2017 Baseball Forecaster: & Encyclopedia of Fanalytics

By Brent Hershey, Brandon Kruse, Ray Murphy and Ron Shandler

The industry's longest-running publication for baseball analysts and fantasy leaguers, the 2017 Baseball Forecaster, published annually since 1986, is the first book to approach prognostication by breaking performance down into its component parts. Rather than predicting batting average, for instance, this resource looks at the elements of skill that make up any given batter's ability to distinguish between balls and strikes, his propensity to make contact with the ball, and what happens when he makes contact—reverse engineering those skills back into batting average. The result is an unparalleled forecast of baseball abilities and trends for the upcoming season and beyond.

• Language: English
• Publisher: Triumph Books
• Release date: Dec 15, 2016
• ISBN: 9781633196957

Book preview

BASEBALL FORECASTER

AND ENCYCLOPEDIA OF FANALYTICS

Copyright © 2016, USA TODAY Sports Media Group LLC.

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form by any means, electronic, mechanical, photocopying, or otherwise, without the prior written permission of the publisher, Triumph Books LLC, 814 North Franklin Street, Chicago, Illinois 60610.

Triumph Books and colophon are registered trademarks of Random House, Inc.

This book is available in quantity at special discounts for your group or organization. For further information, contact:

Triumph Books LLC

814 North Franklin Street

Chicago, Illinois 60610

[phone]

www.triumphbooks.com

Printed in U.S.A.

ISBN: 978-1-62937-309-6

Rotisserie League Baseball is a registered trademark of the

Rotisserie League Baseball Association, Inc.

Statistics provided by Baseball Info Solutions

Cover design by Brent Hershey

Front cover photograph by John Hefti/USA TODAY Sports Images

Author photograph by Kevin Hurley

Ron Shandler’s

BASEBALL

FORECASTER

Editors

Ray Murphy

Brent Hershey

Associate Editor

Brandon Kruse

·  ·  ·  ·  ·  ·

Technical Wizard

Rob Rosenfeld

Design

Brent Hershey

Data and Charts

Matt Cederholm

Player Commentaries

Ryan Bloomfield

Rob Carroll

Matt Cederholm

Matt Dodge

Alec Dopp

Brent Hershey

Brandon Kruse

Ray Murphy

Stephen Nickrand

Kristopher Olson

Greg Pyron

Brian Rudd

Paul Sporer

Jock Thompson

Rod Truesdell

Research and Articles

Patrick Davitt

Ed DeCaria

Matt Dodge

Arik Florimonte

Brad Kullman

Brian Slack

Prospects

Rob Gordon

Jeremy Deloney

Tom Mulhall

Injury Chart

Rick Wilton

Acknowledgments

Producing the Baseball Forecaster has been a team effort for a number of years now; the list of credits to the left is where the heavy lifting gets done. On behalf of Ron, Brent, and Ray, our most sincere thanks to each of those key contributors.

We are just as grateful to the rest of the BaseballHQ.com staff, who do the yeoman’s work in populating the website with 12 months of incredible content: Dave Adler, Andy Andres, Matt Beagle, Dan Becker, Alex Beckey, Bob Berger, Chris Blessing, Derrick Boyd, Brian Brickley, Brant Chesser, Doug Dennis, Greg Fishwick, Neil FitzGerald, Brandon Gavett, Sam Grant, Rick Green, Phil Hertz, Joe Hoffer, Scott Holmes, Ed Hubbard, Tom Kephart, Chris Lee, Glenn Lowy, Troy Martell, David Martin, Harold Nichols, Frank Noto, Josh Paley, Joseph Pytleski, Nick Richards, Mike Shears, Peter Sheridan, Skip Snow, Matthew St-Germain, Jeffrey Tomich, Nick Trojanowski, Michael Weddell and Mike Werner.

Thank you to our behind-the-scenes troopers: our technical dynamic duo of Mike Krebs and Rob Rosenfeld.

Thank you to all our industry colleagues, a truly impressive group. They are competitors, but they are also colleagues working to grow this industry, which is never a more evident than at our annual First Pitch Arizona gathering each November.

Thank you to Dave Morgan, Chris Pirrone, Ryan Bonini, and the team at USA Today Sports Media Group.

Thank you for all the support from the folks at Triumph Books and Action Printing.

And of course, thank you, readers, for your interest in what we all have to say. Your kind words, support and (respectful) criticism move us forward on the fanalytic continuum more than you know. We are grateful for your readership.

From Ray Murphy The annual process of constructing this book should be a fire drill, but the great people involved transform it into a symphony. Rob Rosenfeld’s work on the construction of the player pages sets a tone, much like a double bass. The writers add so much pith in the player commentaries, with the varied yet complementary sounds of a woodwind section. The prospect coverage, the year-long work of our research team, the charts and data slices, all add the oomph of a brass section. Brent and I share the conductor’s role year-round, which is a testament to Brent’s boundless patience. Ron may not actually wield the day-to-day baton anymore, but he will always be our original virtuoso. In my other band, the home one, my wife Jennifer is the perfect blend of tag-team partner and co-conspirator, while Grace and Bridget keep bringing the fun.

From Brent Hershey This project always makes me appreciate of the rewards of hard work. To each contributor, thanks for your individual insights, digging and uncovering the raw materials. To Ray, thanks for your willing collaboration as we divide up the assembly process—fitting these pieces together into a cohesive unit. It’s always a balancing act, and each year we get closer to that mythical centerline. You are a joy to work with. Thanks to Ron, for entrusting us as foremen on this job, and providing additional finishing work when necessary. And to Lorie, Dillon, and Eden: thank you for the after hours delight you each supply, your graciousness, and your strength of support. Our family, too, is building something special.

From Ron Shandler I owe a huge debt of gratitude to Ray, Brent and the entire BaseballHQ family for their continued fine work on this annual project. While we have the process down to a near science—milestone-scheduled weekly from July through November—it is still a thrill to put all the pieces together and sort out those first projections. Mookie Betts, No. 1? Well, that’s exciting!

Thank you to 31 years’ worth of dedicated readers, who wait for those numbers as anxiously as we do. And thank you to my family who continues to provide unending support. Darielle, Justina, Michele and Sue are changing the world through theatre, music and empathy. They are all doing important work.

Despite some buzz to the contrary, I am not retired even though I’ve moved to Florida. I just get to share my office these days with a family of sandhill cranes. A writer has to write; that will never change.

TABLE OF CONTENTS

Devaluation

Encyclopedia of Fanalytics

Fundamentals

Batters

Pitchers

Prospects

Gaming

Statistical Research Abstracts

Gaming Research Abstracts

Major Leagues

Batters

Pitchers

Injuries

Prospects

Top Impact Prospects for 2017

Top International Prospects

Major League Equivalents

Leaderboards

Draft Guides

Blatant Advertisements for Other Products

Devaluation

by Ron Shandler

I once spent several years marveling at a certain player whose underlying power metrics screamed 40 home run potential. Forty is a big number, a feat, a milestone reserved for the very best, and this player seemed poised to take that leap.

Yet, for a variety of reasons, he kept falling short. Maybe it was his 60 percent contact rate and periodic sub-.200 batting average. Maybe it was his streakiness—.164 one month, .289 the next; .171 another month, .344 the next. Whatever the reason, the magical 40 remained elusive.

He came closest in 2014:

The 2015 Forecaster said, He could be a monster. UP: 45 HRs.

But, the forecasting gods were unkind and 2015 turned out to be a bust. Naturally, our recency bias depressed his 2016 expectations to near nothing. His average draft position ranking (ADP) coming into last season was No. 305, which in real terms meant, Undraftable even in a 13-team mixed league.

In the 13-team FSTA/SiriusXM experts league, he was selected in the fourth reserve round, or No. 339. That meant I did not deem him worthy enough to be among my first 26 picks even though I had seen 45-HR upside just one year earlier.

You can probably guess where this is going. The stubborn few who kept the faith last March—all six of you—were duly rewarded:

I suppose this is the long-awaited validation for my pre-2016 Chris Carter projections. I don’t know; he only managed the feat by going on an 11-HR tear from September 3 on. Regardless, we evaluate players on a full season’s performance; 40 HRs is 40 HRs, right?

Maybe not. There is one nagging little piece of data that takes something away from the accomplishment:

Despite hitting four more HRs, scoring 16 more runs and driving in six more runners—discounting the minor dips in steals and BA—Carter’s 15-team mixed league Rotisserie earnings were $5 less than in 2014 (and pretty crappy for a 40-HR hitter, all in all).

How could that be? The problem: everyone hit in 2016. When it came to home runs, everyone punched a power ticket:

• Every Upton.

• Every Seager.

• Every Davis.

• Two long-term singles hitters who had career power years at age 35. (1)

• Three declining sluggers with sudden power spikes at ages 37 and older. (2)

• Seven rookies who hit more HRs in their first partial MLB season than in any full minor league season. (3)

• My brother-in-law’s neighbor’s nephew. (4)

• Brad Miller.

Everyone. So in a year when balls were flying over the Green Monster, onto Waveland Avenue and into San Francisco Bay with wanton abandon, the value of any individual homer took a dive.

Rotisserie earnings are benched to the level of offense in any given season, thereby reflecting the context of the day. So, a bigger leaguewide offense will mean a lower value per homer … and every other offensive event. That’s how Carter earned $5 less in a season when he put up bigger numbers.

This phenomenon tends to skew our perspective. I suppose we can’t help but talk about a player’s individual power growth as if it occurred in a vacuum. But everything has to be evaluated within its own context.

So no, this was not a career power year for Evan Longoria (he showed better skill in every season prior to 2014). Brad Miller’s huge breakout was not as big as we think (his expected power index of 116 was just a tick above 2014’s 105). And get this: Todd Frazier’s career power year came along with the lowest xPX of his career. By a lot.

So we can debate whether Chris Carter’s first 40-HR season is an accurate reflection of the skills befitting of that milestone. His xPX has been essentially flat for four straight years. This has been the same player since 2013.

It’s possible that hitting 40 home runs is not as elite of a feat as it used to be.

Power and the balls

Take a look at how the power environment has evolved since around the turn of the century:

The last three columns represent the number of players who hit at least 20 HR each year, the average number of HRs those players produced, and the percentage that their homers represented of baseball’s entire HR output.

In 2014, only 57 players hit 20 or more HRs; their output represented 35% of all the homers hit that year. Just two years later, there were nearly twice as many 20-HR hitters making up more than half of all the HRs hit.

That meant, every owner in a 15-team mixed league could have rostered seven 20-HR hitters (and six teams could have owned eight). Essentially, you would have been hard-pressed not to trip over a 20-HR hitter in your drafts last March.

I actually projected this, right here, one year ago:

… barring a revelation that some external variable changed in 2015, one would expect power to regress off of this year’s spike. But wait … this 2015 correction was the largest single season spike since 1993. Back then it set off a whole new era in power performance. Could we be entering a new cycle?

It’s possible. As you scan all the player boxes in this book, you’ll see many new players being projected for 20 HRs or more, driven by nothing more than normal trends … In all, I count 78 players projected for 20 or more HRs. Last year, only 64 players hit at least 20 HRs. This correction may have legs.

Right idea, wrong magnitude.

There has been a lot written to analyze this sudden phenomenon. Here is an excerpt of what I wrote back on May 19:

Everyone has their opinions about why baseballs are flying out of ballparks at the most frequent rate in the history of the sport. (Most explanations) can be discounted. There’s only one other popularly-held explanation left. But MLB’s Powers That Be will never reveal the truth about the baseballs that are being produced.

Yes, I wrote this on May 19.

May 19, 2000. Seventeen years ago. The more things change, the more they stay the same.

But it’s interesting to compare the phenomenon from the Steroids Era to the current Post-PED-Power-and-Punchout Era. Something extraordinary happened in 1999 and carried over into the following season on an even bigger scale.

From 1998 to 1999, the major league home run rate jumped from 2.08 bombs per game, to 2.28. By the end of May 2000, hitters were slamming an amazing 2.62 home runs per game. At the time, the talk was about how someone had flipped a switch. This discounted any explanation that would lend itself to a more gradual change—warmer weather, smaller ballparks, the impact of expansion or suddenly stronger hitters (steroids had not yet become a part of the popular vernacular). Thus, the conclusion was that the baseballs had to be juiced.

Seventeen years later, we’re once again using a switch-flipping metaphor for the recent power spike. Prior to the 2015 All Star Break, the home run rate was 1.76 per game; after the Break, 2.20. Our surge continued into 2016, the home run rate climbed to 2.32.

The consensus opinion? It’s the balls. (Some attribute the record-breaking summer heat, but it’s tough to switch climate on and off at will.)

But that’s not the end of the story …

The sudden power surge in early 2000 forced us to start reassessing our benchmarks. The juiced ball theory took on greater life after Rawlings refused to allow photographs of the machines that wound the yarn around balls.

But then the surge abated as suddenly as it had begun. By September, the homer rate had faded to 2.06 and finished the year at 2.34. It was as if someone had flipped the switch OFF. Any talk of juiced balls vanished as steroids became a more attractive headline magnet. We never found out the truth about the balls.

The point is, these tectonic shifts can occur at any time, in any direction. They are not necessarily gradual and they don’t have to follow any expected trend. Look at the chart; we can no sooner explain the 2015-2016 spike than we can the 2000-2002 drop during the height of the Steroids Era. And so, neither can we draw any conclusions about the direction this trend will take in 2017.

Balls.

There were 111 players who hit at least 20 HRs last year. A quick scan of the projections here—admittedly, a rough number at this time of year—and I count 101 potential 20-HR sluggers. That level remains pretty significant though it does reflect some minimal regression. Still, I’m not betting the rent that this home run barrage will continue into 2017.

Deceleration

While home runs were a dime a dozen, stolen bases have become scarcer commodities. The number of elite speedsters dropped to its lowest level in over a decade.

These days, in a 15-team league, teams will average about two players with 20-SB potential. Compare that to five years ago when every team could have rostered three players with 20+ stolen bases.

The 2016 Baseball Forecaster spoke to this phenomenon:

Given that steals have always been centralized in a smaller group of elite players, it has made sense to assign a somewhat inflated value to their contribution. But these days it’s even more prevalent. It makes a strong argument for drafting a Jose Altuve or Dee Gordon in the first round, or for $30-plus. I think that approach has become more reasonable to consider these days; just plan the rest of the roster around it.

Will this phenomenon continue?

The first column is the percentage of all plate appearances that have been singles and walks. While this is not a perfect measure of events that create stolen base opportunities, it does provide a rough gauge to track the trend. You can see that this has been declining over time. The 2% decline since 1999 seems small, but that equates to around 4,000 fewer baserunners per year who could potentially put up SBs.

The second column is a bit tangential but interesting nonetheless. It shows the larger Power-and-Punchouts environment in which those potential stolen base opportunities exist. This is the percentage of plate appearances that result in a ball in play—essentially reflecting the global rise in baseball’s three true outcomes (homers, walks and strikeouts). The decline here is even steeper, 5% over the past decade alone.

The third column is Stolen Base Opportunity Pct., which measures how often a stolen base is attempted when a batter reaches first by either a single or walk. This trend has fluctuated within a 2% range over time, but as been in decline from its recent 2011 peak. The last two years’ decline compares with the corresponding spike in power. No point stealing second when odds are a homer is right around the corner, right?

The final column is the Stolen Base Success Rate, or how often a base-stealer is successful in his attempts. While the last two years have seen a bit of decline, steals success is still better than it was 15 years ago. It would be tough to draw a conclusion that the decline in steals has any connection to runner skill.

So, the decline in bags is a decline in the events that potentially create stolen base opportunities. Writer Joe Sheehan also attributes the phenomenon to the decline in singles, not only to create baserunners but also for driving in those runners who’ve stolen second base. The risk of a failed steal outweighs the benefit of putting a runner into scoring position if there are fewer run-scoring singles. Again, odds are a homer is right around the corner.

This environment seems like more of a core trend. It might continue. It might stabilize. Or not. If you look at 2002-2007, you might conclude that we’re just going through a similar trough now.

So I’m not taking bets for 2017. It’s an easy speculation to see how things could turn around. Dee Gordon doesn’t get suspended. Roman Quinn potentially sits at the top of the Phillies batting order. Full seasons for Mallex Smith, Keon Broxton, Jose Peraza, and on and on. There were 28 players who stole at least 20 bases last year. A quick scan of the projections here—admittedly, a rough number at this time of year—and I count 40 potential 20-SB speedsters. And we’re back out of the trough.

Trickle-down trends

The more that we can’t project the environment in which our stats live, the tougher it is to get a read on individual player projections. They are moving targets as well. This book provides a better understanding of skill so you can take aim at those targets with as much precision as is possible. But as Fanalytic Fundamental No. 1 notes: This is not a game of accuracy or precision; it is a game of human beings and tendencies.

Still, no matter how much we might intuitively know this is true, we are going to obsess over who to rank where and how much to pay for Player A, Player B and Player Z. We can’t help it.

QUESTION: If you have the No. 1 seed in your 2017 draft, will you draft Mike Trout, Mookie Betts or Jose Altuve?

ANSWER: It almost doesn’t matter.

For starters, owning the No. 1 seed hasn’t exactly been a stone cold lock anyway.

For 2017, Trout seems like the safest pick, given what appears to be a high floor. But we could have said the same thing about Albert Pujols not too long ago. One day, the Trout Era will end too.

As far as the best player, well, he can come from anywhere:

In fact, our ability to nail anyone in the first round of a 15-team draft is just as imprecise. For years I’ve been parading around the research stating that our success rate in identifying each season’s top 15 players is—using the scientific term—awful.

As in 35.5% awful since 2004.

Several of my industry colleagues have dismissed this research. Their contention is that a player drafted in the first round doesn’t need to return first round value, only some reasonably close level. For instance, in 2016, Miguel Cabrera was drafted No. 13 as a $30 player. He finished ranked No. 19, earning $27. That’s clearly an acceptable result, even though he didn’t finish in the top 15.

However, some might also make a case that the owners of Josh Donaldson (No. 5, $38) should not have been disappointed with his finish at No. 23, earning $26. But in real terms, any player whose end-of-season value is lower than projection means that you are taking a loss on your investment. As much as Donaldson owners may have been satisfied, they still took a $12 loss. I suppose it comes down to your own tolerance level, but I wouldn’t be happy with a $12 loss.

Part of the problem is that the slope of player value at the high end of our rankings is incredibly steep. If the 90th ranked player finishes No. 125, your real loss is only about $4. If your No. 1 pick finishes 36th—the same span of picks—your loss is more than $20. The early round misses have far more impact.

But the bottom line is, those failed first-round picks are not finishing just outside the top 15 anyway. They are barely in the top 30. Look:

In fact, most of the players in the ADP top 15 finish nowhere near their draft spot. Look at the past three seasons:

Admittedly, 2016 was one of our better years. But, for all the pre-season obsessing we do over getting the best seed and trying to decide who to draft where, these lists look like a whole bunch of blind dart-throws.

The point of all this

When you combine …

• the volatility of the statistical environment

• the imprecision of player projections, and

• the uncertainty of the drafting process … you get one hot mess.

Each year, we respond to these variables in pretty much the same way, looking for a silver bullet that doesn’t exist. But we don’t need to resign ourselves to another year of wanton randomness either.

When it comes to the environment volatility, analysts will do their best job at forecasting, but will ultimately fall back on Merkin’s Maxim: When in doubt, predict that the current trend will continue. There are worse approaches, but frankly, there is not much more we can do. Observe, and react.

When it comes to projective imprecision, forecasters will build ever more elaborate models, scratching and clawing for each thousandth of a decimal point in mean squared error. That’s a lot of unnecessary effort.

Your energy will be better spent learning the tools in this book. They’ll help you get a better handle on each player’s potential. Focus on the peripheral metrics and commentaries, not the projections; that’s where the treasures are hidden.

When it comes to drafting uncertainty, tacticians will devise increasingly brilliant strategies to game our opponents, always hoping they respond as lab rats might.

But the year that you try to deplete pitching resources by over-drafting your staff is the year that Chris Archer, Matt Harvey, Dallas Keuchel and Zack Greinke go belly up. The year that you try to corner the market on speed is the year that Dee Gordon gets suspended and Manny Machado stops running. Besides, your league’s winner is still going to be the guy who drafts Jonathan Villar and Rick Porcello anyway.

The best we can do here is look for small tactical advantages. It’s the little stuff that can give us an unexpected edge. I’ve found two such tactics that have helped my process over the past few years.

Categorizing foundation players

History has shown that four types of players generally populate the early rounds in most drafts, or get auctioned for $25 or more:

1. Veteran early round earners are players who’ve been consistently atop the end-of-season leaderboards. These are players like Trout, Altuve, Goldschmidt and Kershaw. These are the lowest risk players to own.

2. New early-round earners are players who earned elite value for the first time last season. These are players like Bryant, Marte, Villar and Blackmon. Recency bias will push many of them in to the early ADPs and $30 bids in 2017, but not all deserve to be there.

3. Possible rebounds are players who were previous top talents and still have the skills to potentially return to the top. In 2017, these will be players like Harper, McCutchen and Stanton. Here again, there is greater risk for a rebound, but some of these types could well bounce back.

4. New season risers are players who have the potential to develop into elite level earners but have never previously achieved that level over a full season. These Draft Day speculations included players like Carlos Correa and Kyle Schwarber in 2016. Expect Trea Turner and Gary Sanchez to be part of this group in 2017.

Consider these four groups as a high-level hierarchy of the types of players to target, the vets at the top and risers at the bottom. Using this logic, I might be more inclined to look at a Kris Bryant or Charlie Blackmon before a Bryce Harper. Heresy? Perhaps. But preconceptions are what often get us into trouble.

Let’s take a look at last year’s top 30 players to see what group each fell into coming into 2016, and how they each fared.

In the chart below, Grp is the Group the player would have been in on Draft Day 2016. ADP$ are the equivalent auction dollars converted from each average draft position. Earnings are the actual dollar earnings at the end of the season. Net is each pick’s net profit or loss.

Only eight of the 30 players turned a profit, but that’s okay. We are not looking for profit in this part of the player pool. Here, the goal is par, and to minimize the risk of big losses. In addition to the eight, another six came within $5 of their acquisition cost. That makes 14 of the 30 (47%) accomplishing about what we needed them to do. Still, a hit rate of less than 50% is nothing to write home about.

But we can do better. Let’s summarize the above picks based on where they fell in our four groups.

Yes, it’s only one season and the sample sizes are small, but the risk of loss did increase the further down the hierarchy you drafted. These results reinforce the need to play it conservative in these early rounds.

Matt Harvey’s ill-fated season drags down Group 4 (without him, the average net loss would have been $9), but I still stand by the oft-advised adage: "Never pay for a