Neil Weinberg FanGraphs Q&A

Neil Weinberg: Hey everyone, this is my first chat since joining FanGraphs, so let me get you caught up.

My job here is primarily to serve as something of a Site Educator. This means that I’m responsible for providing resources to readers about our data, advanced metrics, and how to make use of the features available at the site.

It’s possible you’ve seen me around the various baseball parts of the internet. My Twitter handle is @NeilWeinberg44 and if you want to get in touch about these kinds of questions at times other than 3pm on Wednesdays, that’s where to find me. It’s my job, so take advantage of it.

With that said, I will prioritize “how does this work?” and “what does this mean?” type questions during the chats, but feel free to ask regular questions about anything as well. Although if you’re here for fantasy advice or prospect talk, I’m probably less helpful than some of our writers. And if you’re here to discuss beer, I am literally the least useful person on our staff.

Think that’s it, queue is open and we’ll start at 3pm!

Neil Weinberg: Alright, let’s chat! I’ll probably hang out for an hour and fifteen minutes today, but will go longer most days if there are lots of questions.

Comment From DD
I can never find a direct link on the site for the Live Scoreboard Page. I always have to do a google search. Embarressing I know. Please point me in the right direction. Thanks!

Neil Weinberg: This is a link to Sunday’s. http://www.fangraphs.com/sc…

What you want to do is go to FanGraphs and look for the “Score” tab at the top. Hover over that and then click the current year (so 2014) and it will take you to the current day. Unfortunately, I can’t show you this in action because today is a very dark day on the calendar.

Comment From @BobbleHeadGuru
Why are there two different WARs. Why is fWAR better?

Neil Weinberg: There are two different WARs (actually 3!) because there are different ways to measure certain aspects of performance. Typically, most of the hitting measurements will be the same, but the different sites have different ideas about how to properly measure defense and baserunning…and pitching for that matter. I prefer fWAR, but I recommend looking at all of them. I’ll put together an updated Glossary entry about this in the future to break out the exact details.

Comment From John
Thanks for doing this Neil. If Player A has a 120 wRC+ and Player B has a 80 wRC+, is Player A providing 1 1/2 times more offensive value than Player B (120/80) or is it more complicated than that?

Neil Weinberg: It’s not quite that simple, because wRC+ is a rate stat. So it’s going to depend on plate appearances too. I’ll be updating the wRC+ entry next week, hopefully, so stay tuned for that.

Comment From Pale Hose
I forgot this was going to happen and now it’s here. Wednesday afternoon just got better.

Neil Weinberg: I mean, don’t judge it before you’ve seen it. I might be awful at this. But thanks!

Comment From Zack
Is there anywhere on Fangraphs (or another site) where I can find daily lineups posted? Bonus points for .csv or similar format

Neil Weinberg: Are you looking for lineups the day of? If so, we post them as they happen into the live scores page I linked to earlier. Also, there’s an MLB lineups Twitter account that’s helpful. Can’t speak to CSV, but Baseball-Refernce has a page where you can look at a team’s lineup for each day of the season.

Comment From LeBaseballJames
Hey Neil- looking to get my girlfriend educated on baseball. Since shes not a fan in any way, it should be pretty easy to mold her mind sabermetrically. Any good pieces of literature you’d recommend?

Neil Weinberg: So this is going to depend on her style of learning, but when I was teaching my wife about the game when we first met, I just explained the game to her with sabermetrics built in.

Neil Weinberg: If she likes to learn by reading, The Book is good if she’s stat-savvy. Or pull up some articles from our soon-to-be updated Glossary! Also, BTBS and other sites have some great primers. There’s no one resource, just expose her to lots of information and she’ll pick it up pretty easily.

Comment From @BobbleHeadGuru
Is it possible the Miguel Cabrera is the BEST defensive 1st baseman in the AL? According to Fangraphs, yes. However, does this match with gut feel, no.

Neil Weinberg: So I doubt that he’s actually the best defender in a true talent sense, but he’s actually pretty good over there relative to how he played at 3B. I bet is also helps that he can position himself accordingly now that he has a quality partner at 2B. During his early years at the position, it was a revolving door of questionable second basemen.

Neil Weinberg: This is a good time to point out that defensive stats aren’t perfect in small samples. But they aren’t nearly as bad as people say. The key is not to overreact to a change in those numbers. If a guy is +5 for four straight years and is all of a sudden -15, it’s probably going to regress

Comment From Anon21
More of a feature request: “Advanced” stats above “Standard” stats on Splits pages? I feel like Fangraphs readers are typically less interested in a guy’s home RBIs or whatever than his wRC+ (although maybe the splits page is really geared towards the slavering fantasy hordes).

Neil Weinberg: Not sure if there’s a specific reason we do it this way, but this is a good time to point out that you’re more than welcome to send me requests like this. Not sure how much authority I have, but I’m willing to ask on your behalf.

Comment From Pale Hose
I’m not sure if you are prepared to answer this yet, but how can a pitcher have an ERA lower than FIP and RA9 WAR lower than FIP WAR?

Neil Weinberg: There are a couple of possible ways this can be true.

1) Their home ballpark. ERA and FIP aren’t park adjusted and both WARs are. So if you play in Petco, your pitching WAR is lower than if you had the same stats in Colorado.

2) Innings. WAR is a cumulative stat. A guy with 20 innings of a 1.00 ERA is not as valuable as a guy with 80 innings of 2.50 ERA, for example.

3) Unearned runs. RA9 includes all runs, so if you have 70 ER but 85 total runs, your ERA might look good but you RA9-WAR won’t.

4) Infield popups. I don’t remember when this happened, but infield fly balls are treated as defense independent in fWAR, but we don’t include that in the FIP calculation because FG didn’t invent FIP and that might be confusing!

I think that’s all of the reasons. Great Q!

Comment From The goggles, they do nothing
I WANT TO DISCUSS BEERS

Neil Weinberg: Sorry, try Eno tomorrow!

Comment From Reed
Neil, where would you rank Houston among front offices? Also do you think the backlash over the tanking stuff is just because people don’t ‘get it’?

Neil Weinberg: I’ll say this. I am a huge fan of the overall approach Houston is taking. I love that they are really digging in and rebuilding the org with all sorts of new strategies, etc. However, I have no idea if they are actually doing it well.

Neil Weinberg: I think the backlash was expected. If it looks like you’re intentionally not fielding a good team and are not spending and hoarding draft picks, it seems like you’re trying to cheat the system. But that’s exactly what the system rewards. The Astros are being bold about it and people don’t like it when someone rocks the boat.

Comment From JC
What are the most important advanced stats for hitting and pitching? Why?

Neil Weinberg: The short answer is wRC+ for hitters and, I would say, FIP- for pitchers. wRC+ does everything you want it to; weighs value of the action, park adjusts, controls for context, etc. I would say park adjusted FIP because it tells you more about the pitcher’s performance than any other single stat. But I would encourage you to use other stats too. Look at K, BB, ground balls, everything.

Comment From @outfieldgrass24
Neil, really looking forward to your work here! Question: what’s the case for using wOBA over wRC+ when discussing a player’s offensive output. When would you ever not want that number adjusted for park/defense/etc? I always revert to wRC+ but I see wOBA everywhere. What gives?

Neil Weinberg: wRC+ is “better,” but it’s also a little bit more foreign. We’re used to seeing a period and three numbers, so wOBA is easier to pick up for most people I think. Also, if you want to make your own estimates, working with park factors can be difficult. I can tell you a guy’s wOBA in 15 seconds if you give me his raw numbers. Can’t do that with wRC+, have to ask FanGraphs!

Comment From Will
Difference between FIP and xFIP?

Neil Weinberg: FIP is based on strikeouts, walks/HBP, and home runs allowed. xFIP is based on strikeouts, walks/HBP, and the expected number of home runs you would allow if you allowed a league average HR/FB% (about 10%). So if you allowed 100 fly balls, you’re charged with 10 HR by xFIP. This is because HR/FB% talks a long time to “stabilize” and is rarely telling you much about true talent. FIP tells you want happened, xFIP tells you what we think might happen with average HR/FB%.

Comment From Guest
Hi Neil, very glad to see a Q&A session devoted to these kinds of questions. I have several, but for now, just this: how are batting runs determined at FG? I would have thought they would be wRAA using the wOBA formula–[(wOBA = lgwOBA)/scaling factor] x PA, but I calculated this for several players, and in most cases, it was a little off. And quite a bit off for Tulo, 31.2 vs. 38.8. I used the data available at the ASG break, and the latest coefficients from the guts page.

Neil Weinberg: So wRAA is not park or league adjusted. Batting runs are park and league adjusted. That’s why Tulo’s are so far off. Coors is a great place to hit.

Comment From bassam
What is the best advanced stat to look at if i want to look at command?

Neil Weinberg: This is a tough one, because command depends on what the pitcher meant to do, and we don’t publicly track where the catcher sets up. (I think BIS does!). So you might want to look at Zone%, but there isn’t really one stat that can capture command.

Comment From Matt
What is the best stat: RE24 or RE24-based WAR? Those are the only options.

Neil Weinberg: Hi Matt. RE24 is great if you want context dependent numbers, but some of use like context neutral numbers! I often compare the two!

Comment From @outfieldgrass24
FIP is really the backbone of FanGraphs’ collective pitching metrics (fWAR for pitchers). How does it vary from RA/9 WAR and what are the merits of each? When should I choose fWAR and when should I choose RA/9 WAR?

Neil Weinberg: Both metrics adjust for innings pitched, league, and park effects. fWAR uses FIP (strikeouts, walks, HBP, HR) and RA9 uses all of the pitchers runs allowed. FIP estimates a pitchers ERA assuming they had league average defense, sequencing, batted ball luck. RA9 assumes everything that happened is on the pitcher. The truth somewhere between them, but probably close to fWAR. Keep an eye on both! But some pitchers are consistent FIP-beaters. I’ll expand on this in a week or two in a post!

Comment From yeahimajerk
Any chance of getting a mobile site so I can pull up stats easily and argue with my friends at the ball park?!

Neil Weinberg: So I don’t think this is in the works, but I had a similar conversation with some friends the other day and my advice was to get a bigger phone. Sorry, that’s the best I can do for now. There’s an iPhone app that works for some stuff, or you can pinch and zoom!

Comment From Matt
Why isn’t there a good rate stat for context-dependent hitting?

Neil Weinberg: This is a good question. Sounds like someone could make a name for themselves making an RE24-rate stat.

Comment From Mike
In your opinion, what are the most important advanced metrics to understand for people new to the analytic side of the game?

Neil Weinberg: Love this question.

Neil Weinberg: Learn wOBA. Learn about FIP and BABIP. Learn to speak in terms of runs and wins. If you can do that, you’re probably 80% of the way there.

Comment From David Price
When am I getting traded? Am I getting traded and to where?

Neil Weinberg: You will be traded on July 28th to…um…Seattle?

Comment From Kevin
What is the relationship between ottoneu and fangraphs? Does ottoneu get all of its data from Fangraphs? Where does Fangraphs get its data from?

Neil Weinberg: As far as I know Ottoneu and FG are either partners or FG owns Ottoneu. Maybe I should ask. But the data at Ottoneu is from FG. And FG gets our data from a variety of sources. You can seem some of them at the bottom of the page if you scroll all the way down! Some of the data is out there for free, but we buy a lot of stuff so that you can use it. So use it!

Comment From Jordan
according to projections, miggy is supposed to be better down the stretch, you think he’s going to be a lot better?

Neil Weinberg: I think the projections are about right. I’ve been looking into him trading HR for 2B this year, and I’m not sure if that’s going to keep up. He won’t be 190 wRC+ Miggy, but I think 160 wRC+ Miggy will be around.

Comment From a eskpert
Okay. It’s 1999. You can draft Josh Hamilton knowing he won’t struggle with alcohol or drugs, Jose Bautista knowing he has his swing figured out now, or Albert.

Neil Weinberg: Pujols. Assuming I know about Pujols too.

Comment From Jason
Why does Fangraphs sort it’s WAR leaders (and losers) by a qualified number of plate appearances?

Neil Weinberg: I don’t think people would like it if we had a bunch of guys with 4 or 5 PA gumming up the leaderboards. You can always change it.

Comment From Dr. Met
Do you have have individual game velocity and pitch type charts by inning or even aggregated for the game? For instance, I want to know what pitches, how many and how fast Verlander threw in his last start? I’ve been having to use Brooks for this info.

Neil Weinberg: Doesn’t look like Brooks, but you can access that info in our Game Logs for averages and such. http://www.fangraphs.com/st…

You can also get the plots. http://www.fangraphs.com/pi…

Comment From Mike
How many hours a day do you devote to baseball?

Neil Weinberg: Actually watching and doing stuff? About 4-5 maybe. But I don’t want to tell you how much that would go up if I included just kind of thinking about it or being on Twitter.

Comment From Matt
What is your favorite stat? Not necessarily best, or most useful, just favorite.

Neil Weinberg: Favorite? I’m partial to GB% for pitchers

Comment From TB31
If you go to the League Stats page, the wRC+ for the league is 96. Shouldn’t it be 100 by design?

Neil Weinberg: Click on “NP” in the position tab. It’s 100 for position players.

Comment From Zach
What stats do you keep on your dashboard that aren’t part of the default set

Neil Weinberg: None. I use the default set because I didn’t learn about that feature until I got used to it and my brain hates change.

Neil Weinberg: And there are a lot of good questions in the queue. If I don’t get to them, please come back next week and ask again. I’m just getting used to the whole chat pace!

Comment From Pete
Why is BA with RISP not considered a valid stat for measuring a player’s contribution to winning. RBI are team dependent but what a player does with RISP should be a worthwhile metric.

Neil Weinberg: There is little evidence that getting a hit with men on base is a skill separate from getting a regular hit. More importantly, people like to talk about hits with RISP but never about how you generate those RISP chances. Both are equally important.

Comment From LeBaseballJames
Im confused about FIP+ versus FIP- (same goes for wRC + and -). I understand them and what they measure but is it just preference of which one you would use?

Neil Weinberg: So you should basically use a – stat if lower is better in the regular stat and a + if higher is better in the regular stat. There’s a more involved explanation, but you want to use minus for pitching and plus for hitting in most cases.

Comment From Chris
i’m confused on how to create a custom leaderboard on the leaders page. Do you move the stats you want to the left, right, or neither?

Neil Weinberg: You want to move the stats into the right hand column of the two!

Comment From Guest
Favorite team? Favorite players? Rooting interest?

Neil Weinberg: Tigers. I like utility players and ground ball starts. Don Kelly and Rick Porcello are my guys.

Comment From Hieronymous
Which pitching stat do you like the best: FIP, xFIP, SIERA, or something else?

Neil Weinberg: FIP and xFIP, depending on the question I want to answer. SIERA is a little more predictive than xFIP but it’s a lot more complicated for that extra little bit.

Comment From Boyz in the Hood Question
Jon Singleton long term … can he be the next Rizzo or Goldy?

Neil Weinberg: I don’t think he’ll have that type of approach. Good power, should be a solid player, but those two guys are stars for me.

Comment From Brad
Is OBP a better rate in evaluating a player then slugging % and if so, why?

Neil Weinberg: Getting on base is more important than hitting for extra bases. About 1.8x as important. That’s why OPS is imperfect, it weighs them evenly. wOBA is going to be best.

Comment From Nick
Does wRC+ factor in positional adjustments as well?

Neil Weinberg: Nope. wRC+ doesn’t know anything about position.

Comment From Guest
Neil do you also write for Forbes?

Neil Weinberg: I don’t. Different Neil Weinberg!

Comment From Zach
Re: WAR. WAR isn’t a rate stat, so if a guy with 4 or 5 PA were up there that would certainly be worth seeing! A big article on BtBS recently pointed out that Corey Dickerson, for example, is among the league-leading LFs in WAR but isn’t a qualified hitter so doesn’t show up.

Neil Weinberg: My point was that they would clog up the room around 0.0 WAR. If you want to look only at the top guys, take the filter to 0 PA.

Comment From isaic
Is it possible to add bWAR or WARP to Fangraphs, or is there a restriction or agreement that prevents that? I’d personally love to be able to look at them all side-by-side, and I bet I’m not the only one.

Neil Weinberg: The other sites own their stats. I don’t think that would work out, but I encourage you to look at multiple WARs and pick your own personal cocktail.

Comment From Brian
Are reverse splits over a few months (Ellsbury to name a hitter, Jason Frasor to name a RP) predictive of how they’ll perform the rest of the season?

Neil Weinberg: Almost never.

Comment From CuriousGeorge
What’s your favorite site/source for contract details?

Neil Weinberg: Cot’s is great. We’re adding more contract info all the time, but it’s not as complete.

Comment From AJP
Thoughts on Verlander going forward, this year, next, etc?

Neil Weinberg: going to be good, but he’s done being great

Comment From Christian
I’m not sure if you are familiar with “The Maddux”, but it’s a shut out complete game with less than 100 pitches thrown. I tried to see if anyone has thrown one this year, but I couldn’t figure out how to do it using the custom charts. Is it possible for me to create a chart on here that shows “Maddux” starts?

Neil Weinberg: Baseball-Reference’s Play Index will do the trick. We don’t have that.

Comment From Tucker
Just clarifying your point from earlier: because xFIP uses a base HR rate, would that explain why Masahiro Tanaka has a much higher FIP than xFIP?

Neil Weinberg: Correct.

Comment From Hayden
What are the chances of new video-tracking data ever becoming public, at least to a degree? I want to be able to see it, because I think there’s a paradigm shift in analytics on the way.

Neil Weinberg: I think it’s going to happen. But like others have said, we don’t know until we see it. I think they’re really unsure how to even make it public. It’s a lot of data.

Comment From Tucker
How do you guys at FanGraphs get all of your data? Is there a SQL database somewhere that’s accessible?

Neil Weinberg: We don’t have an open database, and the data comes from a variety of providers. There are a lot of good databases out there for various stuff though, if you’re looking to play around.

Comment From Mayton Doore
Is it time to admit the Shields-Myers trade wasn’t all that bad for KC?

Neil Weinberg: I’m a believer in evaluating trades with the information you have at the time. At the time it looked awful. Now, it still looks bad to me.

Comment From Luke
When do you like to use FIP and when do you like to use SIERA?

Neil Weinberg: I never use SIERA. Too bulky. I use FIP or xFIP when I want a DIPS stat.

Comment From Guest
There are three main sites publishing value metrics that I know of, FG, BBRef and BBPro. You seemed to allude to this in answering another question. Is this correct, are these the main three? And if so, might it not be useful for FG or another site to publish a daily average WAR, for the top players at least? I have actually been doing this for a while for a few players that I follow.

I also wonder if anyone has analyzed the correlations between these three sites. In some cases they differ a lot defensively, and even offensively they are not as close as one would think.

Neil Weinberg: I think comparing the WARs is very useful, but one problem is that our defensive metric for WAR is UZR, which is updated weekly. But in general, making use of more data is good!

Comment From Zach
The other day Dave Cameron said the first three stats he looked at for batters were K%, BB%, and ISO. Why ISO and not wOBA? If two hitters have the same wOBA, aren’t they just as valuable as one another even if all of the hits for one player are singles?

Neil Weinberg: I think Dave goes to ISO because he wants to know about extra base power and a player’s approach (BB, K). wOBA is a better indicator of value, but Dave was talking about looking at how a player is getting there.

Comment From BoB LobLaw
What’s the most important stat to tell if a hitter is unlucky. And a pitcher is unlucky.

Neil Weinberg: Compare their BABIP to their career BABIP. And look at their HR/FB% compare to league and career averages.

Comment From Pale Hose
Does multiplying wRC+ and park factor tell you anything meaningful about performance in a specific context? For example, would Marlon Byrd in Seattle be Marlon Byrd wRC+ times Safeco park factor?

Neil Weinberg: On average a player should have the same wRC+ in every park. That’s not going to be exactly right because of handedness and such. But basically, it should be totally neutral.

Comment From Batch
Hey Neil, do you see any new defensive metrics coming out of the new fielding tracking system that’s supposed to be in place by 2015? Or maybe a way to revise formulas for UZR, DIPS, etc.?

Neil Weinberg: If the data becomes public, I’d expect to see revisions in the current stats but maybe not until 2017-2018. In order to build a good stat, you need a lot of data. Can’t just grab a week of StatCast and know exactly what it means.

Comment From Tom
Looks like the defensive component of catcher WAR is just rSB again this season? Didn’t it incorporate RPP in previous years? How does that work and what has happened to that stat? Maybe I have missed something, but seems like catcher WAR will therefore automatically be lower than previously…

Neil Weinberg: I’m not positive, but I think it includes blocking pitches. Might just not show up in the leaderboard. will have to check

Comment From Guest
are you a Zach Britton fan, then? 78% GB!

Neil Weinberg: Yes, all of the ground balls please.

Comment From David Robertsons Calves
What’s your preference on how players wear their pant legs?

Neil Weinberg: High socks!

Comment From Guest
How should one explain to a WAR skeptic that Ben Zobrist has been one of the most valueable players in baseball the past few years?

Neil Weinberg: Ben Zobrist plays good defense at many positions. Ben Zobrist is a good hitter. Ben Zobrist runs the bases well and never gets hurt.

Neil Weinberg: Here’s my lesson of the day!

Neil Weinberg: When you get information that contradicts your viewpoint, your first reaction should be to question yourself and not the data. Then question the data.

Comment From Jared
Do you think the defensive component of WAR is too strong? Some projections of Andrelton Simmons and Xander Bogaerts have them both as potentially valuable as Miguel Cabrera. If I’m starting a team, I’ll take Cabrera’s bat anyday.

Neil Weinberg: Preventing runs can be extremely valuable. But the best hitter can offer you 70 runs of value while the best fielder can max out around 30-40. I think that’s fair.

Comment From B. Lane
Why is yours called a Q&A rather than a chat?

Neil Weinberg: I think to let people know that it’s designed for inquiries about the site? It was Dave’s recommendation. It’s a chat! Or a Q&A!

Comment From Andy
You just said that YZR is updated weekly. I had kind of thought this, because I noticed that at certain points,some players’ defensive WAR changes drastically overnight, sometimes when they didn’t even play. Can you say a little more about how updating can change a player’s defensive runs so dramatically?

Neil Weinberg: DEF = UZR + Positional adjustment. We update the positional adjustment daily.

Comment From scatterbrian
How much weight is given to Off and Def (and BsR?)when deriving WAR?

Neil Weinberg: (OFF + DEF + REPLACEMENT + LEAGUE ADJUSTMENT)/runs per win = WAR….BsR +Batting Runs equals OFF, FYI.

Comment From Catoblepas
Is there any way to access heat map data in a spreadsheet? It’s hard to do large-n analysis when it all comes in pictures.

Neil Weinberg: Your best bet is baseballsavant.com

Neil Weinberg: There are a lot of good questions left in the queue. If I didn’t answer, please send them again next week. Usually I will chat for longer!

Neil Weinberg: I’m @NeilWeinberg44 on Twitter. Feel free to ask me questions there whenever. And look for updates to the glossary coming Friday, I think!

Neil Weinberg: Thanks for hanging out. Hopefully this was productive!