Neil Weinberg FanGraphs Q&A – 10/1/14

Neil Weinberg: Hey everyone. We’ll get started at 3pm, but the queue is open. Remember, I focus on questions about our data/features/sabermetrics/etc, but regular questions are always welcome.

I’m @NeilWeinberg44 on Twitter if you’re looking for me and I publish updates to our Library on Fridays.

Hope you’re recovering nicely from last night!

Neil Weinberg: Afternoon, everyone who isn’t west of the pacific time zone. Morning to those of you who are.

Neil Weinberg: Let’s see about some baseball. I probably won’t run the usual two hours because there will be like 8000 hours of chats this week, but if you ask good questions I will go until my dog starts bugging me to play.

Comment From 1990JaromirJagr
Woof. the hardest part of my day was choosing my player of the year ballot. thank god the choice was easy with my pick of Mike Morse MVP!

Neil Weinberg: I don’t have a vote, which I’m a little relieved about. Picking the down ballot spaces is very hard. Which I guess is why we have the grading part. Lucroy or Cutch? How would you separate them?

Comment From Guest
it seems like this fall is the ideal time to trade Jose Bautista. He had a .404 OPS, 35 HR, 100+ RBIs ect, but he will be 34 shortly and will have a no-trade right as a 10 year player. What is a reasonable to Toronto, assuming they decide to sell high, ignoring the hole his trade would leave.

Neil Weinberg: I think they should have traded him last offseason, but no one wanted to hear it and they clearly wanted to make a run. I think you can get a really good prospect and a solid one.

Neil Weinberg: Two years for $28 million. Maybe 8 WAR as a nice safe estimate. That’s crazy valuable.

Comment From Santiago
I want to calculate linear weights. How can I do it?

Neil Weinberg: This is too long to explain in a chat, but I think this page does a decent job! Requires lots of computing skills to do it right. http://blog.philbirnbaum.co…

Comment From AndrewY
I feel like KC’s win was a particularly cruel way to beat Oakland. Steals, bunts, average, low power and walks. Billy Beane’s worst nightmare has come true.

Neil Weinberg: I was sort of cheering for the A’s to win with a bunt. Always fun to watch people who didn’t read Moneyball talk about Moneyball. One of my favorite moments of TV was Mayo schooling Reynolds about it at the draft like two years ago

Comment From AndrewY
If you had to have Adam LaRoche or Lucas Duda on your team next year, which do you choose?

Neil Weinberg: Duda?

Comment From Pennsy
FG WAR has Clayton Kershaw as the most valuable player in all of baseball this year. Bill James’ all-encompassing Total Runs formula has Anthony Rendon as the most valuable player in all of baseball this year- but more curiously, Kershaw isn’t even on the top-21 list available to non-subscribers of his site, while Kluber, King Felix and even Johnny Cueto all make an appearance. Why this huge discrepancy between FG and Bill James?

Neil Weinberg: I don’t know what Bill does, but I would assume innings play a big role there. Must have a lower replacement level? Not really sure but that’s the only explanation I can see.

Comment From ralph
Dave talked about differences between Kershaw and Kimbrel in his chat. This bit caught my eye in particular: “Looks like Kershaw’s self-created run environment gave him a 7.9 runs to wins conversion, while Kimbrel’s was 9.0.” Could you expand on that? It seems like a shockingly large difference.

Neil Weinberg: This is a little hard to expand on in a chat, but the basic idea is that Kershaw pitches far more innings per appearance so more of Kimbrel’s R/W is based on the league average than Kershaws. I hope to get the pitcher WAR entry updated next week, but it’s a little trickier to put together because it’s more complicated that position players.

Comment From NotThatGuy
Since WPA is context neutral, RE24 is not and yet they are both on the same scale, can we turn the difference into an average “context score” for the player?

Neil Weinberg: Both are context dependent and they are not on the same scale. So I’m not sure what your question is asking. Please clarify!

Comment From Gila Monster
In voting for Fangraphs player of the year, I ran into a problem. I believe Jonathan Lucroy has a strong case. He put up a 6.3fWAR and also saved 23.7 runs per BP’s framing metrics. This brings him up to a 8.7 WAR or so. Even splitting the difference put him with Trout. The problem is the the “historical context” part. His 2014 season is the 30th best season of all time for a catcher, but it could be much better, but the lack of historical framing data is a problem. Is this a problem with integrating framing into WAR? He could be having a top 10 catching season, but we seemingly can’t tell.

Neil Weinberg: Yeah. We just don’t know. Basically, all framing runs have gone to the pitcher in the past by default. However, I would assume that you can take the average spread in framing runs since 2008, so like -20 to +20 or something and just mentally adjust past catchers. Do I know if Bench was +20? No idea, but I can bet that someone was. Best guess I can offer

Comment From Jeff
When evaluating the value of a stolen base, would you be more inclined to use RE24 or WPA?

Neil Weinberg: Both? RE24 tells you if it’s useful in that inning, but the score and inning (i.e. stuff WPA knows) tells you how much scoring extra runs matters at that moment.

Would you rather steal second with 0 outs in a 4-0 games in the 4th or steal second with two outs in a 0-0 game in the 12th? Really depends on what you care about. Both are useful!

Comment From Bomok
Brett Gardener and a middle relief pitcher for Daniel Murphy, Jon Niese, and Eric Young Jr.
Who says no? and why do they say no?

Neil Weinberg: Yankees, I think. Not sure there is enough coming back for them to want to remove one of their important offensive pieces.

Comment From Pennsy
Who finds success first: Billy Beane wins a World Series or Billy Bean coaxes an active player out of the closet?

Neil Weinberg: I don’t like the use of “coax,” but I bet we have an player who is out before an A’s title because one will happen i nthe next couple years and one has a small chance of happening before then, on average

Comment From NotThatGuy
How valuable are IP really? I feel like Shields gets overrated by WAR largely due to the bulk innings. Posts a league average HR/FB in a homer-suppressing park.

Neil Weinberg: This is fun, because SO many people think innings are undervalued by WAR.

Neil Weinberg: I think there is value in guys who are durable and avoid blowup starts and such, but I don’t think it’s a big jump. I think there is too much pundit talk about guys who pitch deep into games. Nothing against those starters, but 6.2 innings and 7.0 innings on average isn’t a big deal

Comment From Jim
Is Fangraphs considering going from FIP to SIERRA as the main ingredient for pitcher WAR?

Neil Weinberg: No. I mean, unless there are secret SIERA meetings no one invites me to.

Comment From Guest
The problem with the grading system is that it is terrible. I’m not sure any of the players graded above a 40 on the scale Dave set and I’m also not sure I saw enough of a difference in say Kershaw and McCutchen or Bautista and Stanton to separate them on that scale.

Neil Weinberg: This is an argument for why the scale is good. You’re saying the players are identical and if you have to rank them, you’re saying they aren’t. If you can give them both 40s, then you can say they are equals

Comment From Jacob
Alex Cobb current form or Drew Smyly with a cutter and increased change-up usage?

Neil Weinberg: Want to see if the Rays teach Smyly the magic changeup. He needs that pitch. Cobb for now.

Comment From A. Lane
Has anyone looked at how starting pitcher performance decay with turns through the order has change over time?

Neil Weinberg: I don’t know! That seems like 1) something MGL or someone else would have done or 2) something a good aspiring analyst could do pretty easily! I haven’t read a temporal study of it though

Comment From Mike, of Mike and the Mechanics Fame
If a second baseman was league average (not position-adjusted) in every facet of the game–i.e., 100 wRC+, 0.0 UZR, 0 BsR–would that player’s WAR simply be replacement level + positional adjustment?

Neil Weinberg: Basically, yes.

Comment From Pennsy
Roberto Clemente has almost inarguably had a larger impact on Major League Baseball than any other player since Jackie Robinson. First Latin American superstar, often spoke out against anti-Latino and anti-immigrant prejudice, and essentially pioneered the concept of the Ciudad Deportivo which became the instructional academies responsible for many of the game’s biggest names today. Major League Baseball should retire his 21 across the League, right?

Neil Weinberg: I would have no problem with more love for Clemente. Although you have to wonder about the practicality of retire numbers league-wide for every big impact guy.

Comment From Guest
one of the best things about postseason baseball, besides the actual games, is getting to hear Pedro talk about baseball.

Neil Weinberg: I went to bed after the game ended, but people were tweeting his stuff. Pedro is the best.

Comment From Just Sayin
I don’t think Harold Reynolds got Mayo’s message.

Neil Weinberg: He didn’t.

Comment From NotThatGuy
That should have said “Since wRAA (not WPA) is context neutral, RE24 is not and yet they are both on the same scale, can we turn the difference into some piece of useful information?”

Neil Weinberg: Ah! Much better. So…first, wRAA isn’t park adjusted and it doesn’t included SB. RE24 does both. So you want Batting Runs + wSB and RE24. Those are the same scale.

And yes! You absolutely can. I look at that all the time. I think Dave wrote a piece about this too, dealing with Hunter Pence. I did it at BTBS with David Murphy a few months back. It basically tells you how well a player has timed his performance. Which is useful to know for many reasons.

Comment From STiVo
How many Announcers Below Replacement is Ron Darling? Sheesh. (“Josh Donaldson has the best bat speed in the league, pound-for-pound.”)

Neil Weinberg: Funny, the Mets fans I know say he’s usually good for them. I have strong opinions on national broadcast announcers. Usually slam them pretty hard on Twitter, if you like that kind of thing

Comment From Guest
Carlos Martinez going forward is an RP or SP?

Neil Weinberg: Tough one. I’d go for it as a SP, but I would bet he ends up in the pen.

Comment From Pennsy
Given called strikes are rising in today’s game isn’t the value of framing likely higher now than it was in the past? If framers weren’t necessarily ‘worse’ in the past, their skill might have had less relevance in a league environment where fewer strikes were called overall.

Neil Weinberg: I don’t think we know enough about pre 2000s strike zones to know. There’s a correlation causation question there. You might be right, but I don’t know.

Comment From NotThatGuy
Who do you see batting second for NYY next year? Gardner? Not have a floundering Jeter locked in there should really help them (~1 win maybe).

Neil Weinberg: I want it to be ARod, haha. Beltran if he’s healthy seems like what will happen.

Comment From JBD
Whenever I try to explain WAR to a noob, I always get tripped up on the runs-to-wins conversion. I know it comes from Pythagoras… but could you put it in like two sentences the way I should explain it?

Neil Weinberg: R/W is the number of runs a team needs to score, on average, to add one win to their total in that year. So if you score 680 and allow 671, you’re probably a game over .500 on average. Basically.

Comment From Santiago
Are you going to be in one of the Game Chats?

Neil Weinberg: I might! We’ll see how things go. I think we’re scheduled through Friday and will then reevaluate how many we’re going to do each day. If there’s demand and an open spot I’ll jump it.

Comment From Guest
you planning on attending any playoff games?

Neil Weinberg: No. My family had season tickets to the Tigers before I moved in 2012. So I went in 2006/2011, but haven’t been since. Could theoretically drive to BAL or WSH but the overall cost is pretty high. If the Tigers play in the WS I might.

Comment From Bill
Aren’t the WAR totals suppose to add up to 1000 wins across the MLB for the 2014 season?

Neil Weinberg: They do!

Comment From David
Jim has asked that FIP/SIERA question in every chat for the last week. Always gets the same answer too … maybe he’s secretly Matt Swartz!

Neil Weinberg: I love SIERA as an idea. Just don’t think it adds anything above xFIP with the data we have right now. Maybe statcast can change that?!

Comment From Mark
So RA9-WAR uses actual runs credited to the pitcher, correct?

Neil Weinberg: Yes.

Comment From Deluded?
Unnecessarily brash prediction: People will be talking about the Andrew Miller trade as the most important trade of the season after the Orioles win the World Series.

Neil Weinberg: If they win the WS, I bet that’s true! Isn’t baseball crazy?

Comment From Matt P
My understanding is that batters do worse even when the count is 0-2 then when the count is 3-0 when the ball is put into play (no strikeouts, walks, HBPs, etc). Is there any plans to incorporate that into FIP or a different pitching statistic?

Neil Weinberg: Interesting idea! So factoring in the average run value of a ball in play per count rather than total. I think someone could do this! Aspiring analysts, get on it! Not sure how much it would change things, but it’s worth looking at

Comment From Ringtone Composer
Anibal Sanchez to come out of the bullpen in the ALDS?

Neil Weinberg: He’s not in the rotation, so “yes.” But Ausmus is not the kind of manager who “uses his best relievers” so who knows.

Comment From NotThatGuy
What do you think it will take (especially years) for the Tigers to retain Scherzer and (given how the Verlander contract has worked out to date) they will be scared off?

Neil Weinberg: I think Scherzer gets 6/180 to 7/200. So that, minus a few percent. Which they won’t do…I think. There’s no indication he’s staying

Comment From 1990JaromirJagr
thoughts on the game tonight? at first i was big on Mad Bum but now starting to think that the pirates offense is too good in a winner take all

Neil Weinberg: Honestly, it’s probably 51/49 Pirates just because of home field. They are so close in true talent that you’re really just guessing.

Comment From JBD
I was thinking about how my son is going to start playing baseball someday (fingers crossed!). And I was thinking, you know what, we should just remove batting average from the kids’ vocabulary, and use on-base percentage. It’s actually easier to understand. But then, oh god, what if we placed an emphasis on walks in the lowest level of little leagues? Ten hours later, game still going on…

Neil Weinberg: Haha. Right. The key in little league is you want to teach them to enjoy the game more than win at all costs. Same reason you have a five run limit in many innings. Don’t go crazy, but emphasize getting on base rather that “hits”

Comment From Guest
do national broadcasters do any research at all? I know *we* are not the intended audience, but it is pretty brutal sometimes. and I think Harold Reynolds is doing the WS this year, so there’s that.

Neil Weinberg: This is what bother’s me the most. When national guys just get basic information wrong. I don’t mind as much if they talk BA and RBI, but gosh, at least do the work to prepare.

Comment From Bill
How come team WAR doesn’t tie to the teams actual number of Wins over replacement level. It seems to me like that would make logical sense.

Neil Weinberg: Add a team’s WAR to about 48. That will do it. You’re never going to get exactly right, but on average, the correlation between team WAR + 48 and Team Wins is very high

Comment From ralph
True or false: To be as great as it was, last night’s game needed the A’s being up 7-3 with the game looking like it was over.

Neil Weinberg: False…but that’s a matter of taste. The actual play last night was lousy. It was exciting, but it wasn’t high quality play. I would have preferred a 2-2 games with great defense and such, but that’s just my preference.

Comment From Trouble with the Curve
Can you think of a pitch more pleasing than Clayton Kershaw’s curveball?

Neil Weinberg: Fister’s two-seamer? Preference thing, though.

Comment From ralph
What would you have set the odds at for the Royals winning with Alex Gordon having a lower WPA than any of the other Royals hitters? Because that’s what happened. (Though Nix’s one PA almost surpassed Gordon’s 6 in negative WPA.)

Neil Weinberg: Haha, that’s fun. Very low.

Comment From Mark
Follow-up: wouldn’t RE24 make more sense than R, to divvy up inherited runners more appropriately? Or am I missing something that you’re going to write about in the Pitcher WAR entry?

Neil Weinberg: You could absolutely make this argument. I wrote about this a little at BTBS two weeks ago regarding Cueto and Kershaw. For most pitchers, it doesn’t matter much, but for a few per season you’re dealing with +/- 0.40 runs per 9

Comment From STiVo
Mookie Betts over/under 2015 WAR: 3.5?

Neil Weinberg: Just under.

Comment From Guest
Do Red Sox sign Sandoval or Headley?

Neil Weinberg: Hmmmm, certainly possible.

Comment From Analyst/Therapist
If you were starting an expansion baseball team tomorrow and you could either have Chris Sale or Anthony Rendon, which would you pick?

Neil Weinberg: Rendon. Want to build around the position player.

Comment From Yost
In T-ball for 3 and 4 year olds, is it more important to work on getting a good jump when stealing a base or their bunting skills?

Neil Weinberg: I didn’t get this because I don’t read the names very much. But I smiled.

Comment From JBD
It seems that Mookie Betts is extremely quick to make adjustments? Is there a good way to quantify this?

Neil Weinberg: You could measure this, I’m just not sure how with our current data. But that seems to be true given basic observation

Comment From ralph
Is this a good way to think about how Kershaw gets to a relatively low runs/win number? A run saved is more valuable than a run scored, so as you approach a 0.00 ERA, the runs-to-win conversion starts valuing runs saved more and more?

Neil Weinberg: It’s really more about the pitcher having control over how much of the game they participate in and their influence over how many runs their team needs to score. Mike Trout can’t get 9 at bats just by dominating the way Kershaw can extend an outing by pitching well

Neil Weinberg: Still questions in the queue but fire off some more if you have them!

Comment From You Cannot Fight in Here, This Is the WAR Room
More like to make the hall of fame: Stanton or McCutchen?

Neil Weinberg: Cutch.

Comment From Pirates Hurdles
Are you surprised at how Yost is getting crushed for making the right move with Shields and Melvin is hardly being talked about for leaving Lester in far too long. Its like bizarro world.

Neil Weinberg: Surprised? No. But yeah, the degree of bashing is wrong

Comment From A Former NFL Fan Who Is Fed up with That League
What would happen if a bedrock tenant of sabermetrics changed over the long term, like, say 30 years. (e.g., momentum became a thing.) Would we catch this, or would it go unnoticed, because people aren’t constantly testing these kinds of assertions?

Neil Weinberg: This is true in all aspects of “science.” We do so little real replication. I often tell aspiring writers who ask for advice to do replication with new data.

Comment From JJB
Does Bob Melvin survive after the poor Aug & Sept and some brutal decisions last night?

Neil Weinberg: I think so.

Comment From Guest
watching the game last night gave me some hope that Showalter can maneuver my team to a victory over a better team managed by Ausmus…

Neil Weinberg: Well Showalter will maneuver better. Just a question of if he’s in a position to make the difference

Comment From Lucas
Is it at all possible that the tigers trade for Jason Heyward this off-season to help with their atrocious outfield defense?

Neil Weinberg: That would actually be amazing. I’m not sure they have a good offer though…

Comment From mtsw
Is it encouraging or discouraging that the public reaction to last night’s game was basically “LOL NED YOST” even though a lot of his bunting and base-stealing was actually good strategy given the game situation?

Neil Weinberg: I think it’s good that people’s default reaction is anti-bunt rather than pro-bunt, but we still have a lot of work to do in identifying the gray area

Comment From Mike, of Mike and the Mechanics Fame
I know (because I read it on the Internet!) that just because we’re in a low run environment doesn’t mean that small ball becomes more important. But I can’t seem to convince myself of this. Can you convince me?

Neil Weinberg: The value of a single run is higher, but that doesn’t mean small ball will produce more runs. If you play for one run at every chance, you won’t score more runs on average than if you never do. If you can use it only when it’s the right call, you will be in good shape

Neil Weinberg: Haha, my dog is trying to tell me to wrap this up. I’ll keep going for a bit but he’s not going to let this go much longer without sitting on my keyboard.

Comment From mtsw
If you’re Showalter, do you save Gausman for long relief duties or bring him in in high-leverage situations as the 2nd best RHP in the bullpen (assuming Gausman doesn’t get a start)?

Neil Weinberg: Leverage. I mean….don’t use him Buck….

Comment From Arts and Sciences
When is Jose Iglesias going to play again?

Neil Weinberg: On track for OD, but a long way to go

Comment From 1990JaromirJagr
if im trying to explain the rationale for 1000 total WAR to a non statisically inclined person, how would you word it?

Neil Weinberg: It’s based on replacement level. So we think replacement level is about 48-114 for a team. If everyone was replacement level that’s 1440 wins. And because we know there are 2430 wins, we need 1000 WAR (I obviously rounded!)

Comment From Guest
as someone who never watches him, just how bad is Castellanos at third? will be have to move off eventually?

Neil Weinberg: He *was* bad this year. But he has the raw tools to play there at a -5 type level with proper improvements. Can he do it? Tough to say. Reports are that he’s a hard worker, but if he doesn’t get better next year he’s moving off.

Neil Weinberg: Alright, sorry to only run 70 minutes today but I was gone all weekend and the dog really wants to play. Plus you have Jeff and Paul in four hours! If you want to chat or follow along during the game, I’m @NeilWeinberg44

Neil Weinberg: Try not to die watching baseball games.

Neil Weinberg: You wouldn’t think I would need to say that, but here we are. Take care.