Attempting to Forecast Yoan Moncada Statistically

by
March 2, 2015

The general consensus on Yoan Moncada is that he’s among the best prospects out there. The newest Red Sox prospect’s bat speed and power both grade out as plus, and scouts believe he has the physical tools to be an asset on defense as well. Our very own Kiley McDaniel weighed in on Moncada several times this past winter, and said he’d rank him in the 5 – 12 range on his top 200 list.

My wheelhouse is forecasting prospects’ future production using minor league stats. Admittedly, this might not be of much use for a player like Moncada, who has nary a minor league plate appearance to his name. But rather than throwing my hands in the air and deferring entirely to the scouting reports, I decided to use the scarce data available to glean at least some insight into how good Moncada might be as a big leaguer.

With the exception of a few established veterans like Jose Abreu and Yoenis Cespedes, most recent Cuban imports didn’t jump right to the majors. The majority spent time in the minors first — just as Moncada’s expected to do in 2015, and probably 2016. In the last decade, I found 19 hitters from Cuba who logged at least 100 plate appearances in Double-A in their first year stateside. Using this admittedly small sample of players, I looked to see how the inputs to KATOH — BB%, K%, ISO, BABIP, and SBA% — translated from the CNS to Double-A. Unsurprisingly, I found positive correlations across the board. Applying these translations to Moncada’s Cuban stats, we would expect the following performance from Moncada in Double-A next year:

BB% K% ISO BABIP SBA%
Moncada’s Double-A Projection 8% 19% .133 .322 13%
Relative to League Average -1% +0% +.006 +.017 +5%

It’s a little hard to picture what this player looks like based purely on what’s listed above. To help paint a more vivid picture, here are some players (of all ages) who had similar lines in Double-A last season:

Name AVG OBP SLG HR/ 500 PA wRC+
Alen Hanson .280 .326 .442 10 111
Yorman Rodriguez .262 .331 .389 9 103
Rangel Ravelo .309 .386 .473 10 142
Taylor Featherston .260 .322 .439 15 116
Jason Martinson .236 .305 .358 10 85
Taylor Motter .274 .326 .436 16 111
Mycal Jones .255 .337 .347 3 96
Cody Stanley .283 .340 .429 14 119
Stephen Bruno .276 .346 .393 3 110
Darrell Ceciliani .289 .331 .405 8 103
Shannon Wilkerson .266 .317 .357 2 88
J.T. Realmuto .299 .369 .461 9 132
Rymer Liriano .264 .335 .442 17 122
Kyle Johnson .259 .344 .384 5 104
Cutter Dykstra .274 .349 .391 7 107
Josh Richmond .247 .325 .399 11 104
Average .271 .337 .409 9 110

Basically, we’re looking at a player whose bat is maybe a tick or two better than the average Double-A hitter. So if we anticipate production from Moncada in 2015, what might we expect from him beyond next season? Running his projected stat line through the KATOH machine, I came away with the following forecast through age 28.

MLB >4 WAR >6 WAR >8 WAR >10 WAR >12 WAR >16 WAR WAR thru 28
89% 41% 29% 24% 19% 19% 9% 6.1

If Moncada’s transitions transition from Cuba looks anything like what we’ve seen from past players, KATOH would be a fan, but perhaps a little lower on him than the scouting reports suggest. His projected 6.1 WAR through age 28 would place him 44th on KATOH’s top 200 list, right between Henry Owens and Kyle Schwarber; a real prospect, but not quite up there with some of the other names mentioned.

For kicks, let’s throw around some names of players whose age-20 season were similar to Moncada’s anticipated season. To qualify for this group, a player needed to be a 20 year old in Double-A, and fall within one standard deviation of Moncada’s projected performance for each metric. The two players marked with an asterisk (*) are still younger than 28, and will likely add a few more WAR to their totals.

Year Name BB% K% ISO BABIP SB% WAR thru 28 MLB Games WAR/150 G
2007 Andrew McCutchen* 8% 18% .117 .294 14% 33.9 880 5.8
2003 Grady Sizemore 8% 16% .176 .329 7% 29.3 892 4.9
2010 Brett Lawrie* 7% 20% .167 .340 20% 7.9 345 3.4
2005 Melky Cabrera 6% 15% .124 .307 10% 9.3 1,066 1.3
2000 Felipe Lopez 5% 22% .119 .327 9% 7.0 873 1.2
1990 Mark Lewis 5% 15% .174 .301 8% 0.0 726 0.0
1996 Edgard Clemente 9% 23% .182 .351 5% 0.0 114 0.0
1999 Luke Allen 8% 21% .183 .346 10% 0.0 8 0.0
1995 Willie Romero 7% 15% .132 .318 18% 0.0  0 0.0
2015 Yoan Moncada Projection 8% 19% .133 .322 13% ? ? ?

Overall, this is a pretty exciting collection of players. McCutchen and Sizemore were both elite hitters, and each of Lawrie, Cabrera and Lopez became perfectly productive big leaguers. Sure, there are some duds mixed in, but flaming out completely is risk that exists for virtually every prospect.

Keep in mind that Moncada’s KATOH forecast isn’t just a projection, but a projection of a projection, which adds another layer of uncertainty. To make matters worse, this analysis is based on a sample of only a few players, who may or may not accurately represent the transition from Cuba to the US. Still, I think there are a couple of interesting takeaways here.

For one, Moncada’s Cuban stats suggest he has decent pop for a teenager. Despite being just 18-years old, he hit for a respectable .133 ISO in his final year in the CNS. Based on how past players have transitioned, we might expect his performance to translate to roughly league-average power in Double-A. Coming from a 20-year-old, this would bode very well for his future offensive upside.

Secondly, Moncada’s plate discipline appears to need some work. Based on my analysis, we’d expect him to post strikeout and walk rates slightly worse than average in Double-A next year. By no means does this constitute bad plate discipline, but given the rest of his numbers, it might be his biggest offensive weakness; it’s also completely normal for a player of his age, especially being projected against higher-level competition. Even if Moncada’s plate discipline is currently less than stellar, it’s certainly not unreasonable to expect him to make marked improvements in this area of his game.

In sum, if we put stock in these translations and Moncada were to spend 2015 in Double-A, we might expect him to hit something close to league average, if not a bit better than that. His walk and strikeout numbers would be a bit worse than average, but he would make up for it by hitting for power and running an above-average BABIP. From a 20 year old, this would be a very encouraging development. Most prospects who are that far-along at age-20 find their way onto top 100 lists, and many go on to become useful big leaguers.

Minor league stats can be useful, but should never act as a replacement for scouting. This is especially true for a player like Moncada, whose numbers came two years ago against competition that falls well short of what he’d see in the majors. Due to this imperfect data, Moncada’s KATOH projection is little more than a back-of-the-envelope calculation, so take it for what it’s worth. But caveats aside, the statistical projection seems to mostly agree with the scouting consensus: Yoan Moncada looks like one of the most promising young players in the game.





 

Chris works in economic development by day, but spends most of his nights thinking about baseball. He writes for Pinstripe Pundits, FanGraphs and The Hardball Times. He's also on the twitter machine: @_chris_mitchell None of the views expressed in his articles reflect those of his daytime employer.

12 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
tz
9 years ago

Great stuff once again Chris.

Do you know what the error bar would be (say +/- one standard deviation) of Moncada’s forecasted AA stats? With only 19 comps in the sample, it’s hard to gauge the possible forecast error.

1
Chris Mitchellmember
9 years ago
Reply to  tz

Thanks. I didn’t try to calculate error bars explicitly since my sample of players was so small, but they’d be pretty large. The transition from Cuba to the US is a big one, so its hard to predict how the stats will translate.

0
Least interesting man in the world
9 years ago
Reply to  Chris Mitchell

What was the R^2 on that regression? That would give us some idea to how much wiggle there is in the CNS to double-A translation you have.

0
Chris Mitchellmember
9 years ago
Reply to  Least interesting man in the world

I did a separate logarithmic regression for each stat. The R^2’s were all around .35-.65.

0