In response to yesterday’s post about the risk-reward balance of prospect valuation, commenter “Hunter fan” made the following observation:
Another reason lists could be so pitcher heavy is that apparently pitcher, for some analysts, are easier to project. Sickles just did a few articles on this. Top pitchers, almost without exception, he rated as A or B+ prospects. The position players were all over the place, with several good position players being B- or C level prospects.
Just food for thought.
Pretty interesting food, actually. I’m not sure why I hadn’t thought of this earlier, but on the surface, this comment seems to pass the smell test. Kids in high school can throw in the mid-90s, and that’s a pretty easy thing to scout. They can throw nasty breaking balls that no one can hit. Likewise, there are always a few college arms who command their pitches so well that they’re considered to be nearly Major League ready before they ever sign a pro contract. With these types of premium pitching prospects, there’s not really a lot of projecting to be done – the forecasting involved is more along the lines of whether they’ll be able to stay healthy or not.
This isn’t really true of hitters. Very few high school kids possess the physical skills of a Major League player. A large part of hitting is how well you can drive the baseball, and game-usable power often doesn’t develop until a player is in his early-20s. Likewise, hardly any teenagers have the understanding of which pitches to swing at and which ones to lay off, as that is learned through experience and repetition – things that only come with time. Underdeveloped power and overly aggressive hitting strategies mean that very few young kids are capable of stepping right into the Major Leagues and performing well offensively. The position players who do make the big leagues before they can legally drink are usually those whose athletic abilities allow them to contribute in the field or on the bases – a disproportionate number of teenage position players in the big leagues have been shortstops and center fielders.
Amateur pitchers flash the kind of skills needed to become top-notch Major League arms, and in many cases, those skills can translate directly to the Major Leagues in a short period of time. Amateur position players can show athletic ability and good hitting mechanics, but rarely do you see a HS or even college hitter who is already displaying the kinds of skills necessary to hit well against big league pitching. Position players require a bit more projection – how will the player fill out as he gets older, can he adjust to facing good breaking balls on a regular basis, etc…
It makes sense, and would help explain why teams have been willing to invest so much money in high draft picks on pitching prospects despite knowing that their flame out rate is significantly higher. As we talked about yesterday, our confidence in our evaluations goes up as we gain information, and if scouts feel that they have more information about a pitcher’s Major League tools than they do a hitter of the same age, it’s natural that they could prefer the pitcher, even with the greater chance of injury. Of course, a lot of things make sense when you think about them, but are debunked once you actually start looking at the data. So, let’s try and look at some data.
We’ll start off by just borrowing some research from Jeff Zimmerman, who looked at the relative percentage of players from Baseball America’s All-Time Top 100 list that showed up on the 2010 leaderboards. There were 13 pitchers that posted a WAR of +5.0 or higher in 2010, and all 13 appeared on BA’s list at one point or another. There were 18 pitchers who posted a WAR of between +4.0 and +5.0, and BA had identified 72% of those. They got 70% of the 20 arms who posted a WAR between +3.0 and +4.0. This is a pretty good showing for BA, as prospect evaluation isn’t easy, and they consistently nailed a large majority of the best arms in baseball.
On the hitting side of things, they didn’t fare quite as well. Again, 13 players posted a WAR of +6.0 or higher, and BA identified 77% of those. They stayed in the mid-70% range for every tier down to +3.0 to +4.0 WAR, and as Zimmerman’s final table shows, the only real difference in the results seems to be at the very top. In most tiers, they got close to the same percentage of hitters and pitchers, but when it comes to +5 WAR or higher players, they did significantly better with pitchers.
We’re dealing with pretty tiny samples, though, and only one year of results from one source. We can’t draw any conclusions from this information. So, let’s look back at Sickels’ rankings, where the source of the original comment came from. A couple of weeks ago, John posted a review of his prospect grades for pitchers who rated in the Top 25 in baseball in WAR last year. Of the 25 guys who posted +3.9 WAR or higher, Sickels had placed a B- or better prospect grade on 23 of them – only C.J. Wilson and Doug Fister fell through the cracks, and they’ve both taken some pretty anomalous career paths to success. Twelve of the 25 pitchers received an A- grade or better. By and large, Sickels was able to identify most of the best pitchers in baseball last year while they were minor leaguers.
He then looked at the top 30 position players from a year ago, and the results were somewhat less impressive. Most of the guys on the list got B- or better grades at some point, but Sickels acknowledges that he missed on Matt Kemp and Robinson Cano, notes that Matt Holliday was a late-bloomer, and underestimated the abilities of Brett Gardner and Shane Victorino, who were pegged as fourth outfielders. We’ll give him a pass on Jose Bautista because what he’s doing is basically unprecedented and no one could have seen this coming.
He still got most of the best hitting prospects, just as Baseball America did, but the proportion of misses is higher and more severe.
Again, not enough information to draw any firm conclusions, so I decided to look at how very young kids perform in the big leagues. I grabbed all player seasons since 1982, and then filtered out only the ones where the player was 21 or younger and had at least 50 IP/150 PA. There were nearly identical samples of each population – 132 pitcher seasons, 123 hitter seasons. That was not planned, but it’s a pretty nice coincidence, as it gives us very similar samples to deal with.
The 132 21-or-younger pitchers were pretty pretty good (thanks to guys like Dwight Gooden, Bret Saberhagen, Fernando Valenzuela, and Madison Bumgarner), throwing just over 15,000 innings and posting an ERA- of 99, supported by an equal FIP- of 99. Overall, this combination of above-average pitching over a decent number of innings added up to +212 WAR, or an average +2.53 WAR per 180 innings pitched.
The 123 21-or-younger hitters also weren’t bad. Alex Rodriguez, Albert Pujols, and Ken Griffey Jr were all monsters at an early age, and helped lead the group to +192 WAR in just under 49,000 plate appearances. That averages out to +2.36 WAR per 600 PA, just a bit lower than what the young pitchers put up.
Again, the gap isn’t large enough for us to draw any real conclusions. There have been some really good young pitchers, and some really good young hitters. The young pitchers have been slightly better, and may be slightly easier to identify, but the information isn’t overwhelmingly in support of the idea that premium young pitchers are just easier to identify than young hitters.
The idea still makes sense, though, and I wouldn’t be surprised if a more thorough study ended up supporting the claim. It’s officially on my to-do list for this year, and is a topic I’d love to see discussed in more depth.