- FanGraphs Baseball - http://www.fangraphs.com/blogs -

Big Leaguers, Prospects, and Uncertainty

It’s no secret that I don’t think the Kansas City Royals made a very good trade last night. In my view, the price was just too high, and the Royals weren’t in a position where their team needed to give up that kind of future value to improve their chances of winning in 2013. Reasonable folks can disagree, of course. There’s a case to be made that the Royals are closer to contending than I think they are, and if KC can overtake Detroit for the division title, then the reward may justify the cost. Win-now moves can be worth it, and as teams like the Nationals, Orioles, and A’s showed last year, pre-season projections aren’t written on stone tablets and handed down from on high.

But, this morning, I’m not reading many arguments in favor of this trade that come from that angle. Instead, the defense of this trade from the Royals perspective is coming mostly from a different angle. Here’s Jeff Passan’s take, for instance:

While Shields is a known quantity – six straight seasons of 200-plus innings, a strikeout rate that approached one per inning last year and battle scars of the AL East to show for it – there is little allure in the expected. The fetishization of prospects is a baseball-wide malady, and it’s why sentiment skewed decidedly in the Rays’ favor. Granted, it should – Myers has the sort of talent that wins awards, Odorizzi looks like a mid-rotation starter, Montgomery is a high-ceiling left-hander and Leonard comes with the one tool, power, that everybody wants – but not nearly to the degree it did.

There’s a reason Tampa Bay turned down Myers for Shields straight up. There’s a reason Oakland turned down Myers for Brett Anderson straight up. Despite the scouting reports that glow and the awards he won this year, the 22-year-old Myers remains a risk. He is a safer one than most – his .314/.387/.600 line with 37 home runs between Double-A and Triple-A last season portends stardom – but any number of players have aced the minor leagues only to lag behind early in their major league careers.

Shields “is a known quantity”. Myers “remains a risk”. The Royals just traded a grab bag of who-knows-what for an ace, turning potential into performance. Myers might be good, but Shields already is. This argument gets trotted out there every time a team trades young for old. Unfortunately, this argument simply doesn’t stand up to scrutiny.

There is no question that prospects do not always develop into quality Major League players. You can look at any past prospect list and identify dozens of guys who never made any kind of contribution in the big leagues. Remember Joel Guzman? Or Greg Miller? Sean Burroughs? Ruben Mateo? Pablo Ozuna? Roger Salkeld? The list could go on forever. It is easy to rattle off the names of busted prospects who prove that the Royals very well might not have given up any future value in this move. A skeptic could look at the bust rate of even top prospects and determine that projecting guaranteed stardom for any of them was a fool’s errand.

And that skeptic would be correct, to some degree at least. As this terrific analysis of prospect evaluation shows, the overall bust rate of players ranked in Baseball America’s Top 100 from 1990-2003 was 70%. Seventy percent. Most prospects fail, and this is the kind of information that validates Passan’s criticism of the “fetishization of prospects”. Minor league success does not always, or even usually, translate to Major League success.

But it’s lazy to lump every prospect into one barrel and pretend that they all have the same odds of success and failure. Myers may be part of a group that has a 70% failure rate on average, but that does not mean that he personally has a 70% failure rate. From that same study, you’ll note that the type of players who succeed the most often are position players who rated in the top 20 prospects in baseball – guys like Myers, in other words. Those types of prospects succeeded 61% of the time. No other group of prospect made good more often than they busted. In fact, no other type of prospect succeeded even 40% of the time; the top 20 pitching prospect had the next best success rate, coming in at 39%. The top 20 position player had a rate of success that was so far ahead of any other type of prospect than there’s no real reason to lump them together with all of the others.

Of course, even with the kind of prospect that has the best outcomes, we’re still looking at a 60/40 success/bust rate, so it’s still correct to say that Myers is no sure thing. 60/40 isn’t quite a coin flip, but it’s pretty close, and would suggest that there’s about as good of a chance that Myers turns into nothing as he does an impact big league regular. And, of course, if Myers turns into nothing, then the Royals probably come out ahead in this deal. So, if we just stopped there, you could argue that this deal isn’t far off from a coin flip, and hinges entirely on whether or not Myers (and prospects in general) are overhyped and might not turn into anything in the time frame that the Royals needed to upgrade their roster.

But that conclusion requires a false pretense, and one that is implicit in nearly every article about what James Shields brings to the Royals; the idea that his performance is a “known quantity”, and that established big league veterans do not come with uncertainty of their own.

Yes, Shields has shown himself to be a highly durable and effective starting pitcher, and over the last two seasons, he’s been one of the game’s best starting pitchers. But just as we should be aware of the bust rate of prospects, we should also be aware of the bust rate of pitchers who have carried big workloads in the Majors.

For instance, 26 pitchers headed into the 2012 season having thrown 600+ innings from 2009 to 2011, and 22 of them put up similar or better performances over that stretch to what Shields has done from 2010 to 2012. Of that group, a significant portion of them either spent most of the season on the disabled list or struggled with dramatically diminished performance. A year ago, Roy Halladay, Tim Lincecum, Dan Haren, Jon Lester, Chris Carpenter, Ricky Romero, and Matt Garza would all have been described in similarly glowing terms as Shields is now, and while Ubaldo Jimenez started coming off the rails in 2011, his track record before his collapse was even stronger than Shields. Basically, a third of the pitchers who were Shields-esque (or better) from 2009 to 2011 provided minimal value in 2012.

You can repeat this exercise for pretty much any year you want. Go back and look at the track record of “proven” big league innings eaters, and you’ll find similar results. Pitchers get hurt, even ones who have a long track record of not getting hurt. Big league players perform badly, even ones who have a long track record of performing well. There is no such thing as a “known quantity” in baseball. Every player’s future is uncertain. There are varying degrees of uncertainty, of course, but the idea that a player’s future performance is a complete mystery while in the minors and then fully understood once he has success in the Majors is just completely and entirely wrong.

The best any of us can do is combine all the available information to create a forecast of future performance, then understand the variability around that forecast as well as possible. Certainly, the Royals have more information about Wil Myers than we do, and it’s reasonable to argue that their forecast for his future performance is going to be better than anything in the public domain. If they’re right about Myers being a bit overhyped, maybe his success/bust rate is 40/60 rather than the other way around, and failure is more likely than not.

But let’s not pretend that future performance variance only exists around Myers, or around prospects in general. James Shields comes with a significant amount of bust potential himself, much of it simply due to the fact that he throws a ball with his right arm for a living. Even if he stays healthy, we have to acknowledge that his past problems with the home run might return. That his exodus from a team that shifts on every play might lead to more balls falling in for base hits. That he just might pull a Ricky Romero and fall apart with no real explanation.

Wil Myers is a risk, but so is James Shields. So is Albert Pujols. So is Justin Verlander. So is everyone.

When we start using labels like “prospect” or “proven veteran” to describe players, we lose that reality. Myers and Shields both have the chance to be good, bad, or anything in between. Let’s not let the terms we use to describe player types obscure that fact. The Royals didn’t trade a lottery ticket for a paycheck; they traded a few lottery tickets for a scratch-off card. They probably did reduce their overall performance risk for 2013, but it didn’t go to zero. Let’s not pretend that we know so much about projecting the future of Major League players that we create an artificial divide where one does not exist.

Prospects come with uncertainty, but so do Major League players. Everyone is a risk. You weigh that risk against the potential rewards, and you figure out which trade-offs are worth making. Once you cross over into treating some players as non-risks, though, you’ve stopped evaluating players properly. And then you make trades like the one the Royals just made.