Effectively Wild: A FanGraphs Baseball Podcast - Effectively Wild Episode 474: The Rest-of-Season Projections Test

Episode Date: June 19, 2014

Ben and Sam identify the players who’ve most overperformed or underperformed their PECOTAs and discuss what they’ll do next.

There is no language in our lungs to tell the world just how we feel. No, no, no, no, no bridge of thoughts. No mental inlay. No letting out just what you think. Good morning and welcome to episode 474 of Effectively Wild, the daily podcast from Baseball Perspectives, presented by the BaseballReference.com Play Index. I am Ben Lindberg, joined as always by Sam Miller. Hello. If anything sounds off today, if our rhythm or our chemistry is not what it usually is, it's because Sam called me on Skype instead of me calling Sam. This is unprecedented.

Totally threw off my game. It's actually not. It doesn't really change anything. Technologically, it's the same. Yeah, right. It's part of my routine. I call you. Messing with the success after 473 episodes.

All right. So I have a topic. Before we get to that topic, I have some barely relevant research that I think is interesting. So I want to lead off with that. with that. We had an exchange over email today with a podcast listener who began his email by addressing it to Sam and Ben. And he mentioned that usually he starts it with Ben and Sam, but this time he chose to be different. You said that your sense was that it was usually Sam and Ben. So I searched my email for emails sent from listeners to podcast at baseballperspectives.com. And I discovered that Ben and Sam leads Sam and Ben by a score of 72 to 48, which you said was fascinating. I don't know whether you

were serious or not, but I was fascinated. So I wanted to find out why this might be, why Ben and Sam might be the more popular choice. I'm assuming that it's not that I am the more popular person. There is likely an explanation here. So I put out a call on Twitter for linguists after spending an hour or so perusing linguistics papers about high front vowels and low back vowels and semantic order and all kinds of concepts that I was only vaguely familiar with. Turns out that our friend and podcast listener and Grantland contributor, Michael Ba Bauman is married to a linguist and she provided an answer her name is Kate I asked her if there was a linguistic explanation for why Ben and Sam might be the preferred construction to Sam and

Ben she gave me several reasons so the first reason she suggested that this might be is that we might be suggesting a certain order, that precedence matters. So if we have our names listed somewhere in print in the podcast description, that might subconsciously or consciously influence the way that people address us. And that could be it. The podcast descriptions usually say, Ben and Sam discuss whatever the topic of the day is. So possibly people are picking up on that. She also mentioned, of course, Ben comes first alphabetically. She says, which wouldn't be a huge conscious motivation,

but subconsciously could influence it since it is a list. She thinks that would be a small factor. She also mentioned that Ben ends in N, and people tend to shorten the articulation of and in situations like this into something like Ben and Sam. So having Ben first allows its final consonant to run together with and slash N for faster articulation. Sam and Ben takes a little longer, she thinks. She also says that the position of an item in a list can signify its importance or relevance to a conversation. So depending on what the person is talking about, the more relevant person might be listed first or second, depending on the emphasis and context.

Lastly, she mentions that there is a subfield of linguistics called pragmatics, where people look at how context contributes to meaning. For example, the ability of the two guys in Dude, Where's My Car to have an entire conversation with the word dude, or the implications of a statement versus its literal meaning. For example, have you got any cash actually means can you loan me some because I clearly have none. So anyway, there are lots of reasons that people say the same phrase or word differently at different times other than just place of articulation or ease of pronunciation economy. So it might also come from other contextual reasons that would only become clear with more study. So if anyone wants to do more study on this subject, please be our guests. on this subject, please be our guests.

Michael chimes in that he thinks of us as Ben and Sam, possibly because he thinks of me as the play-by-play man and you as the color guy. That's interesting. Ben, I have to say, first I have to say something slightly off topic. That's probably a good thing my connection is terrible and I think it's because I called you

so I think we should hang up and you should hang with me alright I'll do that and we're back how's this? better so far see possible

all very interesting. The play-by-play on the color is interesting. Uh-huh. You do, you know, you read the credits at the end. Right. Yes, I tell people to email podcasts at Baseball Perspectives. You don't generally start the show any more often than I do, but you are clearly, you have the role of

the administrator, so that makes sense. I feel like those are all really interesting and they all make sense to me. One other thing, and maybe it's a subset of this, is that you are generally the one who puts the call out for questions. You say, send us emails for the email show approximately 3,000 times more often than I do. And you're also, you know, I assume you're doing this on Facebook. Occasionally I will see that you tweet it, whereas I am happy to, I'm personally happy to just go back and answer the unanswered questions from 17 months ago. And so it's conceivable that people are literally responding to you and then tacking me on because I'm part of the team. Yeah, okay.

Well, I'm glad we looked into it. It's also somewhat interesting that it's not 110 to 2. Right. It's fairly even. It's statistically significant, but there are a lot of people who choose the other, and I wonder what the linguist would say about that. Because we didn't ask. people who choose the other. And I wonder what the linguist would say about that. Because we didn't ask. I mean, there's 40-some people, what, 30-some percent, 35%, 30% are choosing the opposite. So what is it that's driving that decision? So we need to do a targeted survey of those people to ask about their address choices.

Hey, Ben. Yeah. I have a bonus play index oh okay well hold on uh so so that was so thank you to michael and kate for answering that michael said that he couldn't imagine a scenario where sam becomes aware of this on the air and spends less than seven minutes thinking out loud about it that was less than seven minutes though i, I think. So we came under that. But thanks to them. And now anytime you email us in the future, you can be incredibly conscious of how you're addressing that email.

Okay, bonus play index. So Koji Ohara allowed a home run, a solo home run on Wednesday. Koji Ohara has now allowed three home runs this year and has also allowed three runs this year and I wondered somebody might have wondered I can't remember if I think I think somebody asked me this so I don't know why I said I wondered somebody I think wondered this to me and I decided to look it up um maybe I maybe I did think it up I honestly have no idea the question is has anybody done this for you know a long time uh for a season

for instance what's the the question that i looked was what's the the most innings a pitcher has pitched in a season giving up no oh no nobody did one of this to me somebody alerted me to the fact okay so uh so what's the most innings that somebody has pitched in a season allowing no home runs uh outside of solo shots um and the answer is currently Koji Ohara at 33 innings. He is the record holder. He's got to dodge a lot of landmines, though, in the next three months to keep that record. The current record before him was Pat Neshek, who threw 19 and two-thirds innings in 2012 and allowed three runs on three homers.

The record for most home runs, instead of sorting by innings, sorting by home runs, is Fred Wenz, who in 1969 pitched 11 innings and allowed seven runs on seven solo home runs. Which, the amazing thing about that, and this really is truly amazing, in those 11 innings he walked 10 and gave up nine hits and somehow with a whip of like 1.8 or something he managed

to not allow any runner to score uh every runner that he allowed uh that was not a solo home run was stranded. That's pretty amazing. And John Frascatori gets a bonus mention for being number two in both these lists. 16 in the third innings, four home runs. Also, guitarist for the Red Hot Chili Peppers, if I'm not mistaken.

And one more thing, because this is the rather arbitrary season boundaries, which, of course, we know Koji Ohara probably won't make it to the end of the season, but it doesn't diminish what he's done. So looking for the longest stretch, I'm with something like 99.8% certainty, confident that the record holder for this is Tom Seaver.

And are you ready for this ben ready 79 innings without allowing a non-solo home run to score 70 79 innings from late 72 to early uh to uh to early may of 73 uh 79 innings and i think something this is do you know what his era over that i mean how many home runs was it uh he allowed um eight eight home runs in that stretch uh in so uh eight runs and 79 in so just basically bob gibson's era is like like one point uh well i guess actually not it's even better than that it was that the portion that the biggest bulk was in April, and he had a 1.1 ERA in April. But actually, he had even less than that. He had 0.91, I want to say. Okay. Might be wrong about that.

Good. That's right. So that's the answer to that. So Uehara is not even halfway there, believe it or not. Actually, I might be – yeah, I think Uehara is at – if you go back to last year, to September, I think he's at 38 or 39. So he's just about halfway there. Okay. All right. So that concludes the banter portion of this podcast.

Okay. All right. So that concludes the banter portion of this podcast. The actual topic is projections and rest of season projections, which are just, which are a hot topic right now. Oh, by the way, I guess we should mention that, that the Marlins signed Brad Penny. And we should probably mention that Joey Votto struck out against Travis Snyder, which adds to the position player pitching tally. Uh-huh. We should also mention that Bartolo Colon got a hit.

A double. Mm-hmm. And scored from second on a single, right? If it was a single, it was a joke single. Right, it was. I mean, it could have been that the batter wasn't sure whether he could go to second because he wasn't sure whether Bartolo would have vacated by that point no yeah it was it was a double yeah okay good he was uh yeah 20 21 22 seconds for bartola to circle the bases in a sprint uh not not not really in a sprint but

in anger he ran in anger and it took him 22 ish seconds all right what are we talking about projections yes so this is this has been a hot topic lately. There was an article at Mitchell Lichman, also known as MGL, prominent sabermetrician. He wrote a couple of posts at his blog last week, I guess it was, doing a study on players who have exceeded their their preseason projections by the most. their preseason projections by the most. So he looked both for position players and for pitchers, looked for the guys who 200 plate appearances into a season had overperformed and underperformed their projections by something like 40 points of weighted on base average, which is a lot.

As he said, that could be the difference between an average player and a star player or a bad player or an average player. And he looked to see what those players did for the rest of those seasons in which they were overperforming or underperforming their projections. And he compared their rest of season projection, which is, you know, updated in season. And you can find rest of season projections at Fangraphs, at Baseball Prospectus. He used the steamer projections at Fangraphs. And he looked to see whether the rest of season projection was more or less predictive of those players' rest of season performance than their performance to date in that season.

Because we'll often hear, you know, the recency effect, whatever it is, we weight recent performance quite heavily in our minds, sometimes too heavily. And if a guy has 200 play appearances where he is amazing and he's much better than we thought he was, often we will accept that he is better than we thought he was. We'll, you know, maybe we'll come up with some reason why he's better than we thought he was. Maybe we'll come up with some reason why he's a different kind of player now. If it's a pitcher, he's throwing harder, he added a new pitch, he did something with his mechanics.

If it's a hitter, maybe he's healthier, maybe he's changed his stance, something. And we'll accept that he's reached a new talent level he has a new true talent now so he uh compared you know if you look did those players continue to perform as they had to that point in the season or was the the updated in-season projection as good a a predictor of how they would play and of course, the rest of season projection takes into account previous seasons and how those players did then. And what he found was that the rest of season projection was better, that if you take into account

the previous season's performance, that will be closer to how those players perform over the rest of the year than if you just took their actual in-season performance. And this is true really up until the very end of the season. He found maybe in September, by September maybe, there's some sign that the current season performance underrates the player. But on the whole, certainly to the point in the season where we are now, as a group, as a complete population, the overperformers tend to perform like the projections say they would, not like their current season projections say we would. So there's been some discussion about this, you know, how much do we trust this? Because of course there's, even if that's true for all players as a group, it might not be true for individual players. It's certainly possible that someone could identify one of those over-performers or under-performers who will actually continue to over-perform or under-perform. And so there's

been a back and forth. There was a Twitter discussion between Dave Cameron and Keith Law about this. There was a long Facebook discussion yesterday on Kevin Goldstein's Facebook between writers and front office people talking about this. Basically, how much you should buy into this, how much you should just accept that projections know all and that we shouldn't try to pick out the players who will beat those projections. So I guess my first question to you, and we will do a little test of this ourselves. That is the meat of this topic. But what is your feeling about this?

If we assume that this is the case, that on the whole, will will play as they are projected do you feel confident relying on your own analytical skills whether it's looking at certain stats looking at at eyewitness observations do you feel comfortable in many cases going against that and saying that a certain player will outperform the projection or underperform the projection uh well i mean i feel comfortable with it in the sense that my opinions have no significance to the world. There are no stakes whatsoever to me being wrong. So I feel very comfortable.

However, if your opinion mattered. You know, I think that generally speaking, it would have to be something pretty convincing for me to trust myself over a projection. It would not be very difficult for me to trust Kevin Goldstein over a projection, however. I think that this is a situation where both sides are absolutely true. situation where both sides are absolutely true. The bulk of these projections are smarter than human eyes or human emotions or human instincts. Therefore, it is in most cases, and perhaps arguably in all cases from our perspective, wise to trust them instead of going freelancing and trying to find some reason not to trust them however um it is also indisputably true that they don't take into account some of the most important factors of the game and of performance and therefore if you are one of those select few

who is trained to do this as well as has decades of practice doing this as well as and i think this is significant too, as well as actual stakes to the extent that there is penalty to being wrong. And therefore, you can't just toss off some opinion whenever you want to. You have to actually think through your biases and get it right or else you're going to lose your job. I think in those instances, it would be foolish, foolhardy to take projections 100% at their word. So yeah, if Kevin and Keith, for instance, are seeing something, this came up about Marco Estrada, if for instance

they both saw something in his performance that is something that we know isn't included in projections, I'd be perfectly happy taking them at their word. I wouldn't argue either side. I think both sides are legitimate. And if I had to make a decision on which one to follow, I don't know, I probably would use, you know, I probably would decide by who was yelling loudest, maybe. Yeah. Right. So that's one of the things that I mentioned in my discussion, because of course, Kevin was questioning the wisdom of just relying on the projections. But Kevin has some scouting experience and he has a whole scouting staff, people who are paid in part for their ability to beat statistical projections. Yeah, I would think, though, that even in those cases, if I were them, too, I would start with the presumption that the projection system is smarter than maybe even than me.

I would still require a fairly high bar of evidence to go against it. Because I think that the important thing for everybody in baseball to realize is the limitations in what they know. And there's a reason that these projections are useful. It's because they strip away the desire, the tendency, the inclination to trust ourselves, our own minds, more than we should, and the inclination to overreact to things that we think are more important than they probably are. Yeah. I always bring up the mechanical

adjustments quandary and how I never know whether to, whether that means anything. If you see a guy, if he says, I changed my stance, I changed my swing. I feel much more comfortable. And maybe he has a little hot streak. Does that mean he is a better player now? Will he beat the projections or, you know, will he, will it turn out that that wasn't actually the reason for his performance? Or maybe he won't be able to maintain it just because a guy looks good when the last time you saw him and seems to be doing something new or different or better doesn't necessarily mean that he can continue to do that or that it will continue to work as well. So it's kind of a dangerous thing where trusting the projections blindly can get you into trouble and trusting your eyes and the recent observations blindly can also get you into trouble

in a different case. And maybe if you work for a team, maybe there are more people you can consult who can give you a more intelligent answer here, more informed answer than we have on the Internet. It might be a different equation for us than it is for Kevin or anyone else who has professional scouts who can at their beck and call, who can provide opinions on these people. can at their beck and call who can provide opinions on these people um and and sky kalkman our friend has has suggested that we try this exercise with baseball prospectus's scouting staff see whether bp scouting people can can beat the projections uh which is maybe an experiment that we will do but in the meantime let's play that game ourselves we We will have a classic Effectively Wild test here that we will know the answers to by the end of the season. So in this case, I have found the players who have overperformed and underperformed their preseason Pocota projections by the most.

So the guys who have had at least 200 plate appearances. I've got hitters and pitchers using ERA for pitchers and true average for hitters. Again, if you're not familiar with true average, we mention it from time to time. It's the BP all-in-one offensive statistic. It's on the traditional batting average scale. So 260 is average, even though 260 is no longer actually average in the major leagues. 260 is average. 300 is good. 220 is bad. You know, that sort of scale you're familiar with. And it takes into account everything a player does on offense.

And it's park adjusted and league and era adjusted and all of those adjusteds. So I have the overperformers and the underperformers. And I have the rest of season projections for these players-performers, and I have the rest-of-season projections for these players, also generated by Pakoda and updated daily. So how should we do this? Should we just take the over and under on the rest-of-season projections? Should we actually specify numbers that we think they will have? What do you think is the best way to do this?

I think we should do over under. I don't think we necessarily need to say that we are as good at projecting things as a projection system. We are merely seeing whether we can spot a bias in a projection system or a hole in a projection system. We don't need to demonstrate that we are also as good as a system that took many years to develop. We're just seeing whether we can poke holes in it. Right. Of course, if we actually use numbers, then I guess you could figure out more accurately whether we –

because we might pick someone to be over and we'll win by one point or something, and then we'll pick someone to be under, and he'll actually, you know, he won't be under. He'll be over by 78 points or something. Yeah, so how many of these are we doing? Because, I mean, if we're doing like six or something, then, yeah, we're not going to be able to demonstrate much knowledge or lack of knowledge. I mean, right.

However large a sample we do here, I mean, unless we do every player, it's probably not going to be statistically significant. it's probably not going to be statistically significant. So I was just thinking we'd just go quickly through the top and bottom 10 or something. All right. Wait, the top and bottom 10? That's 20. Well, are we just doing over-under? Because that would be pretty quick. All right.

over under, because that would be pretty quick. All right. Okay. So hitters, the number one over performer this year with a minimum of 200 plate appearances is Lonnie Chisholm. His preseason projection was 267. True average.

He has been three 50 to date. His rest of season projection has, has been bumped up by 11 points since the preseason. So it's now 278 over the rest of the season. Over or under? I will say under. Okay. I will also say under.

I guess we should probably alternate here so that we don't influence each other too much. So, yeah, I will say under. I don't think Chisholm is suddenly a super hitter. Next guy on the list. By the way, I'd like to think that after 474 episodes, we've influenced each other enough. I mean, we should be completing each other's sentences.

You're right. Next guy is Carlos Gomez. Preseason projection, 256. Actual, 337. Rest of season, 264. I guess it's my turn oh okay i was i was gonna say over also i i would have taken over on the pre-season projection as well so i would have taken over if the pre-season projection was 264 yes yes uh right gomez is one of those cases where picota is looking at many years of data he did not used to be a good hitter. He is now a very good hitter.

So, yes, over on Gomez. Okay, next, Jonathan Lucroy, preseason 265, actual 341, rest of season 275. Your turn. That's, uh, that's a good one. Uh, I will say rest season is two 75. I'll say under.

Yeah. I'm going to take under two. I mean, I would, I would go right around there, but I would take under. Okay. Next Adam LaRoche preseason 268, actual 343, rest of season 271.

I had absolutely no idea that Adam LaRoche was having that kind of year. It's my turn. And I will take the under on 271 because I think I once wrote an entire article maybe about how Adam LaRoche was just the most average player. I guess for a first baseman, I guess that is about average, but I will take under on 271. Well, if I'm not mistaken, if the over-under was always 271, he would have in his career been over exactly half the time and under exactly half the time.

Yeah, he's right. I'll say over. Okay. would have in his career been over exactly half the time and under exactly half the time uh-huh yeah he's right uh i'll say i'll say over okay he does have a 285 career but on the other hand he's 30 34 almost 35 um okay next louis valbuena uh who over actual 330 pre-season 257 rest of season 261 over over yeah i'll take over all right next i feel like no i feel like i influenced you there yeah you i definitely heard a inclination to under before you said it i don't i feel like i would need to look up how good good valbuena is to even he's never he's never been over 261 if that helps 28 years old 28 and six months 248 career but he was 260 last year and 330 so far this year yeah i'll stick with the over all right next seth smith preseason 272

actual 342 rest of season 277 i'll go first i'll take under i also will take under although uh it's i mean there's not there's just not that much difference between 272 and 277. So, yeah. But, yeah, I also take it under. Okay. This is an interesting one. Nelson Cruz, preseason 283, actual 346, rest of season 286. All right. I'll take over.

Me too. All right. All right. I'll take over. Me too. All right. Andrew McCutcheon, who is following up an MVP campaign with an even more MVP campaign. Pre-season 308, actual 371, rest of season 310. And I will take over.

I will also take over. All right. Michael Brantley, pre-season 260, actual 318, rest of season 269. So his has actually bumped up quite a bit. It is. Enough that I hesitate, but I'll say it over. Yeah, I'll buy the Brantley breakout to that extent. All right.

And last over performer is Evan Gattis. Came into the year with a.282 projection, is currently at.340, rest of season.286. Say those again. Say them again. Preseason.282, actual.340, rest of season.286. I will take the under. I'll take the under. I'll take the under too. All right.

Okay, so that's the 10 top performers. The bottom guys, starting with the biggest under performance, Jed Jorko, who was projected for 277, has been 191, and his rest of season projection is 265. Under. Under. This feels to me like a classic case of we shouldn't totally buy the slump because I don't know of any. We talked about Jorko before and I hadn't really looked in depth into his numbers then and I haven't since then.

So I'll take the over. Fair enough. I would note I mean there's He's missed a lot of time Who knows if he comes back There will be the couple day adjustment When he comes back As we know when you miss some time

You come back and it takes a few days Statistically at least And he was only 264 last year So anyway But you took over Bad bet by me. Okay, next, Brian McCann, preseason 284, current 221, rest of season 278. I'll take over.

Me too. Joe Maurer, preseason 306, so far 24 243 Rest of season 298 I'm going to take I'm going to take under I'm going to take over Jackie Bradley Preseason 273 Current 215

Rest of season 260 Oh wow 260 is a big drop I was definitely going to say wow. 260 is a big drop. I was definitely going to say under, but that's a big drop. The drops are bigger, of course, for guys with less data. Yeah, and for guys like Jackie Bradley, who are so controversial. Pocota knows how many hot takes there are out there. It's part of the algorithm.

I'll take the under. Me too. Next, Dominic algorithm. I'll take the under. Me too. Next, Dominic Brown, preseason 281, thus far 223, rest of season 272. Under. Me too. Next one, Carlos Gonzalez, preseason 300, current 244, rest of season 296. This is a hard one because he's a health guy. Right, that's the problem.

Who knows if they got all the tentacles out of the fatty mass. I would say over, but I would acknowledge a healthy degree of uncertainty. Yeah, I'll take under on that one. Alright, next, Yonder Alonso. Preseason 272, current 217, rest of season 265. Under. My hacking mass first baseman, so I'm sticking with the under.

Why did you pick him as your hacking mass first baseman? Do you recall? He sucks. Okay. Alright, well I came into this kind of a blank slate on Yonder Alonso, so I'm Do you recall? He sucks. Okay. All right. Well, I came into this kind of a blank slate on Yonder Alonso, so I'm getting influenced by you, I think. But I'll take the under. Next, Will Venable, preseason 267, actual 213, rest of season 262.

Whose turn? I'll go. I'll take over. Over. Yeah, I'm over. He's been over 262 every single year of his career. Every single one. Six for six, Ben.

Uh-huh. All right. Usually by planning. David Fries, preseason 277, current 2244 rest of season 272 oh so this one I would definitely defer to somebody else I have no opinion on this

I'll take the under just I mean people were down on freeze coming into this year I'll take the over okay alright you're cheating. You can't use a coin. That's going to be a better predictor than me.

All right. And last underperformer, Alejandro De Aza. Preseason 269, current 221, and rest of season 263. I will say under. And I remember, partly because I remember at the beginning of the year that Pakoda was seemed to be bananas on Dehaza

to start with. Yes, right. Under. Okay, so that's all the hitters. Can we do lightning round pitchers? Sure. Okay. Wait, what's the difference? This will be the same thing thing but we'll go faster um okay uh pitchers would you rather start with underperformers or overperformers doesn't matter okay then we will start with underperformers and we'll we'll end on a positive note okay so the biggest underperformer

clay buckholz um projected for 361 uh has been 702 and rest of season projection 3.92 uh over yeah that that seems like like a gimme over really? I think so he's confounded he's confounded everyone for years so to me that's actually probably one of the three toughest that we've had

yeah I don't know yeah I guess just seeing him be such a mess has influenced me. All right, Justin Verlander, preseason projection 2.84 ERA, to date 4.98, and rest of season 3.03. Over. Me too.

You convinced me with your article about Justin Verlander. Next, Colby Lewis. Colby Lewis came in with a preseason projection of 3.86. He has been 5.97. His rest of season projection is 4.23. Over. Yes. Yeah, okay.

Colby Lewis, his last start was not so good. Yeah, I'll take the under. Eric Stoltz. Wait, hang on, hang on. I think we just said the same thing, but you called it the other. Oh, sorry, yes. Yes, right. I was thinking of, yes.

In this case, the over is the bad. Over is bad. I'll take over. Okay, Eric Stoltz, preseason 4.03 ERA, actual 5.76, Rest of season, 4.42. I will take the under. I'll take the over, but this is tricky because my best bet is that he is gone from the club in three starts. I don't think that he'll give us a sample big enough. Yeah.

Okay, next guy, Franklin Morales. 4.24 projected. 5.83 so far 4.67 projected rest of season boy i am much much less uh i guess much less optimistic about pitchers or maybe just much more easily swayed because i'm just going with the over on all these. Over. Yeah, me too. And you'd think that we should be more easily swayed by pitchers. We've talked in the past about how we change our opinions about pitchers more quickly because it seems like they can change who they are more quickly. But MGLs... They can switch back very quickly too.

Right, yeah. And MGLs study, if I recall correctly, showed pretty much the same thing with pitchers as with as with hitters so not much of a difference there okay um next brandon mccarthy preseason projection 3.66 actual era 5.18 rest of season projection 3.82 your turn i'll take under and mccarthy i will also take under to me brandon mccarthy looks i i've probably watched i don't know 35 of his innings and that guy looks incredible this year yes he does he looks unhittable i have no idea how they're like i think a lot of these runs are clerical errors right yeah um yes he's he's been a bad luck guy, hard luck guy. Tony Singrani, preseason 3.06, actual 4.52, rest of season 4.2.

Wow. His projected ERA went up a run and a quarter. Yeah, that's a lot. I mean, I guess he only had, how many innings did he have before this year? Not a whole lot. And basically no minor league career. Right. he only had how many innings that he had before this year not not a whole lot he had uh and basically no minor league career right hardly any uh so i'll take the under for for because it's such a big jump and i'm gonna put some faith in pakoda um so i don't actually have much of an opinion about how good singrani is but i will i do have an opinion on how good pakoda was three months ago so i'm gonna i'm gonna defer to pakoda and say i'll take the under yes me too all right matt kane projected for the same preseason projection as as singrani 3.06 he has also been exactly the same as singrani so far 4.52 uh but his projection

obviously has not moved as much 3 3.13 rest of season. Over. Yes, I will take over too. I like Matt Cain. I'm not worried about Matt Cain particularly, but I'll take over on that number. Tim Lincecum, I think we probably barely have to talk about this one. Over.

Yeah. Yes, his projected rest of season was 3.48. That's an easy over. Okay. Last pitching underperformer, Juan Nicasio. Preseason, 4.5. Actual, 5.92. Rest of season, 4.7.

Okay. Your turn. I can't say I have a strong opinion either way on Juan Nicasio. 4.7. Okay. Your turn. I can't say I have a strong opinion either way on Juan Nicasio. I'll take the over. I'll take the under. Okay. Alright.

And now wrapping up with the pitchers who have exceeded their projections by the most. Number one on that list is Scott Casimir, whose preseason projection was 4.57. He is at 2.05, and his rest of season is 4.16. You know, the Pocota Baseball Perspectives annual comment on Scott Casimir, written by Rob McKeown, was all about the things that Dakota doesn't know about Scott Casimir. And I've been congratulating Rob every so often this year. So I'll take the under.

Yes, as will I. Next, this is an interesting one. Mark Burley, preseason 4.36, actual 2.28, rest of season 4.28. Wow. That guy just cannot move the needle. Under? Man, I want to take under because I love Mark Burley, and he's one of my favorite things about this season so far,

and he just had another pretty strong start today. But if you look at like his, his ex-fip, it is almost exactly his, his rest of season projection. He's, he's had good, good home run for fly ball rate luck, even though he's pitching in a good home run park. So, and, and he's not getting more strikeouts. He's not really showing better control. He's just kind of Mark Burley-ing,

but getting good sequencing stuff going on. So as much as I love Mark Burley and as much as I love his streak of never getting hurt, I'll take the over. Really, I'm taking the on the nose, on the dot, but I'll take the over. Really, I'm taking the on the nose, on the dot, but I'll take the over. All right. Next, Dallas Keuchel, preseason 4.68, actual 2.63, rest of season 4.29.

I know I'm supposed to—I feel like I'm slightly backlashing against the idea that I'm supposed to think that Dallas Keuchel is like a legitimate Cy Young candidate for the next five years. And so I think that I want to backlash against that and say the over. But the number is high enough that I'll still take the under. Yeah, me too. I mean, I think last time I looked, he was leading the majors in ground ball rate and also striking guys out and also out walking, and he went doing all the good things. He's doing great. He's doing great, yeah.

Okay, another Astro, Colin McHugh, preseason 4.7, actual 3.03, rest of season 4.66. The thing is that at that point, I mean, it's hard to be bad. Well, I guess from this point forward. Sorry, I'm just thinking through the scenario. I'll take the over. I'll take the under. 4.66 is high.

It's very high. I'll take the under on that.66 is high. It's very high. I'll take the under on that. This is a pitcher's era. By the way, I'm using minimum 50 innings here, which is why my cue is included here. Okay, next, Julio Tejeron. Preseason, 3.93. Actual, 2.31.

Rest of season, 3.87. Wow. Only cut.06 off, huh? Yes. Because of that FIP, because of that, 3.87. Wow. Only cut.06 off, huh? Yes. Because of that FIP, because of that FRAW. Uh-huh. 3.87 is pretty high for a park like Atlanta. Yeah.

So as much as I'm probably more pessimistic about him than you would like, I'll still take the under. Yes. We've been accused of being overly pessimistic about him before. But I'll take 3.87. Yeah, I'll take the under on him. Willie Peralta, one of your favorites. 4.58 preseason, 2.98 actual, 4.55 rest of season.

3.98, actual, 4.55 rest of season. Dude, I would take the under on 2.95 or 2.98 or whatever number you just said. You love Willie Peralta. I do love Willie Peralta. Yeah. That's an easy choice for me, so I'll go with the under. As will I. Really?

Because it's not like his numbers are all that great on the CRA. I mean, yeah, 4.55. That's high. Tim Hudson, preseason 3.4, actual 1.81, rest of season 3.29. So the deck is kind of stacked here because he gave up seven runs in five innings today. Oh, did he? Okay. Well, I was going to take the over for what it's worth.

I would also take the over. Yeah, 3.29. Yeah, in that park, I think going to take the over for what it's worth. I would also take the over. 3-2-9. Yeah, in that park, I think I would take the under, but can I punt this game? I mean, given that he starts so far behind, I'll take probably the over, but otherwise I would take the under. Nobody's going to check this.

It doesn't even matter. People check things. We get updates. i'm taking the under on hudson okay jordan lyles pre-season 5.08 season to date 3.52 rest of season projection has not budged much 5.04 uh under me too yeah all right uh almost almost there two more to go danny Duffy, preseason 4.39. I don't know whether he was projected as a starter or a reliever or what. He's been both. But 4.39 preseason, 2.83 thus far, 4.18 rest of season. It's a tough one. It is a tough one.

I'll take the over. Gosh. I will take the under. Okay. And last one, Chris Young, the tall one. Preseason, 4.96. Actual, 3.40. And rest of season season 4.63.

Over. That's pretty much a push for me, but I'll take the over. All right, so that's it. That's 10 of each type. Maybe someone will compile these for us. If not, then I don't know. Now let's draft them. Right. Okay. So that was fun. We'll see whether we have any success in beating the projections.

Although even if we do, the sample is still too small to say that we have any actual talent for beating projections. So there's 40 coin flips here. How many do you need to get to feel like you did something? I need to get 28. Okay, I was going to say 26. 28 and 12 would really be something. Yeah, I'd be pretty honest about that. 26 and 14, I'll probably dumbly boast if I get 26, but I won't. I'll get 18.

Yeah. Okay. All right. Well, so that was a fun exercise for us, at least hopefully for some of you. That concludes the baseball portion of this podcast. I have one more update on the linguistics portion of this podcast. I received an email from Kate, Michael's wife, while we were recording, who thought of two more things. She says that language in the brain is a dynamic system, so any or all of the things I listed could be affecting someone's choice of phrasing all at once, like a bunch of different weights settling onto a scale, eventually causing a decision between the two. Also, phrases like this one tend to eventually settle into one order over another, given enough use and enough people agreeing on it.

By agreeing, I mean following into the pattern of using one over the other. The eventual winner is probably mostly determined by sheer frequency of use and exposure. So if Sam wants to fight back, for example, he could start flooding your shows with Sam and Ben. So we'll see whether now that people know that Ben and Sam is the more popular choice, that it becomes a dominant choice or whether anyone wants to rebel and fight back. Okay. I also sent out this question to multiple people who responded to my Twitter inquiry. We also got a response from our friend Ken Arneson,

who writes excellently about baseball, sometimes about the A's, sometimes for the BP Annual. He also has some linguistic experience. He says it could be a sociological explanation in that I am the editor-in-chief of Baseball Perspectives, so I'm higher in the hierarchy. So people put my name first. I don't know if i buy

it he guesses that if it has something to do with linguistics similar to what kate says it's because ben and sam is a simple mouth movement palindrome sam and ben on the other hand makes your mouth change position four times to pronounce its consonants there are six consonants five different ones in either phrase two of them are bilabial three of them are alveolars. Bilabial means they are made by placing the lips together. Alveolar means they are made by placing the tongue on or just behind the alveolar ridge on the roof of your mouth. S and N and D are alveolars. B and M are bilabials. So the sequence with Ben and Sam is bilabial, alveolar, alveolar, alveolar, alveolar, bilabial. Your point of articulation changes twice. The sequence with Sam and Ben is alveolar, bilabial, alveolar, alveolar, bilabial, alveolar. Your point of

articulation changes four times. I like that answer. So that was much more thorough investigation of the topic than I intended to make. We started the show when Bartolo Colon crossed home plate. He is just now completing his walk back to the dugout, which means that we've been talking for far too long. So that is the end of this podcast. Please support our sponsor, Baseball Reference. Go to baseballreference.com, subscribe to the Play Index using the coupon code BP to get the discounted price of $30 on a one-year subscription.

We will be back with another podcast about baseball and probably not about linguistics tomorrow.