March 10, 2005

"Pretty Even??"

When I posted the latest poll ("How do you rate the Schmidt trade?") on our web site, I expected to see one vote for either "Pretty even" or "Great for Cleveland!" -- from Mike Stein, of course.  I wouldn't have been surprised to see two votes for "Pretty even", given Chris Luhning's ill-explained position on our message board.  But when I saw three votes for "Pretty even" and one for "Great", I knew it was time for a special edition of the FTDOTC page.

Now, I understand that there are certain people who will say that ANY trade is fair, no matter how ridiculously lopsided it may be.  Chipper Jones for Ryan Dempster?  That was fair.  Jim needed to dump salary.  Barry Bonds for Brad Wilkerson?  Yeah, that was okay, too.  I mean, Wilkerson's a decent hitter, and Stein saved a lot of money with that trade.  The Edmonds trade?  So what if Edmonds wasn't worth nearly $11.5m?  Litchfield got John Olerud out of the deal.  These people simply cannot be reasoned with.  Of course, that won't stop me from trying.

First, let's take a look at what Cleveland gave up.  Using Tony C's 2004 MLB spreadsheet, we can sort by innings and get every pitcher in baseball with more than 150 IP.  Using that subset, we can then see that Schmidt is ranked #9 among all pitchers in VORP, #9 in ERA, #4 in DIPS, #3 in batting average against, #5 in OBP against, #3 in SLG% against and #3 in OPS against.

In other words, he's pretty good.

I assume that I don't have to make a case for Schmidt being among the top starting pitchers in baseball, but given some of the comments I've seen on the message board today, I think it's probably best to point out just how good he is.

Trading ace starters isn't unprecedented, of course.  In fact, it happens a little too often in this league.  Using VORP as our standard of measurement, I can find nine starting pitchers with a VORP of at least 60.1 that have been traded over the past two-plus seasons alone:

• Roger Clemens (2005): Traded twice: Once for Tim Hudson, Ryan Klesko and Jerry Hairston, and once for Brandon Webb and Guillermo Mota.
• Mark Mulder (2004): Traded along with Tony Graffanino and Karim Garcia for Geoff Jenkins, Junior Spivey and Jhonny Peralta.
• Roy Halladay (2004): Traded twice: Once along with Troy Percival for Matt Clement and Billy Koch, and once along with Ramon Ortiz and Ramon Vazquez for Jose Reyes, Justin Duchscherer, Jung Bong and Joe Mays.
• Esteban Loaiza (2004): Traded along with Keith Osik for Shea Hillenbrand, Claudio Vargas and Jason Stokes.
• Tim Hudson (2004): Traded along with Gary Bennett for Victor Martinez.
• Derek Lowe (2003): Traded along with Armando Rios and Horacio Ramirez for Tim Hudson and Reggie Taylor.
• Barry Zito (2003): Traded along with Troy Percival, Steve Karsay and Khalil Greene for Casey Kotchman, Kris Honel, Ben Broussard and Brad Penny.
• Roy Halladay (2003): Traded twice: once along with Mark Redman and Dan Plesac for Matt Morris and Javy Vazquez, and once along with Ryan Anderson for Ichiro!, Drew Henson, Hanley Ramirez and Corey Bradford.
• Tim Hudson (2003): Traded along with Reggie Taylor for Derek Lowe, Armando Rios and Horacio Ramirez.

That's twelve trades in total.  Without getting into a bunch of hardcore number-crunching, the eyeball test tells me that nine of these dozen trades look pretty fair in retrospect.  For the most part, teams trading these ace pitchers received one of three things:

1. One or more players upon which to build for the future.
2. Salary cap relief for the following or upcoming season.
3. An exchange of pitching talent for the current season.

In all twelve trades, these teams received a combination of the three.  The Schmidt trade, then, is the only trade of an ace pitcher over the past three seasons in which the team trading the pitcher received only ONE of the benefits in the list above.

Timing-wise, three of the dozen trades above were made in the pre-season, giving the teams trading their star pitcher extra money to spend in the auction/draft.  Two of the trades were made in Chapter Three, four were made in Chapter Four and two were made at the deadline in Chapter Five, giving the teams trading these pitchers the maximum 20/20 hindsight available.  The Schmidt trade, then, is the only trade of an ace pitcher over the past three seasons in which the team trading the pitcher gained NO hindsight whatsoever.

Now, perhaps the package that Cleveland received in this deal was so overwhelming that no salary cap relief or exchange of current talent was necessary.  Perhaps this package of talent was so incredibly enticing that they couldn't possibly wait another chapter.  So let's take a look at what Cleveland received.  (Note: We'll discount Ryan Dempster and Kyle Farnsworth, as those two fliers hardly merit a mention.)

Clifton Phifer Lee is 26 years old, and will be turning 27 in August.  His MLB career stats to date:

241.2 IP, 235 H, 37 HR, 109 BB, 211 K, 4.88 ERA

And his career minor league stats were:

380 IP, 305 H, 39 HR, 175 BB, 423 K, 4.14 ERA

Or, broken down into rate stats:

Major League: 8.8 H/9, 1.4 HR/9, 4.1 BB/9, 7.9 K/9
Minor League: 7.2 H/9, 0.9 HR/9, 4.1 BB/9, 10.0 K/9

Obviously, what we have here is a decent-looking left-handed pitcher with a live arm, who walks a lot of batters and also whiffs a lot of them.  At his age, he may have already entered into his peak performance years, or like some late-blooming lefties, his best years may be in front of him.  At this point, there is no way to tell for sure.

And that's the problem with young pitchers (or old pitchers, for that matter.)  Like Forrest Gump and his box of chocolates, you never know what you're gonna git.  Lee may be on the verge of a breakthrough in his career, or he may be nothing more than a future #4 journeyman.  At this point, who knows?

Let's see if we can predict the future by using the leading tools at our disposal.  Sometimes, you can get a feel for which direction a player's career is heading by looking at how he most recently performed and comparing that to the previous period of time.  So let's look at Lee's second-half numbers from last year.  A vast improvement from the first half to second half can sometimes portend better things ahead.  But here's how Lee's 2004 went:

April: 28.2 IP, 27 H, 17 BB, 20 K, 3.14 ERA
May: 31.0 IP, 23 H, 16 BB, 30 K, 2.61 ERA
June: 29.0 IP, 34 H, 11 BB, 22 K, 5.90 ERA
July: 29.1 IP, 35 H, 12 BB, 27 K, 7.06 ERA
August: 31.1 IP, 39 H, 14 BB, 28 K, 7.18 ERA
September: 22.2 IP, 25 H, 11 BB, 28 K, 7.94 ERA

Ouch.

Now, perhaps Lee ran into the dreaded "dead-arm" syndrome at around the all-star break.  Maybe he swallowed a bad ice cube somewhere around the beginning of June, and it affected him through the middle of September.  Maybe he just wasn't conditioned to pitch more than 60 innings.  Maybe he worked extra hard this winter and is ready to go the distance in 2005.  Maybe monkeys will, in fact, leap out of my ass at any given minute now.  It could happen.

Unfortunately, I don't have predictions from all the leading prognosticators, gurus and psychics.  But I do have access to Baseball Prospectus' PECOTA numbers.  If you're familiar with PECOTA, you know that they try to match up players to the most similar types of players in the past, then estimate how each player will perform, based partly on the player's own history and partly on the trends of the player's most similar peers at similar junctures of their careers.

According to PECOTA, Lee's top five most similar peers are Denny Lemaster, Denny Neagle, Bob Ojeda, Ted Lilly and Dan Schatzeder.  Here's how PECOTA sees Lee's 2005 season:

What this means is that Lee's most likely performance for 2005 is a 4.76 ERA and a VORP of 19.4.  There is a 10% chance his VORP could be as high as 48.3.  There is also a 10% chance it could be as low as -12.0.  The fact that there is such a wide disparity here (from an ERA of 3.35 to 6.52) means that PECOTA doesn't know what to make of him.  At this point in his career, it is equally likely that Lee could either step up and become a nice #2 starter or become a waste of roster space.

In other words, even a system as complex as PECOTA deems Lee to be a bit of a gamble; worth taking a flier on, given his upside potential, but not worth gambling too much, given his downside risk.

PECOTA is also not very optimistic about Lee moving beyond 2005.  It projects VORPs of 15.6, 13.6, 9.2 and 8.0 from 2006-2009.  It also projects an attrition rate (i.e. the rate of a player's playing time decreasing due to either injury or poor performance) of nearly 50% by 2008.  That's not good.

Of course, it's possible that PECOTA could be wrong, and Lee will outperform that 90th percentile forecast.  It doesn't happen often, but it happens.  The question is: What are the odds of that happening?  Probably not great.  But the bigger question is: Would you pony up Jason Schmidt if the bet on the table was whether or not Cliff Lee will beat that 90th percentile forecast, knowing what we know today?

Of course, even if Lee meets his 75th percentile forecast, that's still pretty damn good.  Trading Schmidt for a $100,000 pitcher with a 3.95 ERA and a 34.4 VORP would be a pretty good trade.  The problem is, those numbers are not guaranteed at this point, because we have no hindsight advantage.  (But I'll get to that problem in a minute.)

Among the arguments cited in favor of Lee are the fact that he ranked among the top 30 prospects back in 2003 (according to Baseball America), and he owned a lofty 10.0 K/9 rate in the minor leagues.  After all, top-ranked lefty pitching prospects who post strikeout rates like that are a guaranteed lock for future success right?

Well...here are a few other lefty pitchers around the same age as Lee who owned high minor league K/9 numbers and also ranked among BBA's top 30, along with their minor league numbers:

Bruce Chen: 7.9 H/9, 2.7 BB/9, 9.7 K/9, 3.52 ERA, now age 28
Ryan Anderson: 7.7 H/9, 5.4 BB/9, 11.9 K/9, 3.94 ERA, now age 26
Matt Riley: 7.9 H/9, 3.6 BB/9, 10.3 K/9, 3.48 ERA, now age 26
Brandon Claussen: 7.9 H/9, 3.1 BB/9, 9.4 K/9, 3.38 ERA, age 26
Wilfredo Rodriguez: 7.7 H/9, 4.8 BB/9, 8.8 K/9, 4.09 ERA, age 26

So, apparently, minor league strikeout rates and Baseball America's collective opinion is no guarantee of future stardom.

Another argument employed by Stein is that because past gambles like Roy Halladay, Jake Westbrook and Brandon Webb paid off for him, taking a gamble on Lee might pay off, too.  The problem with that argument is that neither Halladay (5.9 K/9) nor Westbrook (5.3 K/9) owned stellar strikeout rates in the minors, which contradicts his first argument.

Also, by the time they had celebrated their 27th birthdays, Halladay had already won an MLB Cy Young award, Webb had come close to winning one himself, and Westbrook had enjoyed a terrific (though unlikely to be repeated) year (14-9, 3.38 ERA, 216 IP.)

Clearly, there is no way of knowing for sure what to expect from Cliff Lee in 2005 or beyond.  All of which makes him a VERY risky pick-up at this point in the season.  Even if Lee got off to a great start (as he did in 2004), there is no guarantee that he won't completely flop the rest of the way (as he did in 2004.)

Therefore, if Lee really was the one and only target Stein was after, it would have been in his best interest to wait at least one more chapter before pulling the trigger, to see if Lee at least escapes spring training without injury.  And surely, Bobby Sylvester would have just as happily made this trade in August as he would have in March, as this trade all but guarantees a championship.

But that's not the only reason to have waited.  Waiting also allows other owners to get in on the bidding.  Perhaps another, much less riskier, player would have been offered.  Or, better yet, risk could have also been lessened by taking several players instead of just one (and yes, I am completely discounting the minimal impact of the other players involved in this trade.)  Perhaps a bidding war would have ensued between division rivals, fearful of having to face Schmidt the rest of the way.  Perhaps someone would have panicked over the prospect of finishing behind Billy Baseball in a division race and offered a long-term impact prospect like Delmon Young.  Now, we'll never know.

Aside from those two reasons above, waiting would also give the Rocks more of a chance to win some games early in the season and possibly avoid penalties next off-season.  Having Schmidt for only one chapter, and replacing him in the rotation with Lee, certainly won't help that cause.

In conclusion, I have three good reasons why this trade was grotesquely lopsided:

1. Cleveland should have received much more in return for a top-tier pitcher like Schmidt.  In recent BDBL history, every team that has traded a similar pitcher has received far more in return than what Cleveland received in this trade.
2. Lee is no sure thing.  In fact, there is no way to tell, given even the most sophisticated prognostic tools, how well or poorly Lee will perform in 2005 and beyond.  There are literally dozens of other pitchers in both the major and minor leagues who are comparable to Lee, and all are equally risky.  There is nothing special about Cliff Lee.
3. Cleveland gained absolutely nothing by pulling the trigger on this trade when they did.  In fact, there is no rational reason in favor of making this trade so early.

My question is: What GOOD reason is there to proclaim this trade as "fair" -- or, laughably, as "terrific for the Rocks?"  Knowing what we know right now, what GOOD reason is there to believe that Cliff Lee will, without a doubt, become an all-star pitcher in 2005?  Knowing what we know now, what GOOD reason is there to think that Stein made the wise decision by making this trade in March rather than June, July or August?  I anxiously await some answers.