comments I frequently read here, made in recent year drafts. And keep in mind that Erik presented data that says hitters are more predictable than pitchers. Kozma seems like more a defensive pick than a hitter, but point is risk is relative and pitchers are high on the risk totem pole.
All draft picks are risky. Put 3 names in a hat and pull one out.
Oh, absolutely. More fail than succeed. but some picks are less risky than others, which is the purpose for my post.
“the last significant HS shortstop drafted in the 1st round outside of the top 10 picks was – Royce Clayton in 1987.”
you’ve defined an exceptionally narrow group of prospects and you’ve defined the group in no particularly principled fashion. maybe 15 guys have been picked between 11 and 30 and were from high school and were drafted as shortstops in the last 20 years. several draft years in that time had NOBODY who was a high school SS who was picked between 11 and 30. in many years, few ss were picked and most were college aged. so, not finding a decent player in there is not that surprising. of those 15 or so guys, most made the majors. some were reasonable players. willie greene played for a while for the cubs, for example.
by constructing such a convoluted set of criteria and without providing a context, your blast at the kozma pick is pretty meaningless.
I’d be curious what the standard deviation is on the WAR averages.
email me and i’ll fwd you the spreadsheet.
I don’t disagree with your analysis Erik, but saying we should or shouldn’t draft a player because they fit a certain general category is much like picking stocks because of how they performed in the past. In fact, it makes more sense possibly with stocks because at least with stocks you might argue that the same people who produced past results are still involved in current and future performance.
I have read many accounts that top end position players much fewer compared to years past. Also, many supposed top-tier college pitching prospects have not performed well either. I wish a Christian Friedrich (from last year’s class) would be sitting at #19 this year but that appears not be the case in 2009. These things I would hope would give the team additional reasons to consider a high school pitcher in the first round.
I guess my point is that each player still needs to be scouted on their own merits and after that the Cards should be applying all the mathematical formulas and projections they have in order to determine the best player. Sorry for the long-winded post.
i wouldn’t deny that. i never said “never, ever draft a high school pitcher”. you can’t be that dogmatic. but your scouts, cross-checkers and probably the industry too should feel extremely, and i mean extremely comfortable with that particular pitcher’s overall future potential, and their mechanics.
If jake turner is available at 19 and the Cardinals draft and the Cards pass on him to take mike minor, i’ll freak out just as much as the rest of you.
the safest strategy is draft a hitter first and then load up on pitching. in this class that may be difficult to do.
No amount of mathematical logic will deter me from yelling “HS PITCHER” from the mountaintops
its just so sexy
History is worth considering, but if it plays as the biggest factor when your number is called on draft day, then you aren’t doing a very good job. If Jacob Turner busts, I think it will have more to do with Mr. Turner not being a major league caliber pitcher rather than someone from 1996 busting.
I’m reading “Moneyball” again and I have to ask: does anyone really think the whole drafting collegiate players exclusively plan is going well? Beane’s farm system was in shambles, which is why he started acquiring other teams’ blue chip prospects two years ago. There was a part where he mentioned that he wanted to fire all the scouts and let DePodesta’s computer run the draft. Well, the Cardinals tried that and the results were a disaster.
I just don’t think there are definite answers as to who you should and shouldn’t draft.
Thanks for the heads up on some of the data. I think why most people argue the Cards should draft high school pitchers early is because we haven’t in the past and look where it has gotten us (granted we’d look a lot better had we hung on to Haren…).
nmstar makes a good point though, is the data against drafting high school pitchers over college pitchers early that different that we should say “stay away from high school pitchers early”? I’d rather just give the preference to who we think will develop into a better pitcher (easier said than done I know). I would be concerned if the organization eliminated an entire cohort of players basically. I’m not saying they are doing this but it would be unintelligent even given the risk of drafting a HS pitcher early.
I think the biggest problem the Cardinals face right now is we haven’t been doing a great job at judging pitching talent and developing it. Not sure how we do a better job of that.
fpslackers,
FYI Haren was a college pick out of Pepperdine.
The other issue is that there is so much that can go wrong with a young kid’s arm on his 5 or so years travels to the majors.
(6 or 8 years if you are counting kids out of Latin America as well.)
for what ever it’s worth, i looked up HS pitchers from 00-04. it’s too early to call, as some guys could boost or drain their averages, but the average WAR for a HS P is .8.
Still a 72% “bust” rate, but as I said, guys still have time to turn it around…or tank it.
Swirls, I was mentioned Haren has the only good starting pitcher we’ve really drafted and developed in a long time, didn’t say he was a HSer.
Sorry – my bad.
You can cite average WAR and all the other stats you want, but the conclusion (stay away from HS pitchers) is flat wrong. The conclusion should be that the risk analysis conducted by ML teams for the past decade was flawed when they evaluated HS pitchers, and that it led them to overvalue this type of prospect.
If the Cards feel they have found a way to improve that analysis, they shouldn’t hesitate to take a HS pitcher.
Last year the only HS pitcher taken in the 1st (non-supplemental) was Ethan Martin. You could count Casey Kelly too since now that’s the direction the Sox are going with him, I guess. Or Gerrit Cole, even though he didn’t sign.
In 2007, Parker, Bumgarner, Main, Alderson, Beavan,Withrow, Porcello.
Pretty promising bunch two years later amd while the final returns are obviously pending, maybe teams are doing a better job at picking the right guys.
If the other 30 teams are running away from a strategy right when the risk analysis is improving, that may be the time to embrace it, because it implies that there may be an inefficiency to exploit.
So I have disagree with “never draft a HS pitcher”. That’s too simplistic. The Cards should take the best player available after factoring upside and risk to the best of their ability. If that brings them to a HS pitcher, so be it.
so, what’s behind this?
do HS pitchers suffer from, as jeff suggests, an extra couple years to get injured?
are teams drafting HS pitchers from a place of greater ignorance, relying primarily on subjective scouting, while in drafting college players, more reliance can be placed on college stats, making for better info and a better draft?
does the reliance on scouting mean that more HS pitchers with eye-catching skills (that may carry a higher injury risk) like 97 MPH fastballs get drafted, while fewer pitchers with subtle skills, like control pitchers or groundball machines do?
have college pitchers simply had the time to develop a more mature skillset: more control, more pitches in the arsenal?
Are you using just the first six years, or the actual six years of service time? If it is the former it is going to distort your data from what you are trying to measure.
Additionally, as you note above, HS pitchers have been much more successful in recent years (perhaps due to teams being more conscious of the risks shown in the 90s). Your endpoints end up grossly overstating the case. Look at the good first high schoolers already pitching in majors since the 2002 draft:
2002: Greinke, Kazmir, Hamels, Cain
2003: Danks, Billingsley
2004: (Homer Bailey, Phil Hughes – too soon to call)
2005: Chris Volstad
2006: Clayton Kershaw
2007: Rick Porcello
2008: obviously no one yet
And there are other who still have time to turn into something. If you ran your analysis over these years, I think you might actually find the high school pitchers outperforming the college ones.
Also, I don’t think avg WAR per season is a very good measure of “upside.” But that is a side issue.
Let’s take a look at all of the pitchers selected from picks 15-30 (since that’s where the Cards are most likely to be picking) from 2002 onward….
2002:
HS: Scott Kazmir, Cole Hamels, Matt Cain
Col: Roger Ring, Bobby Brownlie, Jeremy Guthrie, Joe Blanton, Derick Grigsby, Ben Fritz
2003:
HS: Jeff Allison, Chad Billingsley
Col: Chad Cordero, David Aardsma, Bradley Sullivan
2004:
HS: Tim Elbert, Phil Hughes, Steven Waldrop, Eric Hurley
Col: David Purcey, Chris Lambert, Glen Perkins, Taylor Tankersley, Matthew Campbell
2005:
HS: Chris Volstad, Mark Pawelek, Aaron Thompson
Col: Lance Broadway, Cesar Carrillo, Brian Bogusevic, Matt Garza, Craig Hansen, Joey Devine, Jacob Marceaux
2006:
HS: Jeremy Jeffress, Kyle Drabeck, Colton Willems
Col: Brett Sinkbeil, Ian Kennedy, Bryan Morris, Daniel Bard, Kyle McCulloch, Adam Ottavino
I’ll stop there because we won’t be able to tell as much from 2007 and 2008 yet. But look at that list. I don’t think there’s a single top of the rotation starter from the college group. The high schoolers seem to be the better prospects and have a better chance of turning into better pitchers. This is why I’m pro-draft HS pitchers early. Actually I’m not as much pro-draft HS pitchers early as I am against drafting college pitchers early. The better college pitchers are drafted near the beginning of the first round.
Matt Garza is pretty top of the rotation.
07 you have Rick Porcello, 08 Casey Kelly is looking good. I catch your drift. Still too early to call.
I think I’m against drafting any pitcher early. Take a hitter in the first round, then load up on pitching with your sandwich-2nd-3rd round picks.
I can agree with you about hitters vs. pitchers, especially college hitters. But I think it is a bit flawed to consider only the years the pitcher will be under team control. I think it is pretty obvious that the only real chance the Cardinals have to sign a top free agent pitcher is if he is one they developed.
If you look at the best active starting pitchers with 1000+ innings you will find that HS pitchers outnumber college pitchers rougly 2-to-1. College guys are more likely to make it, but also more likely to be ordinary. The goal of drafting a pitcher in the first round, if you choose to do that, should be to end up with a top of the rotation starter, not a middle reliever.
erik, you’re right about Garza who seems to be replicating his good season last year with another one this year. Maybe your strategy is the best. It seems fair more likely to draft a hitter in the first round that’s going to turn into a solid player than a pitcher. Load up on pitchers in the sandwich and second and third rounds and eventually you’ll land one who will turn into something good. My problem with this though is we have been doing this strategy to some extent and it hasn’t really worked.
Lynn, Gorgen, Mortensen, Todd, Kopp, Furnish, Daley, McCormick, Herron, Webber, Wilson are some examples of this in recent years.
Mort and Lynn seem to be the only ones with a good shot at seeing a major league rotation someday. Tough to say if either is really top of rotation material. I’d like to think they could be and both seem to be doing well this year but I doubt the general consensus among scouts would agree.
Hard to say what the best strategy is.
I would not write off Kopp like that. He’s healthy, he’s striking out batters at a good clip and getting ground balls. I’ve been hearing he’s up to 96. He might have more upside than either Lynn or Mortensen.
It’s not college, and it’s not high school. it’s just that the pitchers they have in the system lack “upside”, or they have fizzled.
Erik, IMO you have a risk of upside vs the risk of a bust.
You also have a measured risk in drafting a guy like Kozma with such a low ceiling that if he doesn’t reach his 90th percentile+ is a bust as a first rounder anyway.
So why draft so conservatively that you almost ensure a bust when you combine low ceiling, draft position, and the numbers in general that work against all prospects?
In other words if you draft a low end guy who is statistically going to fail at a high rate … say 80-85% of the time vs a guy with a huge upside that may succeed 20% of the time.
[...] It should be noted, however, that high school pitchers drafted in the first round have a depressingly low success rate. In the ‘90’s, roughly eight percent of High School pitchers became regulars for their team (2 WAR per season) and three percent became stars while under their team’s control. 70% of high school pitchers become busts, and averaged .438 WAR a season. Hitters from either the collegiate or high school ranks had a rate almost double that. (Hat Tip: Future Redbirds) [...]
[...] While they have more upside, the chances of their making the big leagues is next to nothing. This article suggests a high school pitcher has a less than 3% chance at making a significant impact in the big [...]