Going Deep: Learning From Last Year’s Biggest Busts

If you’re anything like me, when October comes to a close, you spend your time obsessing over baseball content and praying the calendar flips to the following year. You complain to your friends about your fantasy baseball teams and the randomness and unfairness associated with fantasy football scoring and injuries. You participate in fantasy baseball mock drafts and build rankings. And each hot stove move reinvigorates you and makes you finally feel something.

Perhaps that’s overdramatic, but I have some good news for you: With only a few weeks until pitchers and catchers report to spring training, it’s time for some average draft position analysis. For this, I compared March 2018 National Fantasy Baseball Championship ADPs to this year’s NFBC ADPs going back to Jan. 1 (23 drafts). I chose NFBC ADP data because the entry fees to NFBC leagues are substantial, such that the data should be a decent proxy for market value among serious fantasy baseball players.

For context, NFBC ADP might be skewed relative to what you’re seeing in your own traditional leagues because NFBC leagues are often different sizes and have different rules affecting ADP. For example, Draft Champions leagues prohibit all transactions, which could give closers a bump as participants can’t chase closers on the waiver wire. Likewise, many NFBC leagues are two-catcher leagues, which give catchers a boost in the rankings. That said, the comparisons I make between 2018 and 2019 NFBC data will be unaffected as both years are subject to these same rule and roster idiosyncrasies. The above observations only matter for comparing overall NFBC ADP with your own rankings.

First, I examined the 20 players that fell the hardest out of the 2018 top 100. I limited results to players that fell out of the top 100 because broadening the scope to all players that fell down NFBC draft boards returns names that are completely irrelevant for most 2019 drafts. For instance, some of the top fallers between 2018 and 2019 ADP among all players were Hector Neris (-403.34), Greg Holland (-402.63), Adam Duvall (-365.69), Bradley Zimmer (-360.86), and Mark Melancon (-353.25) — not exactly the most fantasy-relevant bunch these days.

So who are the unlucky fallers we care most about?

Name	2018 ADP	2019 ADP	2018-2019
Yoenis Cespedes	93	487.43	-394.43
Cody Allen	89	270.78	-181.78
Byron Buxton	48.8	193.45	-144.65
Jose Quintana	69.4	185.43	-116.03
Elvis Andrus	60.1	174.57	-114.47
Billy Hamilton	68.1	174.26	-106.16
Yu Darvish	49.9	155.52	-105.62
Chris Taylor	95.3	200.78	-105.48
Jonathan Schoop	65.7	170.7	-105
Brian Dozier	36.5	140.61	-104.11
Dallas Keuchel	77.6	178.13	-100.53
Corey Knebel	66.1	160.91	-94.81
Dee Gordon	28.8	121.22	-92.42
Willson Contreras	50.1	136.26	-86.16
Eric Hosmer	82.6	168.70	-86.1
Buster Posey	58.3	138.13	-79.83
Chris Archer	55.9	135.48	-79.58
Robbie Ray	47.4	118.30	-70.9
Edwin Encarnacion	57.2	127.65	-70.45
Carlos Martinez	54.6	123.87	-69.27

 

Injuries and Closers and Burners, Oh My!

 

We can instantly remove Yoenis Cespedes from any discussion because he missed most of 2018 and may miss all of 2019 as well. Others can be explained, in part, by injury-riddled 2018s that burned fantasy owners. Consider Elvis Andrus, Yu Darvish, Josh Donaldson, Robbie Ray, Byron Buxton, Wil Myers, Carlos Martinez, and A.J. Pollock. More on some of them in a bit.

Cody Allen and Corey Knebel bespeak the volatility at closer and may caution against reaching for the position in 2019. Honorable mentions go out to Neris, Holland, Melancon, Blake Parker, Kelvin Herrera, Jeurys Familia, Brandon Morrow, Alex Colome, and Arodys Vizcaino, all of whom also fell precipitously from last year’s ADP but were not in the top 100 and thus did not make this list. Once these guys lose their roles, they have little value. How comfortable do you feel drafting closers this year?

A similar niche category of players is the burner, e.g., Billy Hamilton and Dee Gordon. When they stop running or lose opportunities to run, they burn you three times (pun intended). First, you had to invest a high pick on either of these guys. Second, they hurt you in the other hitting categories. Here are their 2018 lines in traditional 5×5 leagues:

Name	R	HR	RBI	SB	Avg.
Dee Gordon	62	4	36	30	.268
Billy Hamilton	74	4	29	34	.236

So if they aren’t running as much as expected, they’re doing nothing to help you. Third, you drafted them expecting a certain number of steals, and consequently, you probably felt comfortable forgoing other sources of steals. So when these guys don’t pan out, you’re in desperate need of steals. For these reasons, I tend to avoid drafting stolen base specialists when possible.

Next, we have players who experienced skills regressions that can be split into three categories:

1. Low ceiling, boring players: Jose Quintana, Dallas Keuchel, Eric Hosmer, and Buster Posey.
2. Veterans who didn’t live up to expectations but may still have something left in the tank: Brian Dozier, Edwin Encarnacion, and Chris Archer.
3. 2017 breakouts who fell back down to Earth: Byron Buxton, Jonathan Schoop, Chris Taylor, Willson Contreras, Andrus (strike two), and Ray (strike two).

 

The Snoozers

 

What can we learn from the boring players? None of these guys should have been in the top 100 in the first instance, and the warning signs were there. When I say boring, I mean they have limited ceilings such that, even if they produced, they’d barely return their cost.

Hosmer had clearly limited upside going into 2018 because he lacked both power and speed. Hosmer barely runs, has never hit more than 25 homers, and maintained an extremely subpar batted ball profile in 2017: 55.6 GB%, 22.2 FB%, 31.3 Pull%. He also made 21.8% soft contact and 29.5% hard contact. Thus, his 25 home runs in 2017 were likely unsustainable, and when his BABIP luck ran out, there was nothing productive left. I get that he hit 25 homers in 2016 as well, but is Hosmer really a Giancarlo Stanton-esque slugger we can comfortably say will outperform his batted ball profile and post a 20-plus HR/FB% for a third straight season? Nope! Unlike Stanton, Hosmer had a mediocre 5.2 Brls/PA% and good but not great 94.1 mph average exit velocity on FB/LD. I use these two inputs to determine the sustainability of a player’s HR/FB% because they are both highly correlated with HR/FB%.

Posey never ran, and his power upside was also obviously limited given his home ballpark and poor 2017 marks in hard contact rate (33%), FB% (33%), exit velocity on FB/LD (92.5 mph), and Brls/PA% (2.6%). His ability to muscle the ball out of the park was clearly declining based on his HRs and HR/FB% over the past four seasons.

Season	HRs	HR/FB%
2014	22	13.4
2015	19	11
2016	14	9.8
2017	12	8.4

It was no surprise to me when he became an empty batting average in 2018.

Quintana and Keuchel are ratios pitchers, neither of whom ever maintained a K/9 higher than 8.40 (Quintana’s 2017 was the lone exception but still a lucky outlier because of an 8.4 SwStr% that was right in line with his career rate). Even at their best, these guys posted league-average swinging strike rates. Keuchel outperformed his 2017 FIP and SIERA by nearly a full run, and even though Quintana actually underperformed his 2017 FIP and SIERA, they pegged him at 3.68 and 3.80, respectively. Why spend a high draft pick on either of these guys who reasonably could give you a 3.70 ERA and 20% K rate? When Quintana and Keuchel weren’t able to contain the damage just by pitching to contact in 2018, the bottom fell out because they didn’t have strikeouts to prop up their bottom lines.

The takeaway from these four players is if someone doesn’t inspire you, don’t draft him.

One name in this year’s top 100 that doesn’t inspire me is Jose Peraza. Here’s a player who had 683 plate appearances in 2018 but only mustered the following: 85 R/14 HR/58 RBI/23 SB/.288 AVG. He pushes the needle in no single category and probably hurts you in two. His 29.5% hard contact rate was 15th-worst among qualified hitters, his 2.0 Brls/PA% was extremely poor, and his average exit velocity on FB/LD was 10th-worst among hitters with at least 150 batted ball events. Yet Peraza’s still trying to partake in the fly ball revolution by hitting 38% fly balls. You can bet he won’t produce many RBI or more than 15 home runs with those poor underlying power indicators.

Worse yet, his desire to hit home runs by lifting so many fly balls will depress his BABIP and, in turn, limit his batting average. Moreover, his elevated 35.2 O-Swing% holds back his 4.2 BB% (O-Swing% and BB% are highly correlated), which explains why his on-base percentage (.326) is so low. And even though he swiped 64 bags in one season of A ball, Peraza’s low OBP and surprising sprint speed (28.8 ft/sec, good for only 70th overall) explain why he stole just 23 bases in 2018 and ensures he won’t have too many SB opportunities in the future.

I say all of this not because I dislike Peraza inherently but because I dislike his price. As with Hosmer, Posey, Keuchel, and Quintana, the warning signs are usually there for players with low ceilings. The best you’re going to get with him is what he did last season, which is … fine. But as with the other four, he could really flop next season.

 

The Veterans

 

Here’s where your ears should perk up. Fantasy players often expect veterans to have peaked and to be ready to fall off a cliff. In some cases, then, there may be value left on the table. These are not guys where it was necessarily apparent they would regress going into 2018; rather, they’re guys who fantasy players perceive to have declined in 2018 based on their age, results, or some combination of the two.

Let’s consider Dozier first. He’s played at least 147 games in each of the past six seasons and is surprisingly only going to be 32 years old. His batting average, however, fell to .215 after hovering around .270 for two seasons. The reason? A .240 BABIP — 60 points lower than 2017 and 40 points lower than 2017. His entire batted ball mix, including his Pull%, was nearly identical in all three seasons. The only difference? A slightly elevated IFFB% at 17.1% in 2018. He even maintained a .260 xBABIP in 2018. He’s never been a high BABIP hitter given his heavy pull and fly ball tendencies, but I think when his infield fly balls regress to the mean, so will his BABIP.

That said, some of the other peripherals aren’t promising. While his 37.3 hard contact rate seems like an improvement over 34.1% in 2017, hard contact was actually up about 3% across the league, nullifying any apparent gains Dozier made. What’s more, his 4.3 Brls/PA% was only good (or more aptly put, bad) for 173rd overall, down from 5.2% last year, and his 91.6 mph exit velocity on FB/LD was even worse (211th overall), down significantly from 2017 (94.4 mph). That’s why his HR/FB% free fell from 16.8% to 11.2% in 2018. With xStats giving Dozier exactly 20.2 HRs in 2018, pencil him in for 20 homers again and don’t expect another 42 home run season. Separately, as an aging player whose sprint speed fell to 330th overall in 2018 (four spots behind Miguel Sano), Dozier probably will continue to run less in 2019. I expect he bats .255 with 20 homers and 10 steals, meaning that the perception of his age-related decline is, at least in my opinion, accurate.

Still, I’m more bullish on Archer. After a third straight year in which he underperformed his FIP and SIERA, many are finally avoiding the pitfall of paying top dollar for him. But that doesn’t mean the ERA indicators are without value. They’re based in large part on his impressive annual strikeout rate, which fell last year to 25.4% but probably shouldn’t have fallen so far given that his SwStr% was right in line with prior years. And providing some hope, his strikeout rate jumped in the second half to 27.1% as he started trading his perennially terrible four-seam fastball for a sinker and changeup. Accordingly, Alex Chamberlain wrote about how reincorporating a power sinker could be the key to correcting Archer’s problems. While I’m not sold, there’s definitely hope for the 30-year-old righty, and I love that you can get him at 135th overall, where the risk is baked into his price.

Finally there’s Encarnacion. I cannot understand why he’s falling so far in drafts, other than fantasy players are assuming he’s going to fall off a cliff because he’s 36 and calling Safeco Field home. Sure, the SwStr% and O-Swing% each jumped a little, resulting in declining BB% and K%. But his Z-Contact% actually improved, which leads me to believe there isn’t any age-related decline as he was still able to make contact in the zone instead of being overwhelmed by MLB pitching. Likewise, Encarnacion hit more than 30 homers for the seventh straight season and at the same pace as years past, which was also supported by xStats (30.2 xHRs). In fact, his VH% and PH% both improved in 2018, he hit 43.7% fly balls, pulled the ball 45.6% of the time, and made 42.4% hard contact (a career high). With a HR/FB% right in line with his career numbers and solid marks in Brls/PA% (7.6 – 44th overall) and exit velocity on FB/LD (94.3 mph – 83rd overall), I see no reason to doubt Encarnacion’s continued ability to smack some bombs in 2019. Let misconceptions of age-related decline work to your advantage and take 30 guaranteed homers at 127th overall.

 

The Breakouts

 

Lastly, there are the 2017 breakouts who failed to repeat in 2018. I’m always skeptical of breakouts, and for each one, there were warning signs in 2017, just like with the low-ceiling guys.

Buxton ran a .367 BABIP, 28.4 K%, and 13.6 SwStr%. The .253 batting average was going to drop because of his K% and unsustainable BABIP when viewed in the context of his 38 FB% and 47.3 Pull%. Thus, he was trying to sell out for power by pulling the ball a ton and hitting a lot of fly balls, which is terrible for his BABIP. It worked for his power in 2017, as he luckily maintained a 14.2 HR/FB%, but it wasn’t going to last given his poor exit velocity on FB/LD (92.8 mph), 27.6 hard contact rate, and 3.5 Brls/PA%. Buxton ultimately was a nightmare in 2018, getting injured, shuttling back and forth between the Twins and minors, hitting no home runs, and batting .156.

Schoop similarly maintained an elevated .330 BABIP in 2017 without the batted ball profile to back it up (lots of pulled fly balls and IFFB but average grounders and line drives). He only had a 21 K%, but it wasn’t unsupported by his elevated 13.8 SwStr%. In addition, his decent but not great 5.0 Brls/PA% and 92.7 mph exit velocity on LD/FB did not support his 17.7 HR/FB%. For these reasons, his batting average fell precipitously in 2018, and the power dropped too. I’m more bullish on Schoop than Buxton given the fact that he profiles as more of a power hitter, which makes pulling fly balls more sensible. I also love his 181.5 ADP, but the fact remains he was overvalued going into last season, and his batting average was likely a mirage.

Contreras’ profile is similar to Hosmer’s. Where a player doesn’t steal bases, hitting 52% ground balls and only 30.7% fly balls portends bad things, such as hitting 10 home runs in 138 games. In particular, it was clear he deserved his 9.3 HR/FB% given mediocre marks in Brls/PA% (4.8%), exit velocity on FB/LD (92.9 mph), and hard contact rate (28.9%), all worse than in the two years prior. He just hit the ball a lot harder in 2016 and 2017, which enabled his 23.5% and 25.9% HR/FB rates in those seasons, respectively. Now that he’s hitting the ball on the ground and making weaker contact, I’m completely out on Contreras even at this price.

Andrus only played 97 games, but his BABIP fell 33 points, resulting in a similar decline in batting average. This, however, was likely a consequence of bad luck in 2018. In both 2017 and 2018, he maintained an xBABIP of .308. But in 2017, he hit 17.7% low drives, right around league average, but he hit 13.3% low drives in 2018. Given the extremely high batting average on low drives and that they’re generally luck-based, they were driving his BABIP decline in 2018. When his low drive luck rebounds, I expect his average to as well in 2019. That said, Andrus’s 2017 power was likely a mirage. His xSLG was 38 points below his SLG. He never hit enough fly balls (31.5%) or hit the ball hard enough (30.5 Hard Contact%, 3.3 Brls/PA%, 91.8 exit velocity on FB/LD) for his 20 homers to be repeatable. Separately, his sprint speed declined 0.2 ft/sec, not enough to make me doubt his ability to continue running, though not quick in the first instance (269th overall). Maybe Andrus can hit 10 homers with 15 steals and a .280 average, which is fine for 174th overall but not good enough for 60th.

Taylor is a little more complicated. He had an elevated 25% strikeout rate in 2017, but it probably seemed too high relative to his average 10.8 SwStr%. Nevertheless, his K% actually got worse in 2018. Similarly, he had an inflated .361 BABIP, but that didn’t regress much either. The only thing that actually declined as expected was home runs, which went from 21 to 17 despite playing more games in 2018. I say “expected” because Taylor’s 2017 15.8 HR/FB% was not supported by his 32.4 Hard Contact%, 5.1 Brls/PA%, and ugly 91.2 mph exit velocity on FB/LD. The SB decline was also unexpected given his high mark in sprint speed, but 20/20 in 2018 was unlikely anyway as a result of his power regression.

Last but not least is Ray, who outperformed his FIP and SIERA by nearly a full run in 2017. He gave up 40.4% hard contact rate and had a .267 BABIP and 84.5 LOB%. The warning signs were there in 2017, but I think his true talent probably lies somewhere between his injury-plagued 2018 and partly luck-fueled 2017. Ray’s strikeout ability is nearly unparalleled in the MLB, so I don’t think he’ll repeat his mediocre 2018 ERA of 3.93, and there might even be some value at 118th overall.

This is not all to say that you should never pay for breakouts, otherwise you’d miss out on repeats such as Whit Merrifield, Tommy Pham, Alex Bregman, etc. Just be skeptical and objective. For example, while I wrote about Adalberto Mondesi’s breakout midseason, I’ll consider him with a skeptical eye when I build my rankings and look at his peripherals rather than just getting excited at what he could do given 600 PAs.

 

Conclusion

 

What are some conclusions we can draw from this?

1. Niche, single-category players often represent poor investments because when they fail to contribute in that category, you’re left in a deep hole.
2. Players with low ceilings don’t inspire confidence and should not be drafted in the top 100.
3. In select cases, veterans represent excellent buying opportunities.
4. Single-season breakouts tend to regress to the mean. Use analytical tools to discern the skilled from the lucky.

 

Photo by Gavin Baker/Icon Sportswire