PECOTA is not infallible.
I learned this the hard way at last year’s League of Alternative Baseball Reality draft, NL version. The 2005 draft marked the third time I used PECOTA to guide my drafting in an experts’ fantasy league, following stints in the 2003 Tout Wars and 2004 LABR National League drafts. Last year was notable for three huge draft-table whiffs: D’Angelo Jimenez for $17, Endy Chavez at $11 and Cristian Guzman at $9. All three of those picks came at values that were actually below the dollar figures spit out by the Player Forecast Manager, a fantasy-specific drafting system that runs off the projections generated by BP’s PECOTA projection system.
The PFM (newly updated today) takes PECOTA’s raw output and adjusts for an array of factors, from position scarcity to category scarcity (saves and steals especially) to many others. But as with any projection system, it pays to inject some common sense into the mix as well. Jimenez showed strong results at the plate heading into the 2005 season, so that flop was tougher to predict. Chavez and Guzman simply stink, no matter how you look at them. Many of the major leagues’ top basestealers bring little more than speed to the table. But as long as they show at least one other skill–doubles power, on-base ability, a knack for finding incriminating photos of their manager–they’re going to keep their jobs, bat near the top of the lineup and generate plenty of steals and runs scored. Players like Chavez or Guzman are so limited in their talents that even the most speed-obsessed manager is going to yank them from the lineup eventually if they continue to dwell below replacement level. So, lesson learned. Don’t draft terrible players.
Heading into this year’s draft, I planned to heed that mantra. At the same time, the PFM consistently values closers and base stealers at higher prices than most other projection systems I’ve seen. It also does a great job of finding gems: I snagged Chris Capuano at $3 last year, jumping on his brisk strikeout rate and first shot at a full season in a major league starting rotation, gaining one of LABR’s biggest bargains in the process. The trick as I saw it, then, was to follow the PFM’s general guidance, but shy away from specialists with no discernible talents beyond their job status as closer or their great speed. As I did last year, I’d go wherever I could find value, looking for players at prices several dollars below the PFM’s projections.
Of course fantasy drafts always involve a good deal of game theory as well. With that in mind, I also brought the Baseball Forecaster’s projections with me. Run by Ron Shandler (frequently discussed in Sam Walker’s fun new book Fantasyland), the Forecaster includes fantasy projections on a wide range of players. Shandler’s projections have been around for 20 years and contain a sabermetric grounding that’s similar to BP’s projection system. For the purposes of this draft exercise, I figured they’d also make for a handy barometer of what other experts might be expected to pay for players. Find the big gaps between PFM and the Forecaster and draft accordingly. For better or for worse, that was my strategy. That, and don’t draft terrible players.
The draft started with a slew of power hitters going by the board. The PFM sees power hitters as widely available in this year’s American League and depresses their value accordingly. That left me reluctant to go after the Ramirezes, Ortizes and Hafners for $30-$35. After a few rounds, Team BP and last year’s AL-LABR champ Jason Grey were sitting with nearly empty rosters. Eventually I jumped in, pouncing on the two areas I figured would be most undervalued (at least according to the PFM): pitching and steals. My first three picks (actual price paid, PECOTA, Forecaster values in brackets):
That quickly I had my staff ace, an ace closer and a basestealing ace who could get half the steals I’d need to win the category by himself, at a total projected profit of $24. Johnson and Ryan own the peripheral skills to be strong picks in any format, rotisserie or otherwise. Podsednik hit zero homers last year and might appear to be a one-trick pony with his speed and frequent stolen-base attempts. But Podsednik also put up a solid .351 OBP last year, with the kind of flashy glove that’ll keep him in the lineup during slumps. This wasn’t Endy Chavez, and I felt good about the pick.
After another lull, I grabbed four more rapid-fire picks:
Polanco and Bonderman were my first two picks where the PFM and Forecaster both agreed I got a bargain. Polanco finished second in the majors in batting average behind Derrek Lee last year and is the rare player who has been able to hit for a high average year after year, a rare feat given that stat’s variability. I’m expecting Bonderman to break out in a big way–a 200 strikeouts, 3.75 ERA or lower way. I have a soft spot for Loaiza ever since he led a Strat-O-Matic team of mine to an improbable championship off his superlative 2003 performance. I like him in pitcher-friendly Oakland with a good defensive outfield behind him. Crosby was a bit of a reach going by the projections, but shortstops were starting to come off the board and this was a gut pick for a breakout year by Crosby, batting third in an improved A’s lineup.
All fine and dandy, but I’ve still got virtually no power. How far can I take the draft-for-value strategy before I’m forced to overspend for one of the dwindling number of thumpers out there?
That’s three Tigers in my first 11 picks. The jokes were already flying–maybe in my yearning for the Expos to return I got confused and thought Detroit was a Canadian city, being north of Windsor, Ontario and all (by the way, fellow Canucks, how ’bout our boys, eh?). Fact is, Granderson has an excellent blend of skills that translate well to both fantasy and real-life success. I’m not thrilled with the Conine pick even though it’s listed as a bargain by both PFM and Forecaster standards. Even pushing 40, Conine’s still got enough on-base ability and likely playing time to avoid the terrible player tag. Call this a case of remorse over filling a corner slot–albeit cheaply–with a mediocre player, when someone better could come up for a similar price later on (Dallas McPherson later went for $3, just as I feared). I still have very little power, and the hour’s getting late.
Much better. I filled my mostly vacant outfield, adding a big-time power threat who I like for 30+ homers this year in Gomes and a power-speed player in Bradley who’s at the right age, in a new environment, where a breakout could be imminent. And hey, it’s another Tiger! Actually Monroe’s an underrated power source who can rack up two-thirds of the homers likely to be garnered by elite power threats, for less than half the price. I also have three A’s at this point. Neither occurrence surprised me much, since those are my two picks for most improved teams in the AL this year and thus a good source of underrated talent.
Two trends emerge from this group of five. First, we see some huge differences of opinion between the PFM and Forecaster in Upton, Patterson and Verlander. Much of this difference of opinion is a statement on playing time. Forecaster sees plenty of ability in Upton and Verlander, but doesn’t expect either player to see much time in the bigs. Its $15 downgrade of Patterson compared to the PFM is partly a statement on Patterson’s horrific 2005, but also on the Orioles’ crowded outfield.
Patterson and Upton could make or break my team. Upton’s a starting middle infielder on this team, so I’ll need to get at least half a season from him (gentlemen, start your unfounded Julio Lugo rumors!). Picking Patterson is a nod to the PFM’s $24 projection, so far beyond any other projection system I’ve seen that it’s goofy enough to be worth a shot, a la Wily Mo Pena in 2004. But it’s also a statement on how ballplayers develop. Very few players progress in a perfect bell curve pattern. The MLB population as a whole will see development in their early 20s, followed by a peak from ages 26 to 29, a plateau for a couple years and a gradual decline from there. But any one given player may be subject to all kinds of fluctuations. Patterson was a roto terror in ’04, with 24 homers and 32 steals. Of course he was also more than 10 runs below replacement level last year, the second-worst mark in the majors. By definition, that makes Patterson a truly terrible player, at least last season.
Patterson is my biggest gamble of the draft. On the other hand, I’ve finished 2nd, 5th and 2nd in my first three years in experts’ leagues. Solid, respectable showings? Absolutely. Does anyone remember who finishes second? Hell no. Call this a go-for-broke pick. Cubs fans, you may now e-mail me your ridicule and scorn.
Oh and the second trend from this batch of players? Two more Tigers! I now own 3/5 of the starting rotation. I’m totally speculating on Dave Rozema as he comes up for a buck.
In describing Soriano in Baseball Prospectus 2006, I wrote: “He’s one of the few players the M’s consider close to untouchable, with both the statistical track record and the stuff to back up his rep. Rotoheads, here’s an August 1 closer if the Mariners are out of the race and flip Guardado, a pretty likely scenario.” I’m all about accountability. So when LABR maestro John Hunt tried to sneak Soriano through late in the draft, I went to $5 to get my guy. If others take my advice and Soriano flops, at least I can say we all failed together. If he succeeds and you go on to win your leagues, I accept charitable donations.
Cuddyer and Laird are moderate-upside players who could hit 25-30 homers between them if everything goes right and/or Rod Barajas goes back to being Rod Barajas. And Brandon Inge? Not only is he my seventh Tiger, he’s also my kind of golfer, a swing-as-hard-as-you-can bomber who doesn’t sweat the ball’s trajectory or the possibility of it rolling all of 30 feet. I have him pegged as a 23-homer sleeper too.
Once I drafted Upton, I figured I might as well roll the dice on the two middle-infield prospects PECOTA most wants to smooch, Kendrick and Pedroia. If any of those three crack the big club’s starting lineup, watch out. Gotay was the same idea, writ smaller. Wigginton and Ambres are useful backups with some upside if they get playing time. Thames is my eighth Tiger, because seven just wasn’t enough.
So where does this team stand? Great pitching with Johnson, Loaiza, Bonderman and Westbrook at the top of the rotation. Top-notch closer in Ryan. Speed to burn with Podsednik, Bradley and Granderson, with Upton and Patterson threatening to make it no contest for steals. Team BP still came up a bit short in the power department, though, especially with the prices others were willing to pay for the big boppers. It’s clear that a pitching or speed for power trade will likely happen down the road. A series of opportunistic deals yielded a runner-up showing last year, so there’s hope.
If nothing else, starting the season without Endy Chavez and Cristian Guzman around is a reason for optimism.