ROTISSERIE: Lessons from an FPAZ exercise

2012 seems like another time. The era of the power pitcher was still nascent, and the “tightball” era had yet to begin. R.A. Dickey won the NL Cy Young award. R.A. Dickey! Mike Trout and Bryce Harper were both rookies.

At its annual conclave of fantasy minds, First Pitch Arizona, BaseballHQ embarked upon an experiment: to draft a set of players that you would commit to for five years. Predicting their 2013 value was not enough—you needed to see well into the future. While the actual selections are of interest, it’s the lessons learned from this exercise that are most important (besides the bragging rights for the winner).

Some of the best minds in fantasy baseball participated, plus this writer. The all-star lineup, in addition to this columnist, included Todd Zola, Rob Gordon, Tim Heaney, Lawr Michaels, Derek Van Riper, Brian Walton, Tristan Cockcroft, Steve Moyer, Nick Minnix, Andy Andres, and Jock Thompson. The draft went six rounds, so 72 players were selected.

Our first step is to try to validate the skills of the people making the selections. Since no projections can ever be perfect, the simplest way to do this is to go by the expert’s own rankings—if their opinions were valid, players taken earlier should have earned more than those taken later. (Note: a complete list of the selections and the results can be found here. Ron Shandler’s write-up of the exercise is here.)

Aside from a surge in Round 4 (thank you, Paul Goldschmidt), they played out as expected. We’ve passed out first test: the experts seem to know what they’re doing.

The Lessons

Age should be a big factor, right? Nobody was going to touch David Ortiz, who was at least 37 entering the 2013 season. More than half the players taken were 24-28; there were four 20-year-olds selected and somewhat surprisingly, 11 players over 30.

However, there were no clear results based on age. In fact, as a group, the 30-31 cohort outperformed the 22-24 cohort, even though in theory the latter group was just entering their peak years and the former group would soon be past their peak. As a group, the 25-year-olds had the best overall performance; however, the 26-year-olds had one of the worst performances.

This defies the traditional aging curve, right? On average, players improve until their mid/late 20s, decline slowly into their early 30s, then fall off a cliff as age 35 approaches. However, age wasn’t the only factor in this exercise—as much as age, the drafters were selecting for talent, reliability, and growth potential.

To some extent, the ability to stay healthy was important. Of the players in the bottom third, almost half missed the equivalent of an entire season, led by David Wright’s 483 days on the DL. However, there was little correlation overall. Clayton Kershaw and Giancarlo Stanton were both in the top 20, despite missing 124 days and 164 days, respectively. Billy Butler and Jesus Montero missed 7 days total between them, yet both were in the bottom 20.

There were no clear trends based on position, either. Catcher and third base returned much lower value than the other positions; however, there were only five 3B drafted in total, and three of them (Wright, Pablo Sandoval, and Brett Lawrie) missed significant time with injuries. There is nothing inherently dangerous about third base, so this is likely small-sample variation.

The catcher position may be a different matter. While catchers do not hit the DL any more often than other positions, playing catcher does take a physical toll. It is interesting to note that two of the six catchers in the draft (Joe Mauer and Carlos Santana) changed positions permanently during the five-year period, and two others (Wilin Rosario and Jesus Montero) were out of baseball, in no small part because their defense was subpar. Catchers are just much harder to project.

Experience should matter, right? Established players should be more predictable, with their base skill level established. Players with less experience should be harder to predict and more volatile. It turned out that players with less than two years’ experience did show a bit more volatility, but the total value earned showed no clear pattern.

But experience, to some extent, implies age. There are few inexperienced 30-year old players, and nobody under 25 with 8+ seasons under their belt. Breaking it down further, some minor patterns emerge.

The first thing that stands out is the 23-26 year-old cohort with 4-7 years’ experience. This is a small group whose value is strongly influenced by Clayton Kershaw (the #1 earner in this exercise), but even with Kershaw removed, the remaining value is still the highest on the board (because you’re curious: Elvis Andrus, Jay Bruce, and Justin Upton). And while the experts clearly targeted players aged 27-30 with significant experience, that group was below average.

One weak result here is the age pattern within each experience grouping. In the largest cohort, 4-7 years’ experience, there is a clear trend. In the others, the pattern is weaker, but we’re dealing with small samples.

Does past performance predict the future? In this case, it appears to, yes. The strongest predictor of 5-year performance was the player’s value in 2012, the year prior. While the overall correlation was only 0.39, indicating a moderate relationship, looking at the extremes showed a clear pattern.

Simply put, players who had already proven themselves prior to the exercise did better as a group.


It may also be helpful to look at players who went undrafted but finished in the top 72 during the exercise. The most common was unproven or pre-breakout starting pitchers (e.g., Max Scherzer, Cory Kluber, Zack Greinke, Jake Arrieta). They weren’t all young guys—Scherzer was 28 with 800+ IP under his belt—but none had really established themselves (Scherzer, for example, had a career 3.88 ERA—certainly not elite).

There were a few older players (David Ortiz, Nelson Cruz, Adrian Beltre) who aged better than we might have expected. And some young players who were either too far away (Nolan Arenado, Mookie Betts) or who became surprise stars (Charlie Blackmon, Jose Altuve, Daniel Murphy).  They look like misses in retrospect, but none would have been in anybody’s Top 100 in 2012.

What have we learned?

The samples here were small, but some trends emerged. First, there’s no magic formula. This was a group of experts, and half of our picks did not finish in the top 72. More than a third failed to return $10 a year.

There are no can’t-miss prospects. Are you dreaming of Vladimir Guerrero, Jr., or Fernando Tatis, Jr.? I give you Jason Heyward ($9 per season) and Jurickson Profar (-$5 a season); both were once #1-rated prospects. Add in Matt Moore (-$4 a season), who many FPAZ participants thought should have been a third-round pick in 2013.

Be aware of peak seasons and recency bias; sometimes a breakout is actually a peak. Regression can be a powerful force. Buster Posey was a $26 player in 2012; he averaged $15 over the next five years. Even Mike Trout, baseball’s best hitter from 2013-2017, never matched the $44 he put up in 2012.

Catchers are very hard to predict, and thus do not make ideal keeper/dynasty foundational players.

Age matters, but it’s only one factor. The same with experience. When combining the two, younger players with more experience is a desirable combination. However, talent, opportunity, and luck are confounding factors.

Performance matters! Players who have established themselves at a given level are more likely to perform in the future. Unproven players are more likely to wash out. It’s an oft-repeated mantra, but it’s important! Never pay for a level of performance that a player has not achieved.

Projecting a year ahead is difficult; projecting five years ahead, as you hope to do for a keeper or dynasty league, is harder by several degrees. When weighing the “now” versus the “later” in keeper and dynasty leagues, you should weigh the “now” very heavily. Next year should be an afterthought. Two years from now shouldn’t even be a consideration. Sure, you may miss out on a Mike Trout now and again. But you’ll make up for it by acquiring known talent and skipping the volatile, shiny new toys that come along.

Oh, and the winner? BaseballHQ’s own Andy Andres. And it wasn’t very close.

Click here to subscribe

  For more information about the terms used in this article, see our Glossary Primer.