Neil Weinberg FanGraphs Q&A – 8/27/14
2:34 |
Remember, this chat prioritizes question about our stats and our site, but I’m happy to take whatever other questions you have. Fantasy and prospect questions will get lower quality answers, though, sorry! I’m @NeilWeinberg44 on Twitter if you want to chat some other time. |
3:01 |
: Alright, let’s chat
|
3:01 |
: I’ll go an hour for sure, but maybe two if you ask many questions. So more questions = more chat.
|
3:01 |
Hey Neil, what are the best metrics to use to predict a power breakout for younger players? |
3:03 |
: To predict a coming breakout or to determine if it’s real? Really some mix of plate discipline numbers and something like ISO. You really want quality of contact data, but that’s not publicly available
|
3:03 |
Do you guys have a full list of players some where that can be extrapolated into excel? I would like to utilize the spreadsheet / hyper link abilities to better conduct an offline draft |
3:03 |
: If you go into the leaderboards and drop the PA minimum to 0, you can export that list.
|
3:04 |
I couldn’t find confirmation of this anywhere but the wOBA and associated wRC+ numbers for minor league players on Fangraphs are based on weights of their individual leagues, correct? |
3:05 |
|
3:06 |
: PS: Looks like we’re off to a slow start with questions coming in, if you want to get something answered now is definitely the time!
|
3:06 |
Doing a machine learning project and have a question. what single stat is best predictor of who will win the game? like if i predict home teams wRC vs away team wRC I assume the higher wRC will win. Is wRC the best stat to do this with? (wOBA, other linear run estimators better?) |
3:10 |
So if the Rockies have a .330 wOBA and the A’s have a .330 wOBA, the A’s are the better hitters. You also want to account for their pitching staffs in the same way. Really you want projected runs scored and runs allowed for each. I’d recommend you run some tests on all of them and see what works! |
3:10 |
Drew Smyly turning into an elite pitcher with the rays? He’s sure looking good so far. How do you think he’ll do in his start today against the orioles who seem to be smashing a few home runs every game here. |
3:11 |
Of course, rumor has it Cobb taught Odorizzi his splitter-ish thing last offseason…so…maybe! |
3:11 |
Do you guys have a different version of wOBA that includes intentional walks? |
3:13 |
1) You could calculate it yourself pretty easily if you want to |
3:13 |
Thanks for the response… wOBA of each team’s batters and wOBA vs for both pitchers are both inputs…. what else would i input to magnify that the COL 330 wOBA and OAK 330 wOBA are not identical? some sort of weighted park factor this wOBA was attained in ie % of time in each park multiplied by that parks park factor? |
3:14 |
: Just a nice park correction would do the trick. Park adjust wOBA…or just use wRC+ across the board.
|
3:14 |
Hi, any chance we see batted ball f/x within the next couple of year? What do you think will be the next big stat/data point? |
3:16 |
I think we’re a couple years away from assigning run values to exit velocity and direction. Something that can tell us how real a BABIP is. I know teams have some stuff like this. If the data is good, we should be able to make it work too |
3:16 |
Lucas Duda is showing a fairly negative defensive rating on fangraphs, but on the other site, he is showing a fairly positive rating, can you elaborate on how these two systems may be looking at things differently? |
3:18 |
|
3:19 |
Are there plans to expand the custom leaderboards to include minor league players? It would be really handy for tracking prospects along with MLers, even if the data isn’t as robust. |
3:19 |
: Someone asked about this last week. Will look into this!
|
3:20 |
Alcantara has some nice tools, but struggling mightily so far, he seems like he could be junior lake 2.0. Poor plate discipline and doesn’t hit for average, although his defense seems like he oculd hack it almost at any position out there. In your mind what do you forsee for him in the cubs future? |
3:22 |
: I’m not a scout, but I think he might just have been promoted too soon. He was never going to be a high walk guy, but his strikeouts went up and the quality of his contact went down when he got to MLB. I think a little more development time is needed, so I wouldn’t worry much. I imagine he’ll end up in the OF or traded. But too soon to say!
|
3:22 |
Neil – the FIP constant is designed to match league average ERA. Is the a purpose behind this other than aesthetics/convenience? It implies that a pitcher that allows all balls in play to produce a league average ERA on average. I am curious, if we compile all MLB innings in which a none of a HR, K, BB, or HBP are recorded and calculate ERA, what would it come out to. |
3:24 |
|
3:24 |
Thanks for the response. If a player has a high ISO, what would make it appear sustainable vs. not sustainable? |
3:26 |
: ISO stabilizes faster than SLG or HR rate, so it’s going to be the first stat to look toward. But there’s always going to be an inherent uncertainty about how their performance will translate
|
3:26 |
any data deceleration rates of pitches? which guys really are pulling the string the hardest? |
3:27 |
: If I understand this question, you want to know about the difference between arm speed and velocity, yes? In which case, we don’t have that data. But there is work being done to create a biomechanics tracker, which might create this data
|
3:27 |
If you were to start a roto fantasy league using only advanced stats, what would the categories be for your league? |
3:29 |
|
3:29 |
is there a quick stat that’s the diff between true avg and BABIP? |
3:29 |
: What do you mean exactly?
|
3:29 |
Why did MLBAM make Pitch F/X available to the public but not Hit F/X? Why might they hold back Statcast from the public? |
3:31 |
: PITCHf/x was never supposed to be public. As I understand it, someone discovered where they were keeping the data to generate the stuff on Gameday and then it was too late to go back. They haven’t released HITf/x because they were ready for that type of thing. They might hold statcast back if 1) the teams want to keep it secret for some profit driven reason 2) it becomes to difficult to release due to the size of the data
|
3:31 |
any chance framing and other periphery C skills eventually translates into a stat? |
3:32 |
|
3:32 |
Why do guys like Harold Reynolds and Matt vasgerian continue to say war what is it good for or why do they continue to say shifts don’t work? Why does MLB network tolerate their close minded attitudes! |
3:35 |
The shift thing is really silly because there is tangible evidence they could choose to look at. But Harold especially comes up with what he thinks is right and then searches for evidence in favor of his point. He always points out when the shift doesn’t work but never mentions it when it works. As for why they’re on the air, they’re charismatic and networks cater to who they think the audience will like. |
3:35 |
You’ve said that you don’t like using SIERA because it’s too complicated compared to xFIP, but isn’t it quite a bit better in some situations, specifically for pitchers with extreme batted ball stats (like Zach Britton and his 77% GB rate)? |
3:38 |
|
3:38 |
Hi Neil. I was wondering what your thoughts were on the traditional “Save” statistic. And more specifically do you believe in the “closer mentality”? |
3:40 |
I do not believe in the closer mentality. I do believe that handling pressure is a skill, but I do not believe that handling pressure in the 9th inning is different at all from pressure in the 8th inning or 3rd inning |
3:40 |
What do you think of Javier Baez… the power is obviously real but with his huge strikeout numbers can he ever really be a superstar? |
3:41 |
: I expect something just short of superstar. All-Star, not MVP. I’m not a scout, but based on the skills he has, I can’t see him ever being that good, even if he’s plenty good.
|
3:41 |
I hate the pfx classification of slider curve etc… what (horiz break, spin rate, etc) would be the best few metrics to use to bucket pitches |
3:43 |
: The classifications are the best we have. Dave is fond of saying their are bendy pitches and not bendy pitches. Really a pitch has several features (h-break, v-break, velocity). You want to clump pitches together that behave similarly, but pitchers are jerks and take a little off here and there to mess with us. You always want to do a sanity check on the classifications.
|
3:43 |
What is the conversion factor between slugging percentage and on-base percentage, as far as overall offensive value? If it’s some neutral run environment. |
3:43 |
: If I understand this right, OBP is 1.8x as valuable as SLG.
|
3:43 |
who’s your favorite team and player? Which teams do you follow the closest? |
3:44 |
: I’m a Tigers fan. My favorite players are Don Kelly and Rick Porcello. I follow most teams pretty evenly because of my jobs, but the ALC gets a lot of attention and also the Orioles because I go on a Baltimore radio show pretty often
|
3:44 |
Thanks for the response! I don’t know if I understand converting wOBA to runs. For some reason I was under the impression that Barry Bonds’ wOBA’s in 2001-2004 were “articifically” low since IBB’s weren’t included. |
3:46 |
To go from wOBA to runs above average, you take (wOBA – lgwOBA)/wOBA scale and then you multiply by PA. So in other words, Bonds’ wOBA doesn’t know he got IBB’d, but you multiple his wOBA by all of his PA, so his IBB are more valuable than the average player’s IBB! If that makes sense! |
3:47 |
How should one treat/utilize defensive numbers, DRS/UZR, for utility players when they have sporadic innings like 200-400 innings at 2B another 400 at 3B over multiple seasons? Should 1300 innings over multiple years be regressed more heavily than 1300 innings from a starter in one season? |
3:48 |
So if you care about a player’s defensive ability at 2B but you have 200 innings from 2012 and 150 innings from 2013, and a UZR/150 of +4, you’re just going to widen the error bars. |
3:49 |
Any chance of having righty/lefty splits on Minor Leaguers’ player pages |
3:49 | : I’ll mention it, but for now check our |
3:49 |
Is there a way to find a per-AB or per-PA context neutral run value for a particular hitter or pitcher? This is kind of useless but may be fun to look at. |
3:50 |
: This is the opposite of useless. This is basically all of sabermetrics!
|
3:52 |
: For hitters, find Batting in our value tab and divide by PA. That’s Runs Above Average / PA adjusted for park and league (and add in .108 if you don’t want the above average part). For pitchers, we don’t have this exactly, but you can create it yourself in a few ways without too much trouble
|
3:52 |
What is the expected BABIP for grounders? Is it safe to assume that a player like Travis D’arnaud is just very unlucky with his extremely low BABIP on grounders (.147) or is it inherent in data that is not yet available (hitter f/x). I would assume that for grounders there is an island between hard hit and very soft that leads to a lot of outs. |
3:54 |
http://www.fangraphs.com/le… Couple notes: 1) Lefties are higher with men on base because of the 1B holding the runner 2) hitters will vary around that level just like they do for normal BABIP. I would guess that .147 is unlucky, but I don’t know by how much without diving into his numbers. |
3:54 |
Should the Jays sell in the off season? Seems like the best possible time to shop Bautista. Coming off a good healthy year. |
3:55 |
: I suggested they trade Bautista last offseason, but no one would listen to me! I think the Jays have to decide what they want. They either need to spend more money or rebuild. They’re stuck in between and that’s the absolute worst place to be.
|
3:55 |
Are there any teams you know of that are installing tracking systems in their minor league parks? It seems to me these things would have a much greater value in helping to quantify the skills of minor leaguers, where statistics are so much less trustworthy than at the Majors. |
3:56 |
: There aren’t any I know for certain, but it’s only about $30K I believe to get something up and running, so I’d be surprised if some teams didn’t already have them. If you see a big black squarish thing behind the plate at press box level, that’s trackman!
|
3:56 |
Don kelly, really? Fair enough, I expected to here miguel after I saw you’re a tigers fan. Man is miguel ever struggling here, is the power coming back, or is this basically who he is as he enters the decline years? |
3:57 |
: I like utility guys. Always been my thing. Cabrera is hurt. Both recovering from his hernia and dealing with an ankle injury. I think this is 10% decline, 20% basic regression from a career year, and 70% injury.
|
3:57 |
is there anyway to generate any kind of defensive metrics (other than garbage like errors and fielding %) from pitchfx data? |
3:59 |
1) Find the average run value of each batted ball based on hit location, and battted ball type (so whats the run value of fly balls hit to this quadrant of the field). that’s basically how it works, with some other nuance worked in |
3:59 |
if the Tigers don’t make it to the WS, which teams would you prefer? |
4:00 |
: Nats, or Pirates if they make it.
|
4:00 |
: Two minute break. Throw more questions!
|
4:04 |
Who is Kyle Lobstein? And would the Tigers be better off rolling with four starters for the last month? |
4:06 |
I’ve thought about the 4 man thing and I think it’s probably too costly to execute over a full month with a potentially injured Verlander and only two off days because pitchers do worse on short rest and only Price has ever done it (has price even done it?). |
4:06 |
Is there any hard-and-fast way FBs and LDs are differentiated? Like is there a trajectory cutoff mark, or is that data somewhat arbitrary? I find it hard to tell the difference sometimes and wonder how the data measures borderline calls on balls hit to the outfield |
4:07 |
: This is basically why batted ball data is only a rough guide. There’s no perfect way to do it.
|
4:07 |
ed on BeerGraphs? |
4:07 |
: I do not know what this means?
|
4:08 |
: Ask more questions, if you want more chatting! We’re running painfully low!
|
4:10 |
Looks like that question got cut off. Do you know how Beers Above Replacement is calculated on BeerGraphs? |
4:11 |
Eno is in charge. There is beer stuff. The web address. However, http://beergraphs.com/bg/10… I don’t know anything about beer! At all! |
4:11 |
What’s my trade value? |
4:12 |
: Not that high. You’re a solid MLB starter, but you’ve yet to show greatness at 26 and are going to get two more arb years. Not cheap, not amazing. Someone who pay for you, but not a huge sum….I think. But I haven’t done a deep dive into Leake in quite some time
|
4:12 |
alex meyer in september? |
4:13 |
: I think so.
|
4:13 |
Is my current performance the best I’ll be, or do I have room for growth. Peripheral numbers don’t love me, but could I be a front-half-of-the-rotation guy? |
4:15 |
: ERA will go up, but could see FIP coming down a bit at his peak. Stay healthy and a 3.5 WAR type year is possible.
|
4:15 |
Jake Lamb…legit home run hitter in the major leagues? perhaps 25+ with a strong OBP? |
4:15 |
: Not sure. I honestly haven’t seen enough of him to comment with any authority.
|
4:16 |
Okay, I know you said no fantasy stuff, so my question would be, what are some metrics that we can look at which will even out next year (and give us a good chance at finding some fantasy sleepers). |
4:16 |
: BABIP is your friend in fantasy. Also, look for pitchers who are moving to a better defensive team or who have new, better fielders.
|
4:16 |
Do you see the Twins competitive in 2016. They should have a lineup similar to CF: Buxton, 2B: Dozier, 1B: Mauer, 3B: Sano, DH: Vargas, RF: Arcia, LF: Plouffe, C: Suzuki, SS: Santana…with a rotation of Hughes, Gibson, Meyer, Milone, Berrios. |
4:17 |
: That’s not enough to be anything more serious than a fringe WC team. But, if those young guys develop and you add in a nice FA or two, you might have something really nice.
|
4:17 |
Alex Gordon said recently that he’s planning on honoring his 2016 player option. If he has a 4ish win season next year, how much $ would he be giving up? |
4:17 |
: Let’s run this quickly, one sec.
|
4:21 |
: I’d say $7+ million in 2016, plus the added risk of not signing a long term deal. If he comes out after 2015 with five straight great years going into age 32, he’s getting *at least* 4 years and $70 million. That’s the floor. If he gets hurt in 2016, that can fall fast. I don’t know why he’s doing this. Even if he wants to stay in KC, there’s no value to exercising the option in my opinion
|
4:21 |
Who busts out of Baez, Bryant, Russell, Soler? |
4:21 |
: Baez/Soler toss up. Think Bryant is the safest of the group
|
4:21 |
is a BJ Upton for Edwin Jackson trade actually a possibility? and who would, er, “win” that trade? |
4:22 |
: Don’t think so. And I’m not sure anyone wins. You’re trading one mess for another. Odds of each turning around seem similar. I think I want Jackson, because maybe he can be a reliever? oof
|
4:22 |
Do you think making there be less time between pitches would alter overall run scoring? |
4:23 |
: Interesting question. I think it would increase run scoring slightly, but just a guy feeling about which player would be bothered more
|
4:23 |
Am I a trade candidate in the offseason? If I finish strong, what do I fetch? |
4:24 |
: Astros should trade Carter. Dangle his HR power in front of the right GM and get a nice little prospect back. No one amazing, but something good I bet.
|
4:24 |
![]() Bottom row are 4 of best RHP throwing changeups to LHB Is there any information based on location alone as to why the top 4 are much better than the bottom 4 or is velocity movement and sequencing just as much if not more of a factor |
4:26 |
: Every RHP wants their CH down and away to lefties. I’m not sure location alone explains it because a good changeup is about the difference between the time and place the batter expects and the time and place it ends up.
|
4:26 |
prediction for world series MVP. go. |
4:26 |
: Rendon
|
4:27 |
When should the Mets look to lock Juan Lagares up? I think they should do so soon before he starts to hit and drives his value through the roof, kind of like the Cards did with Yadier |
4:28 |
: After 2015. His skills are underpaid in arb and he doesn’t hit HR. Plus the fact that defense peaks early, so you might not want him in five years anyway, could definitely wait. But if you want to do it, wait a year
|
4:28 |
Hey Neil! Been waiting all week to ask this question! How can you determine whether a pitcher is really experiencing bad luck as opposed to drop in skill level? For example, if a player has a sharp increase in BABIP and no obvious loss of velocity paired with decrease performance. |
4:30 |
Basically, really try to isolate exactly what’s different and then decide if that seems like luck or skill |
4:30 |
Seems like there’s been a weirdly fantastic number of 3Bs lately. Who would you take if you were a GM: Donaldson, Seager, Rendon, Beltre, or… Frazier, maybe? |
4:30 |
: Donaldson/Rendon, then Seager. Beltre only below them because at some point he will stop being ageless
|
4:31 |
My eyes tell me that, with his arm, Andrew McCutchen should play RF at PNC once Polanco is up for good. Marte in CF and Polanco in LF. Can you confirm or deny with stats? |
4:32 |
: Cutch’s arm doesn’t rate well based on the numbers. I’m not sure if you’re talking about the big left field in PNC as a factor, but the Pirates have a nice problem in basically having three guys who could handle each spot. I’d really go by who is comfortable where for now
|
4:32 |
What’s with Zach “All or nothing” Walters? Clearly his current power numbers aren’t sustainable. Do you see him ever balancing out to become a more well rounded hitter? |
4:33 |
: Yeah, I actually think he might be pretty decent.
|
4:33 |
The Jays have a few young pitchers in Stroman, Norris, Sanchez and Hutchison, as well as a bunch of other pitchers in the lower levels, could the rotation actually be a streng in a few years? |
4:33 |
: Could be? Sure. But it’s pretty rare to hit on a bunch of pitching prospects at one time. I’d bet the under, but it’s not crazy
|
4:35 |
Bryant the safest with 40% k rate?! |
4:35 |
: Bryant has a 27.5% K rate?
|
4:35 |
: This year. Was 27.4% in 2013. Never higher than that
|
4:35 |
i just read that pitch f/x at different stadiums can be different. for example a pitcher’s fastball at home may be 1-2 mph slower than on the road. is this sample large enough to create discrepancies in the data? |
4:36 |
: Yes. Brooks Baseball adjusts for this. Our raw data does not. Use with caution.
|
4:36 |
Who should Nats focus on resigning after 2016? Also what contrct do you think Stras gets in the open market? |
4:36 |
: Let me see who they have due up really quick
|
4:38 |
: Stras is their only big name due up after 2016, right? And I would imagine he’s get something like 7/$200M at that point. Which is an estimate based on many assumptions about him and the market. So grain of salt
|
4:39 |
Does Aaron Sanchez results in the bullpen signal that maybe he is actually better than some people thoght? |
4:39 |
: Pitchers do better in the bullpen. I think he’s a guy who is going to be better in the pen than the rotation. Not saying they should move him yet, but I haven’t changed my view of him based on this stint
|
4:39 |
Zach Walters: The 24 year old Javy Baez nobody is talking about? Walters: .207/.266/.529, 9 HR in 94 PAs, Baez: .198/.233/.465 with 7 HR in 90 PA |
4:40 |
: From what I understand, the Baez bat speed relative to Walters is much different. Although, scouts can be wrong and Walters has made some adjustments. 90 PA is too small to adjust your beliefs much
|
4:40 |
Given his struggles this year, do you see Masterson signing a one year deal to establish some of his value? Thoughts on likely landing spots? |
4:41 |
: Yeah. One year makes sense. Cubs seem like a good fit. Twins? Good question
|
4:41 |
Can you explain why Baez struggles so much to make contact? |
4:42 |
: I assume pitch recognition is a big factor, but I do not know for sure
|
4:42 |
You worried about Brandon Moss’s weird power outage at all? He hasn’t’ hit a home run in over a month, and Brandon Moss’s game is like 90% dingers. But tiny sample size, too. |
4:43 |
: Probably better to think of this as too little power and the previous months as too much. Not super worried, but I wasn’t buying him as a star
|
4:43 |
why no love for corey dickerson |
4:43 |
: He is good! I voted for him as my NL LF in the BTBS ballot.
|
4:44 |
Okay….let’s try that one more time…NEIL. Can you clarify this point you made above ” ISO stabilizes faster than SLG or HR rate, so it’s going to be the first stat to look toward.” |
4:44 |
: If you want to tell how good a player’s power numbers are going to be, ISO will be a better indicator because you can get the necessary sample size faster
|
4:44 |
Does Kyle Schwarber have any trade value? |
4:45 |
: Oh, certainly yes.
|
4:45 |
speaking to your time and place about changeup…. is that why felix hernandez’s is soooo good is because it is the same speed as his fastball but just falls like 6 inches off the table and they swing on top of it? |
4:46 |
: that is one reason. You can ruin a hitter with movement they aren’t expecting, speed they aren’t expecting, or both. Felix throws two pitches at 90 mph that do totally separate things. How do you hit that?
|
4:46 |
Forgetting for a moment all the failings of RBI as a stat…isn’t it kind of strange that hitters don’t get an RBI for knocking in a run if they hit into a double play? Making one out to score a run is acceptable, but making two? No way! That’s up there with running on a dropped 3rd strike, and bunting foul after strike 2 as rules that I think were made up in the driveway. How often does the negative impact of making those two outs outweigh the positive impact of scoring that run? |
4:46 |
: Point one, the rule has no logical foundation. It’s just a made up scoring thing. Let’s play a game through and use RE24!
|
4:48 |
Empty, two outs is .095 RE So actually it is a net negative! Shoot! |
4:48 |
Evan Gattis for Mookie? |
4:48 |
: In a HR Derby, this is a good idea for the Red Sox. Otherwise no.
|
4:48 |
Buxton or Bryant, who has a better career? |
4:49 |
: Bryant but I think both are All-Stars
|
4:49 |
There’s a deli in my town called Weinberg’s. No question, just wanted to say thanks for the killer Reubens. |
4:49 |
: Where do you live? There’s one in Raleigh that I though was just okay 🙁
|
4:49 |
Just read that the Indians have 47 sac bunts this year (The league average is 37!) As an Indians fan, how should I feel about this? |
4:50 |
I talked to someone in the game about their team’s bunting habits and he told me he thinks a portion of theirs were guys going on their own |
4:50 |
Any comps for Bryant? Troy Glaus? Mike Schmidt? |
4:50 |
: Glaus is one people use a lot. Makes sense
|
4:50 |
Is there anyway to calculate the league average numbers for a 5th starter? I suspect that a decent (and consistent) 5th starter is one of the more under appreciated types of player in baseball. (Ask the Tigers!) |
4:51 |
: Sure. Grab every starter and take the average numbers for the fifth best guy in each rotation!
|
4:51 |
What do you think Melky Cabrera could get in free agency? |
4:51 |
: 4/50?
|
4:51 |
Your answer got me excited… Could Schwarber headline a package for a top player? For example could Schwarber now have been traded for a guy like David Price |
4:52 |
: Always depends on contracts, but he’s a very good prospect
|
4:52 |
does the slash line and plate discipline of mike trout this year seem like an appropriate projection of kris bryant next year? or am i getting too excited |
4:52 |
: This seems like a plausible peak for Bryant. Give him three years
|
4:54 |
: Alright, let’s finish this one up. Thanks for the questions. Sorry if I didn’t get to yours. I’m @NeilWeinberg44 on Twitter if you need me.
|
4:55 |
: Also look for new content in the Library on Fridays!
|
4:55 |
remind everyone schwarber is in high A |
4:55 |
: Done. Take care!
|
Neil Weinberg is the Site Educator at FanGraphs and can be found writing enthusiastically about the Detroit Tigers at New English D. Follow and interact with him on Twitter @NeilWeinberg44.