Andrew is part of BP's data staff, and has written the PECOTA Takes On Prospects series, as well as contributing two chapters to the 2015 Annual. Ask him anything.
Andrew Koo: Hi everyone. Just got out of class, and it's sunny outside, albeit still snow-laden. Let's talk!
Uri (FL): I've read a lot of comments by other BP writers - I'm curious for your take on Pecota's projection for Steven Souza?
Andrew Koo: I've received quite a few questions on Souza, so I'll tend to him first here. PECOTA's projection is very optimistic because Souza has performed extremely well in Double-A and Triple-A. When you can master those levels, you usually have a high chance of succeeding in the majors, so PECOTA's level adjustment and projection reflects that. It's a straight projection off his prior stats.
That said, PECOTA works primarily off comps, and Souza doesn't comp well. That means his underlying skills, according to PECOTA, are not as favourable, despite great stats. Russ Canzler, Justin Ruggiano, and Ryan Shealy are his top comps -- nothing too impressive there, so PECOTA's UPSIDE metric rates him as a good prospect, but nothing too special.
Jeff Moore and I have an article coming out tomorrow on hit tools and how PECOTA measures underlying hitting skills, if you'd like to learn about all the under-the-hood stuff PECOTA does.
Rick (Evanston): Any ETA on the Futures Guide? Waiting to hear when I can fork over still more money to you.
Andrew Koo: Our fearless leader Dave Pease says it's a few days away. You can even see the cover on our home page! http://www.baseballprospectus.com/
metsRok (NYC): Can you tell me why Kris Bryant will live up to his lofty projections unlike say a Alex Gordon, who was projected as a 300/30 homer guy before he took a MLB at bat
Andrew Koo: Kris Bryant hit 43 home runs in a declining offensive era and two offense-depressed parks, in AA and AAA. That's really rare. PECOTA confirms the projection by comping him to Giancarlo Stanton. It's as close to a high-probability success as one can get, based on just one season of full-time playing.
Also, to defend Gordon (although I wasn't around for that projection): he's touched .300/25. When we hear about a projection, consider that sometimes, it's for the player's peak.
Alex (Anaheim): Which Yankee prospect is more likely to be called up prior to September: Greg Bird or Aaron Judge?
Andrew Koo: Bird by a feather, as he's reached AA. Judge is the better overall prospect.
Ronson (On the Scratches): Breakout player pick?
Andrew Koo: A.J. Pollock. I'll say he could hit 3-4 wins if it all comes together.
Reggie Jefferson Slanderson (Lifehouse): If the members of one direction were a five man rotation, who would be the ace?
Andrew Koo: I outsourced to a friend, she says Harry Styles. Best singer, reliable, takes the lead. Liam Payne is the no. 2.
John (CT): Are the Houston Astros (the team formerly known as the Lastros) as much of a potential breakout contender as they appear to be? Assuming, of course, all stay healthy.
They have speed, defense, power and pitching that may surprise.
Andrew Koo: I think we're giving them an aura of breakout because of all the rebuilding and Luhnow-ing, though it's not this year for me. There's very little (immediate) impact talent on the present roster. Maybe in a year or two. Prospects take time in the big leagues to reach their potentials.
Ben (New York): I've been yearning to add a more "visual" element to my baseball research for awhile, and I'm under the impression R is a fantastic tool for this (among other things). I don't have any programming experience, however. What's the learning curve like in general for a newcomer to R with no previous experience? And do you have any sources you could recommend? Thanks!
Andrew Koo: For sure! I'm an R guy myself, and found it the hardest to learn of all the languages I know. But programming is all about perseverance for me. You're going to hit walls, and you'll need to grind and break through them. It's excellent you have baseball as a hobby: that passion motivated me to code and grind.
Find a data set (lahman works). Start with basic data manipulation (learn the package `dplyr`). The visual element is the package `ggplot2`. Easy, simplistic, visually appealing.
With regards to resources, try Max Marchi's Baseball with R site. Pull some code from it, and run it line-by-line and see how the data changes.
Also, follow @hadleywickham on Twitter - he's the god of R, and probably saved millions of hours of productivity with the packages he's written (he wrote both dplyr and ggplot2). He'll tweet/retweet useful links often.
Lime (The Coconut): Manny Machado: futre all-star or overrated?
Andrew Koo: Let's not count anyone out at age 22. Still on an All-Star path.
William (New York): Hanley or Tulowitzki this year in a deep league w/ OBP and SLG instead of AVG?
Andrew Koo: Tulowitzki. More power.
Chopper (Indy): I have heard more than a few prognosticators say that Jesse Winker may not be an impact bat. What are your thoughts? Thanks!
Andrew Koo: I think I'm higher than most. While he may not be "impact," the probability he makes MLB as a solid contributor is high; there's value in that. I can't say whether we as prognosticators overrate or properly rate "impact" talent, but I do feel we underrate high certainty.
Nedyo (Wichita): Edinson Volquez: Good, lucky, both?
Andrew Koo: Good but one of the most inconsistent, which perhaps makes it look like he's lucky when he has it going on. The Royals will appreciate the innings.
Nick Jonas (Your Radio): Is Arizona's catching situation really the worst in the league? Tampa Bay's seems pretty bad. So does Colorado's.
Andrew Koo: Yeah. The Rays and Rockies each have catchers from the Padres School of Catching now, which is interesting. I'd bet more on Rivera or Hundley than Tuffy Gosewisch.
Lime (The Coconut): Similar question as Machado but with Nick Franklin: bust or too early?
Andrew Koo: Definitely hasn't got Machado's upside, but let's not count him out after one poor season.
Cal Guy (Cal): Hi Andrew, Since power is becoming so scarce is Gallo's fantasy value increased in a 5x5 OBP league even if he becomes a three true outcomes hitter?
Andrew Koo: Yes, just by the economy of available power. He'll be better in an OBP league than AVG regardless. You'll have to live with the present high upside/high risk still, though.
Curious (Baltimore): Who has the higher ceiling among first base prospects, DJ Peterson or Matt Olson?
Andrew Koo: Olson's less further along the minor-league trail, but I think I'd lean toward him as he's just turning 21.
Mo Betta (Hoodhaven): Best comment in the annual?
Andrew Koo: I actually haven't read much of the book yet because of school. The Moises Sierra one is an all-timer. Adam Sobsey (who I shared the Jays chapter with, quite intimidating writing next to him) wrote Matt Boyd with the perfect changeup metaphor. Of the ones I wrote (not that you asked): Jonathan Gray, LaTroy Hawkins.
Mantis (Tobagin): Rank these prospects according to ceiling please: Trea Turner, Franklin Barret0, Ti'Quan Forbes
Andrew Koo: Barreto, Turner. I've actually have never looked at Forbes.
Graig Coldstein (Taco Bus): Better Nick Jonas single: Jealous or Chains?
Andrew Koo: I would not dance to either of these songs if played at a club. (try again?)
kbs (Key Largo): I need reassurance. 12 team, 5 x 5, OBP, $330 budget. I traded Joc Pederson, $4, & A. Almora, $4, for F. Freeman, $16, & R. Stephenson $4. I gutted the team last year for prospects, so I need everything.
Good, Bad, or Ugly?
Andrew Koo: Good, although dynasty/dollar valuation is not my forte. I'm rather low on Almora (to plug it again, Jeff Moore and I have an article on why, out tomorrow).
rbox (Waterloo): How do you personally feel about Rule 5 in regards to Oscar Hernández. Do you think it positively/negatively affected his career in 2014?
Andrew Koo: I haven't got much of a radar on the PD teams in Tampa and Arizona, though he'll get opportunities regardless. Neutral.
oxpaulo (WV): Do you think Carson Sands has #2 potential?
Andrew Koo: According to our prospect guys, he does. Wish I could offer more.
kylanje (Florida): Dylan Bundy or Masahiro Tanaka, pick one for the next 5 years for fantasy owners.
Andrew Koo: I want to say Tanaka for the upfront value you're getting this year, as Bundy's 2015 is a "maybe" and he'd likely be less impact than Tanaka this initial year. Beyond that, Bundy will probably exceed Tanaka in Years 4-5, so pick the one who fits your window. I'll go with Tanaka overall.
JP (TX): better side of the deal? I will leave out contract details and league factors.
side one: Correa, Finnegan
Side 2: Betts, Odor, Appel
Andrew Koo: Fantasy? Side one. Correa's impact talent.
Vic (Baltimore): 12 team NL 5 x5 minor league draft. Help me decide between Crawford, Schwarber, Winker, and Peraza. In my estimation, these will be the best 4 available. Please rank using ..... to show the gap between the above 4. Thanks so much.
Andrew Koo: Crawford. Winker... Schwarber.. Pereza
Eric (LA): It is entirely possible Tanaka's 2015 is a maybe too. I mean where to you put your odds of him getting TJS this season? 50%?
Andrew Koo: Quite true. I guess I'm not a medical expert to prognosticate that. Can I say I prefer the chance of him pitching this year over Bundy, who will probably tune up in the minors for a bit?
Jim (Florida): Thoughts on Ian Kennedy? Can he be a top 30 pitcher or have we seen who he is? It seems he is solid, durable, effective and could still maybe take one more small step forward.
Andrew Koo: The first three things you said listed -- solid, durable, effective -- are good precursors to a small step forward. Not every top-30 pitcher has that, so he might sneak in there through attrition of others.
dvanhait (Holland): I have asked this question before, but since no one has yet responded to it I considered it safe to try again... what's the scoop on Roman Mendez relief pitcher for Texas? Does he get a shot this year after posting an 2.18 ERA/ 1.12 WHIP in 33 innings of work last year?
Andrew Koo: He'll get a shot, but he doesn't miss quite enough bats to retain those numbers.
johnny 5 (florida): jose abreu, is this guy that good or is he more likely to duplicate his second half numbers
Andrew Koo: I think pitchers adjusted to Abreu in the second half, but he adjusted back, taking more walks, being more selective. That's a sign of a good hitter (which we already know he is), so I wouldn't worry.
Hudson (San Jose): Is there a type of prospect you think PECOTA is especially "smart" about, relative to scouting? Is there a type scouting is especially "smart" about, relative to PECOTA?
Andrew Koo: I'll answer the second question first: PECOTA can't read makeup at all. Makeup matters, even though we can't quantify it presently (teams are working on it though). Defense too, just because sabermetrics haven't collectively figured out quantifying defense 100 percent.
What is PECOTA smart about? Last plug on this: Jeff Moore and I have been working on an article for a while now about hit tools, and the broad range of skills it encapsulates. I won't spoil what Jeff and I posit PECOTA might be smarter about, but yes, there may be a type of prospect I think PECOTA reads better due to possible scouting biases. I'm going to be tackling it from the stat side, Jeff from the scout side. See BP tomorrow.
Andrew Koo: That's all for me today, thanks for your questions! I'm on Twitter @akoo if y'all want anything else, and for the prospect/fantasy questions I didn't get to, relay your questions to Nick Faleris or Bret Sayre. One of the best things about BP is how diverse our staff is. Thanks again!