Quite the entertaining morning yesterday watching the online reaction to some of what Mariners manager Eric Wedge had to say about his veteran players. Many of you here were upset by it, as were some on my Twitter feed and in various other blog comments sections. Some of the more vehement comments centered around his daring to use statistics like “home runs” and “RBI” in discussing player merit, rather than more advanced metrics.
I saw the usual “Fire Eric Wedge!” hysterics, but that’s to be expected. Fans tend to overreact to daily happenings with every baseball team in any given year and that’s nothing new.
...I’ll leave you with this one thought, which occured when I read this comment over the internet yesterday:
“Today is Eric Wedge’s 1316th game as a manager of an MLB team. If you take him at his word, he apparently still believes that it is a worthwhile effort to (1) place a fair amount of importance on Olivo’s veteran status, and (2) reference RBI numbers as a measure of how effective a player is.
He’s had 1316 chances for the light bulb to come on and realize why that’s wrong. If I did my job wrong for 1316 days, I would be wondering why I was still employed. I don’t think it’s a stretch to wonder if Wedge is really cut out for this sort of thing.”
My thought after reading that comment is, is this really the level of arrogance our increased knowledge of stats has brought us to? For me, out of simple humility, the thought process should be: “Wedge has had 1,316 chances for the light bulb to come on and realize he’s wrong. Maybe, the fact that Wedge hasn’t realized he’s wrong is an indication that my thought process might not be as bang-on correct as I think it is. Maybe it’s me who has to re-evaluate. Maybe there is more to the job than I realize and that’s why Wedge has been employed at it for 1,316 games.’‘
But that’s just me. Something to think about.
Reader Comments and Retorts
Go to end of page
Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.
1. Rickey Fredonia Fudge Duckery Precious Twiddle Posted: April 27, 2012 at 01:44 PM (#4117724)Stats don't matter, winning matters. Stats tell you much more than Eric Wedge is apparently willing to admit about who is and who isn't helping the team win games. Olivo was a big offensive liability last year even while hitting 19 HRs, which may have made the yougsters feel less need to do more than they were capable of, but why? Wouldn't it be the opposite.
Is it really true that Ichiro and Ryan aren't providing veteran presence to the club? Is it really true that Olivo would lose all of his gritty usefulness sitting on the bench half the time? If so, then maybe you do need an offensive sinkhole and defensive liability at catcher (and DH) 90% of the time.
Wedge holds a clubhouse together and provides other important managerial intangibles, but he really isn't good at getting his reserves and his 10th,11th, and 12th pitchers enough reps to develop or help a team win.
I don't know, it's kind of strawmanish for my tastes. For example, he repeatedly cites ERA as a stat that stat people(whatever that means) hate but that's not really true at all is it? ERA isn't perfect, but, for starters anyway, you will be right much more often than wrong by relying on straight ERA, especially if you make even a cursory, gut-level "Safeco is a pitcher's park so that 3.50 ERA isn't quite as good as it looks" adjustment. Same things with HRs. Are there any prominent stat voices that object to idea that, in general more HRs are better than less HRs or object to managers saying they wish their teams were hitting more HRs? Even his cherry-picked irrational internet commenter doesn't say anything about HRs
But perhaps my favorite claim, and one that is common in these kind of articles, is that it would be silly to expect a player to care about a "complicated" stat like FIP. (13*HR + 3*BB - 2*K)/IP + C? What a crazy equation! Players prefer simple stats like Batting Average with its straight-forward H/(PA-(BB+S+SF+HBP)) calculation.
Is there a corresponding shortcut among 13*HR + 3*BB - 2*K)/IP + C as there is with PA-BB+S+SF+HBP=AB?
Batting average really isn't a complicated stat, or concept. Trying to make it one by dragging out its components doesn't change that.
AB = PA -(BB+S+SF+HPB) is convoulted. That someone came up with that convoluted equation a long time ago doesn't change that. If I called (13*HR + 3*BB - 2*K) "Pitching Outcome" and could I then say that FIP was simple because it's just PO/IP + C?
As for concept, the concept behind FIP is very similar to the concept behind ERA: fielding can have an effect on pitchers performance so we should try to correct for that.
*EDIT* On review, comment was too combative in tone. Changed some language.
What does "C" stand for?
the speed of light
There is absolutely no reason why you couldn't substitute e.g. 13*HR + 3*BB - 2*K with another constant, call it TO for true outcomes, and then it becomes TO/IP + C. Seriously, try explaining what an AB is to somebody who has never watched baseball before, and you will change your tune really fast. Just because people who know baseball have been dealing with AB's since forever, and have grown used to it, doesn't make it magically less complicated. "Home runs, walks and strike outs" really isn't substantially more difficult to grasp than BA is.
Edit: coke to 13.
And 'C' is a constant they added to make FIP scale to ERA. Ironically, it is there to make it less complicated for people who are unfamiliar with FIP.
a) Of course the fact that ABs have been around forever changes the "fact" that they're "convoluted." For crying out loud, it's part of baseball lexicon -- "he had good at-bats today" is commonly understood (even if often technically inaccurate) while I don't know that (until recently) I had ever heard a player or manager use the terms "plate appearance."
b) "Pitching outcome" would be fine as long as you get it adopted by mlb, b-r, Topps and everybody else who publishes baseball statistics. The reason your detailing of the components of AB is a strawman is because nobody has to do that calculation because every collection of baseball statistics will list the AB. The debate over whether "AB" is a sensible or useful concept was settled 100+ years ago and the answer (correct or not) is "yes".
c) The point you want to make is not about how un-complicated the FIP formula is or how complicated the BA formula is but that Baker is wrong because of course the player doesn't need to know the formula for FIP, he can easily find out what his FIP is on the web. And the reason the player should care about FIP more than ERA is that, near as we can tell, FIP is a slightly better measure of his performance. (Leaving aside whether the player should care about either one.)
I think you're missing something essential in the comparison of these stats. Batting average is a simple record of something that actually happened. Yes, it gets a little complex in that we subtract things like sacrifice flies, but it is a straightforward record of what happened on the field: the percentage of times that a player got a safe base hit in his recognized number of official at-bats. It may be useful, it may be useless, but it's just record-keeping. Eli Whiteside got hits in .197 of his official at-bats last year. That's a fact.
FIP, on the other hand, is analysis. To say that Kyle Farnsworth had an FIP of 2.45 (or whatever) last year doesn't have a whole lot to do with reality; it's somebody's estimate of how he would have done in some alternate universe. It's not a record of anything that happened on the field. It's an analysis of what happened on the field.
I think it would serve sabermetricians well to keep that distinction in mind. Facts are objective; analysis is not.
It isn't convoluted or complicated, if you give it the least amount of thought. PA = times coming to the plate. AB represents times coming to the plate where the batter has the intent and opportunity to hit the ball. BA represents the times a player successfully hits safely against the times he has the intent and opportunity to hit safely. It's really simple. Nothing in your FIP example has such an easy explanation.*
You don't count BBs or HBPs (or CIs, for that matter) in BA because the player's opportunity was denied. You don't count Sac Hits, because he wasn't trying.
There's one flaw, as far as I'm concerned. SF should be counted as ABs, since there's no reason to think players aren't trying to hit safely. That's it.
* This isn't an argument to support BA or denigrate FIP. Only that BA is in fact a relatively simple concept - safe hits: attempts to hit safely.
Huh, I never knew that HR, BB, SO, and IP weren't events that happened on the field, but in fact an analysis.
EDIT: There is a shitload more subjectivity in sacrifices and hits vs ROE than in anything that goes into the calculation of FIP. Those require a much more complicated explanation than anything in FIP.
When does 13 times home runs happen on the field?
Oh Noes!!!11!!1!! How will you ever explain that a HR and a walk are not the same value. That's far too complicated for someone who plays baseball for a living to understand.
in a Bugs Bunny cartoon
Every time Eli Whiteside gets .197 of a hit.
When the Red Sox relievers enter the game.
Or maybe we should stop counting sacrifice grounders.
Except for SF, as you mention. Also except for errors, where the batter tried to reach safely and did reach safely but is treated as though he did not reach safely. Also except for fielders choice, where the batter tried to reach safely, and did reach safely but not the right kind of safely because his reaching safely led someone else to make an out. Also except for when the batter gets out trying to stretch a single into a double, where the batter tried to reach safely, did not reach safely, but was sort of safe for a second and would have been safe if he was a better guesser about the relative quality of his footspeed and the right fielder's throwing arm, so let's go ahead and count it.
Well, that's a non-answer.
Whether deliberately or not, you seem to be missing my point. I'm not saying FIP is too complicated for anyone to understand. I'm not saying FIP is worthless. I'm saying FIP is not a record of what has happened on the field; it's someone's analysis of what has happened on the field.
And in that position, you're wrong.
**this is all independent of the UMPIRE'S subjective judgment, which affects both stats to some extent.
i don't know what coke means but i think a coke for 32 who said what he said while i was composing what i said.
This is a joke, right? Once the starting pitcher gets out of the decision, the rules get pretty convoluted. Granted, probably 95% of the time it's pretty obvious, but it gets kind of complicated in some situations.
I'm only arguing against two things, one that BA is a "record of what happened on the field" while FIP is not and that FIP is a complicated stat.
I would say equating the subjectivity between BA and FIP is not really accurate.
I also don't think the things in #28 are equal to SF at all; that whole post is a bit shrill in its pedantry.
Not exactly, the only analysis is about the relative values of certain events. It's no more of an analysis than slugging percentage. It's also selective about what it's tracking, but again slugging percentage doesn't count walks, HBP, etc.
A very long while. I think ERA+ is much closer. FIP, while growing in acceptance, still ain't thurr yet.
If you're talking about me, I would really much rather prefer not to be lone.
I'm an arse for taking unnamed personal shots, so I'll cop to that. Wish I didn't post it, named or not!
What I meant was the sense of "oh these stat nerds and their FIP, not understanding how the real men of baseball operate" that's coming from the article and a few posts here. I think posters here should be above that.
Spoken like a man who's never read a Dave Cameron piece before.
But the tools that are useful for GMs and their player evaluation teams in the off-season are virtually worthless for a baseball player during an individual season. FIP isn't going to help a reliever place his 2-seamer better. oWAR totals from last year aren't going to help a hitter eliminate the hitch that crept into his swing and blew up his mechanics when he faced that knuckleballer six games back.
When selecting a man (person, but in reality, a man) to be the on-field manager of the team 162 games per year, you have a decision to make. Do you hire someone who speaks (and thinks) in front office/stat speak? Or do you hire a guy that speaks/thinks in player speak. For virtually every case, that decision is a non-controversy, at least within baseball. You hire the guy that can communicate with the players. Stat savvy is a secondary concern. Yes, you'd prefer to have a guy who can process statistical analysis/thinking and then switch to player speak when he goes to the clubhouse, but if you can't find that guy, you take the guy who can communicate and motivate players instead.
Statistical thinkers (nerds in their basements) like to think that they could do a better job managing than Eric Wedge, or Dusty Baker, or Fredi Gonzalez, because, after all, they understand *probability charts* and surf Fangraphs for FIP! That is hubris and ignorance on their part. Dave Cameron (to use the whipping boy for this tendency) would fail so utterly and completely as a ML manager that your eyes would pop out. His players wouldn't know a thing about what he was trying to say, and they certainly wouldn't play hard for a guy that tells them how stupid they are for not understanding that C is just a constant to normalize to antiquated BA presentation standards but couldn't identify a curve out of the release point if his life depended on it.
Hopefully never.
I'm statistically-inclined, but count me in the camp that thinks there is a meaningful distinction between stats that account for things that actually happened versus ones that don't.
Obviously, FIP is determined by actual events -- that's not the problem. The problem is that FIP is a predictive stat. It's derived from the play on the field. ERA is of course subjective in what is earned or not earned, etc., but it's a much more direct relationship. I want statistics to look backward, not forward.
SLG and OBP, again, describe what happened, and do so in a way that's more revealing than batting average. But I don't want to see xBABIP on the scoreboard.
Ok, but how about OPS+ or something else that's at least adjusted for park and league?
EDIT:
I think you are mostly right in #50, Sam, but I don't totally agree. I think some of the advanced stats could be useful to managers and players. FIP is probably not one of them though.
#### you bbtf
I just wrote a long, detailed comment on the article. 13,000+ characters long in fact. I was not happy to find that we apparently have a 12,000 character limit and, after telling me my post was longer than the 12,000 character limit, BBTF was nice enough to erase the 13,000 plus characters I had written.
Your gain, my ego's loss.
A-yup.
HA-HA!
I love OPS+. For old timey baseball statistics, and for baseball purposes, I like OPS, OPS+, triple slash, ERA+, WHIP, and the like. They're useful at describing what has happened. For fantasy baseball, and projection, and analysis, I like FIP, and BABIP, and all of that silliness because its inherently predictive.
Well, kinda. Let's say the point of the article isn't at all clear (my 13,000+ characters in a nutshell). Baker is responding to stathead fan reaction to a Wedge interview with the media. He responds to this by asking "why should Wedge talk in fancy stats his players won't understand and can't use?" Even if we assume that's true (which I'm not sure it is), what does that have to do with what terms Wedge uses when speaking with the media (and indirectly the fans).
There is also some completely unsubstantiated stuff about how baseball managers know how to manage human beings. What that has to do with what objective measures you use to assess performance of those human beings is never addressed. How Chone Figgins' "heart and brains" justify playing him is never addressed. And Eric Wedge's ability to manage human beings is never addressed nor is how Z's assessment of "makeup" has helped to produce a better baseball team (given the Ms are a crappy baseball team, this might be hard). How Wedge's statement that Olivo needs to start hitting because it's a team without an HR and RBI guy might affect Montero (2 BB in 72 PA, sub-300 OBP) and Smoak (5 BB in 71 PA, sub-300 OBP) is not addressed.
Heck, the article starts with a claim that Wedge won't be fired when, in fact, his chances of surviving the season aren't good and his chances of being back next year are very low.
It makes a reasonable point about the utility of video vs. stats (although there are issues here too) but mainly its thesis is "Eric Wedge wouldn't be managing in the major leagues if Eric Wedge wasn't a good major league manager."
Assuming that what you say in this post is what the author is trying to say, your post explains it far better in far fewer words. Given the way article meanders, I am not entirely convinced that it was his point, but I will give it the benefit of the doubt.
How is this any different from pro-grission callers to WFAN saying they could do a better GM job than Alderson, Beane, Daniels, and Epstein?
It's not. I would simply suggest that not being better than callers to WFAN is a pretty low standard.
It's not just that it fails to disprove the assertion that players have no use for advanced statistics, it also fails to address the flip side, as to why RBI is useful to a player. If "FIP isn't going to help a reliever place his 2-seamer better", how does RBI help? Do hitters ever go to the plate thinking 'Driving in the runners here would be bad'? RBI is beyond useless for players, managers, and front office staff.
On the flip side, I actually do think FIP could provide a pitcher with valuable information (albeit less than the individual component stats). Knowing if you are/aren't walking too many hitters or striking out enough, is way more helpful than RBI. And FIP could also be used to gauge whether a pitcher's current approach is good despite not getting results, and he should stick with it, or if he needs to make an adjustment. Why would pitcher's not want to know these things?
Let's try this then: Just because some "statistical thinkers" here or at Fangraphs might tear their hair out watching Wedge, Baker, or Gonzalez, it does not follow that these individuals think they can do a better job. More likely they simply wish someone like Maddon was managing their team.
Yes, exactly. Why bother with FIP when it's the component stats that you're really interested in? What's the point of mashing all the components together to get to a number that reveals much less information than the components themselves?
I don't think the point is that RBI is useful to a player, per se. I think that it's so ingrained in the parlance of the game, from Little League up, that it's not *distracting* or *confusing* to a player. And I think a communication method that doesn't confuse the player - who is really good at muscle twitch reaction and hitting a 95 MPH fastball, and then running 90 feet in mere seconds, but probably didn't do very well in Algebra 101 - isn't something to be undervalued.
Players understand batting average because it counts something positive they did on the field. Modern players are embracing OBP more and more, in direct proportion to the change in baseball culture that views a walk as something positive a hitter did, rather than something negative a pitcher did. Players understand RBI because they view driving in the runner as a positive act that they did.
Players and "baseball men" have little use for FIP or fRAR or whatever because it has no direct relation to something act that they did, with a result that they can clearly see with their own two eyes (like the ball flying off of the bat, or a runner from second touching home when they single him in.)
Oh yes. Joe Maddon. The great white whale of baseball managers, in Ahab-statgeek's world.
It occurs to me that someone like Maddon and Davey Johnson, who can code-switch* on demand between stat and player-speak, is a bit of a rare bird.
*look it up
I didn't read the article as addressing what is or is not 'useful to a player.' The article is about a manager speaking to the press, I think.
What point are you trying to make here?
I'd agree with your point, but the very fact that Wedge thinks that Olivo can supply what he wants would seem to indicate that he's not looking for a good player, he's looking for a player that hits HRs and drives in RBIs. Olivo's always been the kind of player who could look real good based on HR and RBIs and still be a bad player, which seems to be what Wedge is implying would help his team, so I have trouble giving him the benefit of the doubt there.
Really?
Para 5: While it's true that virtually all teams now employ some type of advanced statistical consultants, the idea that players and coaches are sitting around the clubhouse conversing about xFIP, WAR, wOBA and any stats that don't revolve around "Batting average" and "RBI" and "ERA" simply is not the case.
Para 6: I know we all want to think that is what's happening. And some in the mainstream media -- even national-level favorites -- help encourage this type of thinking with stories on the one pitcher out of hundreds who knows the difference between FIP and ERA+.
Para 7: And the reality is, most players still don't have a clue about advanced stats
Para 8: But start talking situational OPS with just about any player and you will get a scrunched up face staring back at you
Para 9: And so, the man placed in charge of these players in the field-level trenches -- the manager -- is not going to have advanced stats as part of his daily lexicon of words when he speaks
Much later: one reason managers don't sit around filling the heads of 25-year-old athletes with an encyclopedia of baseball stats is that they realize their human athletic brains can't handle it.
Then: Baseball is a game of instinct and reaction, where you react to events happening in front of you at 100 mph. ... There is no time to be delving through stat pages in your brain when this stuff is happening
And: So, anyone expecting a manager to start speaking about OPS and wOBA in a conversation about baseball is missing the point.
But those people aren't missing the point. The question is how does Eric Wedge assess the _performance_ of his players. How does he decide what aspects of his players' games to work to improve? What does he value, HR and RBI or OBP? This is my point about Smoak and Montero. Wedge has just said that there is no HR and RBI guy in the lineup -- what are Montero and Smoak for then? Does this hurt their confidence, confuse them about their role on the team (i.e. the human side)? What does it say about Montero's future as a C that Wedge needs Olivo to step up (and keeps Jaso on the roster)? Further Wedge didn't express concern with Olivo's lousy OBP -- how are Smoak and Montero interpreting that? Does Montero have 2 BB in 72 PA and Smoak 5 in 71 (after 55 in less than 500 PA last year) because they have been told or think that their manager cares more about HR and RBI than OBP?
To me here's about all that Baker should have said: Anybody who doesn't realize that any manager interview is a PR exercise is going to be a frustrated baseball fan. Wedge may have spoken in HR and RBI terms because he doesn't know/like advanced stats or he might have spoken in those terms because he thinks they're the ones the majority of the fans will relate to. But what he says to the press tells you little about what he's really thinking. Why do you read me then? I don't know.
Instead Baker writes crap like this:
The guys in charge of players -- who have to lead them into on-field combat
and...
And those human managers all know about baseball. More importantly, they know about managing in baseball and the human subtleties that come with the job.
Really, some of them seem to suck quite badly at the "human subtleties."
Wedge knows what types of hitters he needs and where.
Are we sure about that? Isn't that the crux of the question? Why take that on faith? Why not, god forbid, ask Wedge whether he's concerned about Olivo's OBP? Why not ask him why it is important to have a HR and RBI guy or a guy who can slug 450 that's not all doubles (I don't even understand that one to be honest).
Here's what I know -- as a team, the Ms are walking once per 15 PA. That is awful. There is no way to win baseball games doing that. The young players, the only thing Wedge should be focused on with this crappy team, are leading the charge: in addition to Smoak and Montero, Seager has 1 BB in 61 PA, Ackely 5 in 85, Liddi 2 in 30 (Saunders is doing OK with 7 in 55). If Wedge knows the kind of hitters he needs (and where, whatever that means) then why isn't he developing them? Smoak did better than 1 in 10 last year, so did Ackley, so did Montero with the Yanks but was about 1 for 12 in the minors.
Maybe Baker, instead of assuming Wedge knows what he's doing, should ask him why the young hitters aren't walking and what Wedge is trying to do about that. Maybe Wedge does need to talk to them about OBP, about the importance of not making outs, that it will help the team win more games. He doesn't have to pull out WAR or wOBA or whatever to do that but why are we assuming he knows that's what's important?
Well, the point is about advanced statistics in general, and not FIP per se. And I think the notion that there is nothing valuable for players to be gleamed from anything beyond triple crown stats is beyond ridiculous.
This is basically what I disagree with. Not that a manager should be lecturing players about the advanced stats themselves, but it IS his job to understand the lessons that can be learned from them, and relay them to the players in a way understand.
You are walking to many batters.
You are hitting to many grounders.
You need to swing at fewer first pitches.
You need to throw more breaking balls.
You don't need to rely on advanced statistics to communicate the lessons to the players. But you need to be able to read them, or you won't be able to implement the valuable things they can tell you.
Really? I don't particularly give a crap what sort of stats players talk about in the clubhouse.
Sure, for putting together a team, addressing weaknesses, etc. But I'm not sure this holds for a field manager trying to motivate players or help them improve. The raw data is not only adequate, but likely superior, for that. A manger/pitching coach needs to know that one of his starting pitchers is killing himself by walking a batter per inning, and needs to be able to communicate the importance of fixing that (maybe by showing the pitcher how many of those walks wind up scoring, or how high his pitch counts are after just four or five innings). A manager does not need to know what xFIP is or to give a crap about it in order to understand the importance of the component stats and make use of them.
Accountants didn't just create all these metrics in the business world simply for job safety.
But I don't know that the "big picture" is a useful way to help a player improve his performance. I think you will generally get better results by trying to fix one little thing at a time.
No, I'm simply relying on catcher X to intuitively understand that there is value in getting more strikes called. He's been playing baseball for a very long time. He knows the importance of ball and strike calls without having to see them converted into run values. I will positively reinforce this when he comes into the dugout at the end of an inning by asking how much he enjoyed getting that punch out to end a threat and preserve a lead. I think that will make the point much more effectively than trying to explain how he just lowered pitcher Y's xFIP by 0.07.
Look for Craig Wright's study on the subject. He attributes the drop-off in run scoring to taking too many first pitches.
At one point in time you could have won a great deal of money betting me who took more first pitches, Ichiro or Mark McGwire in his prime. (Ichiro took a lot more first pitches)
I mean, this
To me here's about all that Baker should have said: Anybody who doesn't realize that any manager interview is a PR exercise is going to be a frustrated baseball fan. Wedge may have spoken in HR and RBI terms because he doesn't know/like advanced stats or he might have spoken in those terms because he thinks they're the ones the majority of the fans will relate to. But what he says to the press tells you little about what he's really thinking. Why do you read me then? I don't know.
Is absolutely something we can all agree with, I think. I guess I don't get where you're coming from here ultimately because you seem pretty frustrated.
Huh? What "link to utility" are you looking for? You said the article was about how managers speak to the media. You read it very differently than I do since the first half of the article is talking about how players don't understand advanced stats then it spends another chunk talking about how video is more useful to players than stats. I'm not sure there's a single spot in there where Baker writes about how managers should talk to media/fans. So I was curious how you came to your conclusion.
You must be Registered and Logged In to post comments.
<< Back to main