The Brewster

I was thinking about Mike Trout's HR totals, and wondering how many homers he would have if pitchers came at him like a 100 OPS+ batter. Then of course, you remember Bond's absurd 73 homer season. The most ridiculous part is that he did that in 476 ABs, and would then hit 40+ homers the next few years in around 400 of fewer ABs.

It made me wonder if OOTP models this at all. All the walking of Bonds was probably a bit much. But Trout's a very good example of a realistic approach to a super hitter. And to some degree, pitchers work around the top power hitters in the league. It just makes sense.

There was some talk awhile back about Wareham and how his walk rate given that he can't really make contact is kind of broken and a flaw in the game design. I wonder about the approach to power hitters, too. Is the EYE rating too simplistic of an approach, or does the game adequately weigh how "afraid" of a hitter a pitcher is.

I also wonder if this plays into the BBA's inflated home run totals. It has way more 40+ homer guys than a real league. How much of that is ratings, and how much of it is OOTP potentially not modelling what would happen in real life? If you're on pace for 50+ homers, typically teams stop giving you stuff to hit.

I agree. Baseball is a game of adjustments. Pitchers don't have a "book" on young hitters. Those young hitters may tear it up for a while but pitchers will adjust and find the "holes". Then the hitter will adjust to how the pitcher is getting him out and the game goes on. I don't know how the game handles this. Even if it does.

If a hitter is on a hot streak in real life you see pitch arounds, intentional walks and matchups.

I don't know if any of this baked into the algorithm or any other way that it is accounted for. Maybe that is why we get some inflated numbers in certain stats.

I don't know how much people mess with strategies, but that is a slider. Best way to test is to mess with that I'd think. I know I always keep mine down. But I'm not sure how much people adjust them.

I have used the sliders. It allowed Lionel Crepin to hit 4 homers off my staff on his big game. You would think that after 3 someone would have held up 4 fingers.

bcslouck wrote: ↑

Mon Aug 03, 2020 5:24 pm

I don't know how much people mess with strategies, but that is a slider. Best way to test is to mess with that I'd think. I know I always keep mine down. But I'm not sure how much people adjust them.

I knew there was a slider, but does it do it on its own? And to what degree?

How would you test it?

RonCo wrote: ↑

Wed Aug 05, 2020 1:23 am

How would you test it?

I don't really know. I think you'd have to do something with test leagues. A lot with test leagues.

Really, the question is, do OOTP pitchers adjust correctly from the raw batter ratings based on performance? We know that, in cases like Wareham for example, it doesn't. To take that further, if you make a batter that is say 3/3/3/13/3, he should be almost as bad as a batter that is 3/3/3/3/3. Because pitchers should just throw strikes as best they can and not worry about the guy making contact on stuff he's clearly overmatched by.

This is really the reverse question. Should a 13/13/13/13/13 batter be any better than a 13/13/13/3/13 batter? Should the second batter walk less? Maybe marginally, but at some point, the pitchers shouldn't really be giving either one anything to hit.

So to test stuff, I think you'd have to create those extreme cases. You also have to get an idea of what one point of EYE is worth from an OBP standpoint. Really a 2 part process then. Make a 6/6/6/6/6 batter. And a 6/6/6/7/6 one, and 6/6/6/8/6., and so on. Play a bunch of games so wwe know what one point of eye is wroth.

THen you ratchet up the other stats. 7/7/7/6/7, 7/7/7/7/7 ...
THen 8/8/8/6/8, and so on, because you'll get diminishing returns for those points of EYE as the batter improves in other ways, I'd expect.

After all that testing, you'd ahve an idea of what one point of EYE is worth for a given talent level.

So then you could compare the all 13 with 13 eye and the all 13 with 3 eye. It still wont be perfect, but you'd have enough trend data to make some conclusions.

I sure as hell don't want to do it. My suspicion, is that the game breaks at a point. Just like with Wareham not getting meatballs down the middle until he probes he can put the bat on them, I suspect that there's a point where no one in their right mind would throw a pitch over the plate, but the game will just do it anyway. And base on the fact that Wareham really isn't that extreme, I'd guess that the point at which OOTP deviates from reality is "sooner" than we'd think.

The other reason I suspect it's not modeled correctly is that there's practically no data to model it on. We know pitchers don't go as aggressively at the Mike Trouts of the world, but how many of his walks are because of it? What about great but not Trout. Juan Soto?

You never really know on these uber hitters how good their ability to take a free pass is because pitchers are scared of them. Look at what happened to Pujols BB rate as he lost bat speed and hitting ability. Yes, strike zone judgement is a perishable skill, but not to the extent Pujols has lost it. Pitchers just stopped being afraid of him.

So anyway, it would be fun data to know, but ultimately there's little to compare it to in real life, so how would you know it's modeled correctly?

Ted wrote: ↑

Wed Aug 05, 2020 1:44 am

The other reason I suspect it's not modeled correctly is that there's practically no data to model it on. We know pitchers don't go as aggressively at the Mike Trouts of the world, but how many of his walks are because of it? What about great but not Trout. Juan Soto?

You never really know on these uber hitters how good their ability to take a free pass is because pitchers are scared of them. Look at what happened to Pujols BB rate as he lost bat speed and hitting ability. Yes, strike zone judgement is a perishable skill, but not to the extent Pujols has lost it. Pitchers just stopped being afraid of him.

So anyway, it would be fun data to know, but ultimately there's little to compare it to in real life, so how would you know it's modeled correctly?

Got to thinking about this more. One of the questions is "when to pitchers pitch around hitters?"

Your cases are more related to hitter quality, but I tend to think about pitching around hitters as being more about a combination of situation and quality, weighted perhaps a little more toward situation than quality. One may always pitch around Barry Bonds at his steroid peak, but those are rare hitters.

In that light, it would be interesting to see pitcher performance against various hitter types split by game situation (outs and runners on base).

With a runner at third and two outs, how often does Albert Pujols walk, and on at bats where he does not walk how well does hit hit compared to other at bats? That would be hard to test, but is doable with maybe some heavy lifting modifications on my little script, and a tightly controlled test scenario.

Ted wrote: ↑

Wed Aug 05, 2020 1:40 am

Really, the question is, do OOTP pitchers adjust correctly from the raw batter ratings based on performance? We know that, in cases like Wareham for example, it doesn't. To take that further, if you make a batter that is say 3/3/3/13/3, he should be almost as bad as a batter that is 3/3/3/3/3. Because pitchers should just throw strikes as best they can and not worry about the guy making contact on stuff he's clearly overmatched by.

I think this is still an interesting question, too. One that has no answer that I can "prove" as far as it goes.

In Wareham's case, we say, hey, pitchers would just come into him if they know he can't hit. But the counter side here is to wonder how much a guy who walks a ton creates those walks by fouling off pitches until the pitcher just misses four times? We've all seen cases where a pitcher can barely throw a strike even if they're trying, too. So if a pitcher just throws simple "get it over" fastballs, then what's to say Wareham's BABIP rating isn't then under-stated?

In other words, that low contact rating assumes league average pitchers throwing major league quality pitches. If a pitcher just decides to grove AAA curve balls 86 MPH fastballs in order to avoid that big Eye rating, why wouldn't Wareham's effective contact go way up?

Modelling human interaction is hard.

RonCo wrote: ↑

Wed Aug 12, 2020 2:05 pm

Ted wrote: ↑

Wed Aug 05, 2020 1:40 am

Really, the question is, do OOTP pitchers adjust correctly from the raw batter ratings based on performance? We know that, in cases like Wareham for example, it doesn't. To take that further, if you make a batter that is say 3/3/3/13/3, he should be almost as bad as a batter that is 3/3/3/3/3. Because pitchers should just throw strikes as best they can and not worry about the guy making contact on stuff he's clearly overmatched by.

I think this is still an interesting question, too. One that has no answer that I can "prove" as far as it goes.

In Wareham's case, we say, hey, pitchers would just come into him if they know he can't hit. But the counter side here is to wonder how much a guy who walks a ton creates those walks by fouling off pitches until the pitcher just misses four times? We've all seen cases where a pitcher can barely throw a strike even if they're trying, too. So if a pitcher just throws simple "get it over" fastballs, then what's to say Wareham's BABIP rating isn't then under-stated?

In other words, that low contact rating assumes league average pitchers throwing major league quality pitches. If a pitcher just decides to grove AAA curve balls 86 MPH fastballs in order to avoid that big Eye rating, why wouldn't Wareham's effective contact go way up?

Modelling human interaction is hard.

Definitely it's difficult. I get your Wareham argument. I think the problem iI have is that the ratings allow production of stat lines we just don't really see in real life. The general idea that pitcher don't adjust like they do in real life does show up in the 65 homer, .320 OBP batter, and .400 OBP, .220 avg hitter with no power. I get that perhaps those types of players dont exist in real life either, but moreso I'd say it's that real life players will exploit a glaring flaw. Whereas a game really can't that well. I think you and I talked about this a long time ago with FIP. It's really tough to find a player in real life who k's 4 per nine and is consistently effective. But in OOTP you can just make a guy who is 4/10/10 and he'll be terrific.

I don't think it would be tough to code the idea of pitchers adjusting approach to batters. But probably very difficult to prove you're doing it right. And we haven't even talked about batters adjusting their approach to pitchers. Maybe I'll just let this one be.

So follow up question. The ability to make pitchers that consistently outperform their FIP in OOTP make you wonder about pitcher WAR's reliability for the low K guys. Do we need to worry about batter WAR for the .310, 65 homer guy? Or the .410 OBP, .200 average and no pop guy?

I think the answer is no because of the way linear weights work. But I'm not sure there isn't a thing I'm not thinking of.

I'll do some more noodling, but I think we need to be careful when we say no real players ever operate like XXXX OOTP player. I spent literally two minutes, went to Fangraphs and called up 2019 hitter data and sorted on BB rate.


Danny Vogelbach: Seattle 2019
PA: 558
Avg: .208
BB Rate: 16.5%
K-Rate: 20%
HR: 30
wOBA: .333


Mark Wareham: TWC 2043
PA: 435
Avg: .195
BB Rate: 17.2%
K-rate: 30%
HR: 20
wOBA: 301

These are fairly similar results. Or at least not so dissimilar that you want to throw the game across the room.

RonCo wrote: ↑

Wed Aug 12, 2020 11:30 pm

I'll do some more noodling, but I think we need to be careful when we say no real players ever operate like XXXX OOTP player. I spent literally two minutes, went to Fangraphs and called up 2019 hitter data and sorted on BB rate.


Danny Vogelbach: Seattle 2019
PA: 558
Avg: .208
BB Rate: 16.5%
K-Rate: 20%
HR: 30
wOBA: .333


Mark Wareham: TWC 2043
PA: 435
Avg: .195
BB Rate: 17.2%
K-rate: 30%
HR: 20
wOBA: 301

These are fairly similar results. Or at least not so dissimilar that you want to throw the game across the room.

I think we'll have to disagree on similar. I get where you are going, but 30 points of wOBA is pretty big for me. I get your point though.

And actually, the player that really got me thinking about this is Emilio Morales, who, if I went by WAR only, I would not put in the Hall of Fame. He'll end up with 600+ homers, and a sub 60 WAR. That's bonkers.

And it's because he has seasons like this:
2039 54 HR, .310 OBP
2040 66 HR, .318 OBP
2042 43 HR, .295 OBP
2043 49 HR, .264 OBP

I don't know for certain, but I don't think those types of seasons exist. There's two theories on the flaws. Either pitchers in OOTP aren't pitching around a guy who is that big of a power threat like they should. Or, OOTP has created an unrealistic player. A player talented enough to get the good part of the bat on the ball like Morales at that kind of frequency should either run an average so high that his average alone should prop up his OBP in the .330 range at a minimum. Or he should judge pitch location well enough to walk more.

Morales has a .315 OBP and 600 HR. I don't know Morales's BB rate off hand, but here are some historical comparisons

There are 9 players in MLB history with 600+ homers. The lowest OBP is Sammy Sosa's .344. The average is at least .380, and probably closer to .400.

The lowest OBP in the 500 homer club is Ernie Banks .330 OBP. The 500 homer guys are very similar and maybe a tad better at getting on base than the 600 HR club guys. Again, the OBP in this group is probably near .400.

To get an OBP below .330, you have to get to Andre Dawson at 45th all time with 438 HR.

By the way, as a Cards fan, this is getting funny. Cubs fans really don't know what an all time great bat looks like. All their bests have historically low walk rates to go with their historic power.

To break .320 you have to get to Alfonso Soriano at .319, 412 HR, 54th all time. Seriously, another Cub? And by the way, the overwhelming majority of players with 400 HR have OBPs well over .340. .340 is kind of the low end of the typical guys with 400 HR.

To get at or below Morales BB rate, you'd have to get to Joe Carter at .306, 396 HR, 61st all time. (Edit: in a previous version I mixed up Joe and Gary Carter)

But even the next 20 guys on the list average OBP's well over .340.

Because of that, Morales career 122 wRC+ is low for a guy with 600 HR. The real life 600 homer guys probably average around 150 for career. The 500 homer guys probably 135 or higher.

Sammy Sosa, by the way, has a career wRC+ of 124, so there's that at least. Poor Cubs fans, lol.

But when you look at the specifics, a real life version of Emilio Morales has never existed. Not even close. No one with that kind of power has ever drawn so few walks. So again, either the game isn't modelling pitching around power guys properly, or the game makes players who don't really exist (i.e. guys talented enough to hit so many homers but can't tell a ball from a strike).

Yeah, that's why I said "relatively." 30 points of wOBA is not insignificant--but they are close enough you can squint your eyes and say it's not horrific that the game creates a Wareham who can succeed a bit--especially if the guy can play defense.

------------------------

Regarding the k/9 bit:

Again, I question the base assertion that players cannot be successful at 4 K/9. At question is the run scoring environment and the players around that pitcher. There are MANY pitchers in real life who are pretty good at that range. Here are three just off the top of my head (in other words, without doing any research beyond going to BBRef to get their numbers).


Through the last half of his career, a guy like Tom Glavine built great success on a K/9 ratio of about 4.6 over those years. He used that span to help make the Hall of Fame.

Jamie Moyer also had some successful years down in that range.

Randy Jones won the 1976 Cy Young award at 2.7 K/9 and came in second the year before at 3.3 K/9. He finished his career at a 3.42 ERA (3.56 FIP) and a 3.4 K/9.


Those are starters. I'm willing to bet I can find a bunch of LOOGYs and other relievers who are just as strong in limited innings. It all fits the idea that a pitcher can get along with a whole almost anywhere as long as the rest of his package is something close to elite. So, bottom line, I don't have a problem with a 4/10/10 being a very good pitcher, as long as the game doesn't make a gazillion of them--which it does not (remembering a look I took last night, I think the BBA has only two or three guys who would be considered close to that..actually there are a few more with low stuff, but those are unsuccessful pitchers, and don't fit the conversation).

---------------------------

Regarding FIP vs. ERA performance (and WAR):

I'm not sure we see too many FIP over-performances relative to the MLB, but that's because I don't know what the MLB baseline is. I know about half of MLB pitchers fell on either side of that comparison last year, but have no idea how to compare the data over time right now.

Among my issues with OOTP FIP and WAR is that I also have no idea how OOTP does WAR, so I always have angst with it. The one advantage is that, however it does its calculations, it's the same calculation for everyone--so at least as a relative number it's got some information. I've asked the Devs to publish their WAR calculations, but they don't seem to see that as a real priority.

Other things to think about as I ruminate on your supposition/question ...

2019 NL Pitching: League ERA = 4.38, League FIP = 4.43, 10 of 15 overperformed
2018 NL Pitching: League ERA = 4.02, League FIP = 4.06, 10 of 15

2019 AL Pitching: League ERA = 4.60, League FIP = 4.58, 6 of 15
2018 AL Pitching: League ERA = 4.27, League FIP = 4.23, 6 of 15

So the NL seems to over-perform, and the AL seems to underperform. With no other information, the idea of pitching to pitchers helps the NL performance. Could league-wide variance be influenced by interleague play? Dunno.

MOVING TO THE BBA:

Looking at StatsPlus, twenty of or thirty-two BBA teams have outperformed their FIP E-F is negative). So maybe there's a distribution thing going on, too.

When I put each team's E-F against a wonky little way I use to judge hitters parks and pitchers parks, I find no real correlation to ballparks, but when I do the same using my defensive PAA metric for every team, I find a very strong Correlation of .73. So perhaps a pitcher who plays most of his career behing a good defense will tend to over-perform their FIP. This makes some sense, right?

Of the 100 qualifying pitchers (per StatsPlus), 55 had ERA better than FIP, 45 did not. This is somewhat self-selecting, though, since you would expect less effective pitchers to pitch fewer innings, and therefore not qualify as often. Two of those pitchers (Dan Cannon and Dani Garcia) overperformed by a run or more. In the MLB in 2019, there were three such pitchers (Dakota Hudson, Mike Fiers, and Jeff Samardzija).

So, again, I'm back to looking at the accusation that OOTP is out of whack being somewhat suspect (at least as it is being used in the BBA world).

Lol, we both wrote super long posts at the same time. I'll read yours in a sec. I want to amend my last post. Morales can actually tell a ball from a strike. He's very good at avoiding k's. Which makes him make even less sense. Elite exit velocity, very very good for a power hitter at avoiding K's. Never walks. So he just must run one of the worst babips ever compared to his other stats.. It would be fun to track his BABIP against the all time great power hitters. I bet it's absurd.

Ted wrote: ↑

Thu Aug 13, 2020 12:16 pm

And actually, the player that really got me thinking about this is Emilio Morales, who, if I went by WAR only, I would not put in the Hall of Fame. He'll end up with 600+ homers, and a sub 60 WAR. That's bonkers.

And it's because he has seasons like this:
2039 54 HR, .310 OBP
2040 66 HR, .318 OBP
2042 43 HR, .295 OBP
2043 49 HR, .264 OBP

I don't know for certain, but I don't think those types of seasons exist.

We have had many great conversations about Morales, who I view as deeply over-valued.

Emilio Morales, however, is kind of like Dave Kingman if Kingman could play defense and if Kingman had played in the modern era. Kingman was always also over-valued for that ability to hit HR in a time when HR were not overly prevalent.

Using raw stats is a hard game when translating over eras, though, so of course Kingman had 30-50 HR seasons rather than 50s and 60s.

But, especially when you note that Morales' HR rate is bloated by about 4-5 a year due to his park, you have to come to the conclusion that at the end of the day, he's not really a very good hitter, like Kingman wasn't. He can contribute a long career due to that glove, though. If I were GMing with him on my team, I'd have hit him 8th or 9th, probably. Omaha has had a problem for years, IMHO, because they had a OBP hole in the middle of their lineup.

But he does have a mirror in Kingman.

The real unicorn is Hugh Mangrowthormone, which I'm sure was not really grown by OOTP so much as edited into existence by our Custom Ammy approach.

Thanks for the data, Ron. I was under the impression that a consistent FIP beater was tough to find in real baseball. Year to year, there may be similar numbers of guys beating their FIPs in OOTP versus real life, but I wonder if its the same guys in MLB like it's the same guys in OOTP every year. Assuming it is the same guys in OOTP. It's possible I'm just wrong about this. I would think that given that it's dependent on ratings, FIP beating in OOTP would be the same guys more often, but maybe not.

Ted wrote: ↑

Thu Aug 13, 2020 12:30 pm

Thanks for the data, Ron. I was under the impression that a consistent FIP beater was tough to find in real baseball. Year to year, there may be similar numbers of guys beating their FIPs in OOTP versus real life, but I wonder if its the same guys in MLB like it's the same guys in OOTP every year. Assuming it is the same guys in OOTP. It's possible I'm just wrong about this. I would think that given that it's dependent on ratings, FIP beating in OOTP would be the same guys more often, but maybe not.

If the idea that a pitcher playing with a good defense behind him would consistently beat FIP is true, I actaully wonder if the prevalence of consistent FIP-beaters was higher in the old days when the reserve clause kept good teams together forever (hence would mean a good pitcher would have his defense with him most his career), and less prevalent in the age of Free Agency, when defenders and pitchers are constantly shifting teams.

RonCo wrote: ↑

Thu Aug 13, 2020 12:28 pm

Ted wrote: ↑

Thu Aug 13, 2020 12:16 pm

And actually, the player that really got me thinking about this is Emilio Morales, who, if I went by WAR only, I would not put in the Hall of Fame. He'll end up with 600+ homers, and a sub 60 WAR. That's bonkers.

And it's because he has seasons like this:
2039 54 HR, .310 OBP
2040 66 HR, .318 OBP
2042 43 HR, .295 OBP
2043 49 HR, .264 OBP

I don't know for certain, but I don't think those types of seasons exist.

We have had many great conversations about Morales, who I view as deeply over-valued.

Emilio Morales, however, is kind of like Dave Kingman if Kingman could play defense and if Kingman had played in the modern era. Kingman was always also over-valued for that ability to hit HR in a time when HR were not overly prevalent.

Using raw stats is a hard game when translating over eras, though, so of course Kingman had 30-50 HR seasons rather than 50s and 60s.

But, especially when you note that Morales' HR rate is bloated by about 4-5 a year due to his park, you have to come to the conclusion that at the end of the day, he's not really a very good hitter, like Kingman wasn't. He can contribute a long career due to that glove, though. If I were GMing with him on my team, I'd have hit him 8th or 9th, probably. Omaha has had a problem for years, IMHO, because they had a OBP hole in the middle of their lineup.

But he does have a mirror in Kingman.

The real unicorn is Hugh Mangrowthormone, which I'm sure was not really grown by OOTP so much as edited into existence by our Custom Ammy approach.

I'm not really sure you can compare Kingman. The different between a 30-50 HR and a 45-60 HR guy is night and day. There is an order of magnitude difference in the number of guys who've done it. The talent level is so dramatically different. I think you can strikeout swing your way to Kingman's power numbers, but you really have to be a complete hitter to get to 600 homers, or put up 50 on a regular basis. And as I mentioned in my amendment, Morales really doesn't strike out. Questions about how OOTP does WAR aside, you're talking 20 career WAR for Kingman versus 55 for Morales. They really aren't in the same realm. That's what makes Morales so weird. I don't think he can actually exist. You jsut can't have the taleent to hit homers like that and not hit for an absurdly high average, or walk a ton (either because pitchers are afraid or because you yourself have the strike zone judgement). I don't believe it's possible for it to ever play out that way.