Relative Ratings and Platoon Splits: Batters

[Ted]

I've suspected for awhile now that OOTP's relative ratings do not handle big platoon splits well. What I wondered is, "Is it weighing the likelihood of a player with a big split to not face the weak half of their split heavy enough when giving overall scores?" This would be hard to do, but I think it probably isn't done much at all. There's also the question of whether it should be, or if that is something to be left up to users. For a long time, I think OOTP has not done a good job in the way it handles platoon advantage. It basically leaves it up to us understand that a left handed batter 6 is better against a right handed pitcher than a right handed batter 6. The problem is the degree to which that player is better. It is a LOT. I've felt this should be baked into the ratings somewhat. But unless you hold the ratio of RHP to LHP perfectly constant, that gets really tricky.

I also wonder, how much is my perception that it's dragging down it's overall grade too much on any player with a split related to my understanding that the player won't be facing their weak side much, or does the game really over do it. Here's the player who is my inspiration.

http://montybrewster.net/BBA/HTML/news/ ... 26218.html

Anyway, that got me wondering if the overall ratings are wholly inadequate to evaluate players with big splits. I started with batters, and here is what I did. In a default MLB league with our settings for seeing ratings (1-10, show ratings above 10, relative ratings, etc) I made a left handed second baseman who is the middle of 50 for his overall rating. He's a 7 defender at 2B with 7 speed. I feel that's fairly "meh" to not throw the calculations off much. Being a 2B will make his batting ability to get to 50 a bit lower than say a 1B, but I don't think it will affect the data much. I'll check.

To be a middle of the road 50, all his internal ratings for BABIP*/GAP/POW/EYE/AVK were set to 107 and 108 (see table below).

(* For those of you that don't know, THERE IS NO INTERNAL CON RATING. CON is a function of BABIP and AVK, and to a lesser degree POW. To maintain any given CON, if you hold power steady, increasing AVK means you HAVE to decrease BABIP, and vice versa. This is critically important when evaluating players and everyone needs to understand it. It's also not how baseball actually works and it a pet peeve of mine.)

All 100 was 35
All 106 was 45
All 107 to 108 was 50
All 109 was 55
All 112 was 60

So I settled on 107 to 108 (108 for pow and gap). This gave me a 6/5/6/6/6 left handed 50 rated 2B.

The first thing I note is that there is NO CHANGE when you make the batter right handed, confirming that OOTP does not take handedness into account much if at all when assigning overall scores. You just have to know that left handed players are "better", in that if you use them to their platoon advantage, they will massively outperform RHB versus RHP. But that's not baked into the game at all. (FYI there was no change with switch hitters either).

Okay, so now I started making splits. In the next three cases, I only changed the versus left splits. The versus right splits stayed 107 and 108 and the player stayed 6/5/6/6/6 versus right handed major league hitting.

Case 1: a little worse against LHP (Player is 6/5/6/6/6 against AAA LHP), change all LH splits to 101 -> vsl becomes 5/4/5/5/6 in bigs-> Overall to 45.
- This seems reasonable. Slight detraction from my "middle 50" player makes him just below a 50.

Case 2: a lot worse against LHP (player is 6/5/6/6/6 versus AA pitching), change all LH splits to 85 ->vsl becomes 4/3/4/5/5 in bigs-> Overall to 35
- Seems drastic, right? This is where OOTP not taking into account that players platoon is a real problem. This guy is essentially the same player as the original version. The game is basically saying he is a"worse player" because he has to be platooned, which is true, but I think it this system discounts the understanding of uses that LHB won't play against their weak side as much.

Case 3: abysmal versus LHP (player is 6/5/6/6/6 versus LHP at low A level) change all LH splits to 63 -> vsl becomes 2/2/3/3/4 in bigs, overall to 25.
- This is kind of nuts. The player's utility is basically the same as case 2. He's NOT going to face LHP in either case.

Okay, so here's where it gets wacky. In each of the preceding cases, when I change the better handedness from a left handed bat to a right handed one, the overall rating DOESN'T CHANGE. That's right. A left handed bat with 6/5/6/6/6 splits against RHP is rated the same as a right handed bat with 2/2/3/3/4 splits against left handed pitching. Notably, that right handed bat is still 6/5/6/6/6 against RHP, but these are VASTLY different players and the batting handedness is not taken into account at all.

So that's a big flaw with the rating system. And with relative ratings, where the increase or falloff once you get outside the big "median cluster" is pretty dramatic, the misleading nature of the overall rating of players with huge splits is rather impressive.

Let's try a couple increases. We're back to left handed batters.

Case 6: Increase all LH splits to 109 - > overall still 50. Ratings 6/5/6/6/7 versus LHP
- not a big change. Makes some sense. All ratings 109 barely made the guy a 55 if you remember from above.

Case 7: increase all LH splits to 112 -> overall to 55. Ratings 7/5/6/6/7 versus LHP.
- Still seems reasonable.

Case 8, swap these LH and RH splits from case 7, that is, 107 and 108 versus lhp, 112 versus RHP. - > player is a 60 and 7/5/6/6/7 versus RHP
- Okay, so here's where we start to see what the game does. It doesn't take batter handedness into account. What is does is rate a player better for hitting RHP well more than it does for hitting LHP well. That's part of the problem with the overall rating. It kind of makes sense because there are more RHP's, but it's a secondary way of rating platoon advantage. It approximates the effect, rather than explicitly define it.

And again, there was no difference in making the batter LH versus RH for these. This means there is more to be gleaned here. Case 7 for a left handed batter is an LHB with reverse splits, but who is not worse against RHB, just slightly better against LHB. (These players are very rare and frankly can perform kind of strangely because the reverse split of a batter tends to get overpowered by the pitcher's platoon advantage).
The RIGHT HANDED version of case 7 is the typical OOTP RHB with a slight positive split against LHP.

So, the next logical question is what happens with lower RH splits?

Case 9: back to 107 and 108 versus LHP and AAA level versus RHP (101). This is the reverse of case 1. Player become 5/4/5/6/6 versus big league righties. Overall goes to 40.

So what we've confirmed here is that the game changes the overall potential MORE for changes in versus right splits than versus left splits. The exact opposite splits from case 1 (i.e. worse against LHP) made a 45, not a 40.

I reversed the splits from case 2, and the game called the batter a 20.


Study results for an MLB league:
1) Batter handedness is not factored into the overall rating.
2) The ability to hit right handed pitching is weighted more heavily than the ability to hit left handed pitching. The amount is significant!
3) For hitting LHP like a lower level player (bigs to AAA to AA, etc) the game removes points from the overall rating that is something like 5 for the first level, 10 for the second, etc. The increase "demerit" with each level makes sense in a relative ratings environment where the game is concerned how CLOSE you are to the median.
4) For hitting RHP like a lower level player (bigs to AAA, etc etc) the effect on the overall ratings seems about double that of the verus LHP. I.E something like 10, then another 20, etc

Takeaways:
The overall ratings for platoon players is misleading in that it expects them to play against all players. In some ways, this is right. However, it causes the overall ratings for players with big splits to be of little value, as those players will overwhelmingly be in platoons. Or at least they should. But what we have now is a way to estimate how good a batter with a platoon split could be. There is the problem of this being an MLB league versus our league, but I can't edit players in our league. Just messing around shows our AAA is really talented. Half the players don't change from AAA to AA, and the half that do don't change from AA to A. And only half the A players change from A to short A. So while the concept is the same, the scale is different. I'll basically add/sub 5 points for each level on the LH side, and 10 for each level on the RH side. This needs some testing but I don't have time right now.

Let's look at an example of my player above to see how you can use this. You change the relative ratings until the weak side looks like an average ish hitter at that level. I'm going to use kind of "mostly 5's" to have a save/conservative estimate.

http://montybrewster.net/BBA/HTML/news/ ... 26218.html

Julio Medina is 6/7/6/5/6 against RHP and 4/4/4/3/5 against LHP. That's a big split, and probably explains why the game is only calling him a 30, which seems really low. At ratings relative to AAA he's 5/5/5/4/5 on his weak side. At ratings relative to AA he's 6/5/5/4/5, which is probably about an average AA bat. At ratings relative to single A he's 6/5/5/4/6. That's still a pretty average A bat, just like a 6/5/5/4/6 bat would be pretty mediocre to average at our level. When you change him relative to short A, he becomes 6/5/6/4/6 against LHP. Now we're starting to get maybe a bit better than average.

What I think this means is that the game is saying while Medina's 6/7/6/5/6 bat is decent against RHP, his versus LHP bat is A level. So looking above, we had to move four levels down before his bat was better than average. It was average at three levels down. So that's 15 points, making his platoon only ability overall rating 45, and that seems about right.

This kind of analysis will be fuzzy. But it can give you an ideal how much the game might be killing you overall rating of you guy with a big platoon split.

I'll try to do some more testing later on players in our league and come up with a more reliable model, but it's time to chase kids ten years younger than me around a basketball court for two hours in a attempt to continue doing something I love despite the inevitability of aging.

[Ted]

I was debating whether you should decrease levels until their ratings look average, or until their weak ratings equal their strong side ratings. Above I did the former, because my test case was an average player. I think in reality it may be the latter. It makes more sense that if you want to eliminate the game's concept of platoon advantage you'd have to make the player the same versus both handedness of pitcher. Oh well. More testing later.

[RonCo]

Running coments are I read this...which is very good, BTW.

REGARDING CONTACT vs. BABIP vs. AVK:

When I hear you talk about this, it's like you're asking CON (batting average) to be a skill. That's what I hear when I hear you say "if you want to hold CON steady..." Mathematically, this is true. CON = f(BABIP/AvK/HR). Where I fall off the rails is where you say "this is not how baseball works."

I suggest you rethink that. Because for hitters, unlike for pitchers, BABIP is actually a real life skill--as in, the best BABIP players in one season tend to be the best BABIP players the next season. For pitchers, BABIP has essentially no predictive basis. For hitters it does.

Hence in real life baseball: BAVG = f(BABIP/AvK/HR) is actually exactly how baseball works. But to "calculate" how it works in any single result, you'd use the Batter's BABIP for the batter's side of the equation and the League Average BABIP for the pitcher's side.


REGARDING HANDEDNESS:

When you see no change in overall ratings when you change handedness, that doesn't really mean much. All the game cares about is the ratings. The ratings map into performance. So all the game is looking at is "what is the batter's ability at hitting LHP/RHP in this skill type. I believe the game will likely calculate overall ratings based on roughly a 70/30 weighting of handedness values, but I don't know.

And I wouldn't say you have to know LHB are better than RHB, because that isn't totally true. What you have to know is that pitchers have splits, too. And to go further, it's good to know that the RH platoon advantage--like in real life--is smaller than the LH platoon split...I've posted those around here a few times. Bottom line, in general (and making numbers up for batting averages):

LHB vs RHP = .285
RHB vs. LHP = .280
RHB vs RHP = .270
LHB vs. LHP = .265

It is, however, fair to say that as far as OOTP is concerned, a batter with 107 BABIP vs. LHP and RHP is equally skilled at hitting both handed pitcher. And, in fact, for BABIP (since in general all pitchers are the same) this will 100% be true and is probably true in real life baseball. For AvK, LHP will tend to be harder on LHB than RHB, and visa versa, so if a batter actually has 107 AVK against both, his final performance will tend to wind up worse against that platoon split because the pitchers will tend to be better. Of course, in the end, it will depend on which pitchers he faces.

REGARDING EXAMPLES:

Yes, it seems like OOTP doesn't know how to show you that a guy with horrible splits will still be really good as a platoon player. This is one of the "problems" I have with Rob Thomas, my LHB 3B who puts up 3 WAR but is listed as a 45.

REGARDING WEIGHTING OF PITCHER HANDEDNESS:

It doesn't surprise me that the game is weighting vs. RHP more than vs. LHP, and as I said above, I suspect it's 70/30 because of the raw numbers of RHP vs. LHP.

REGARDING OVERALL FINDINGS

1) Batter handedness is not factored into the overall rating.

Yes, but not really. Handedness is factored into player creation (hence the creation of separate ratings against each hand...the other way to design this would be to just give each player a base rating, and a split offset. Since you've proven that adjusting the ratings themselves alter the overall rating, handedness is included in the overall ratings.

2) The ability to hit right handed pitching is weighted more heavily than the ability to hit left handed pitching. The amount is significant!

Yes. I think that's right. This is probably the problem regarding platoon split and overall rating. The game is giving players overall ratings as if they are fulltime players and are going to see RHP twice as often as LHP.

3) For hitting LHP like a lower level player (bigs to AAA to AA, etc) the game removes points from the overall rating that is something like 5 for the first level, 10 for the second, etc. The increase "demerit" with each level makes sense in a relative ratings environment where the game is concerned how CLOSE you are to the median.
4) For hitting RHP like a lower level player (bigs to AAA, etc etc) the effect on the overall ratings seems about double that of the verus LHP. I.E something like 10, then another 20, etc

I could be wrong, but I think these last two are just quantifiers of the second finding.

REGARDING RELATIVE RATINGS IN MINORS

I know people like them, and to a small extent I suppose they can act as an indicator a guy is ready to go to the next level, but I'll admit that I pretty much never use them. I like your fiddling with them backward, though. Bottom line, Medina has no career if he can't platoon.

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


Anyway, yeah, this is a good study...and the basic take away is that there are a lot of very useful players sitting at the 20-45 ratings in our league if you can find ways to platoon them. This is effectively why I never really cared much for stars, either.

After taking a little while to look at how the game is working, if you have ideas on how to better represent players I think the dev team would be interested in thinking about it.

In the meantime, most of us would probably benefit from trying to adjust our idea of what a good player is from, say 45-50 down to be willing to consider that a 30-35 is playable, and sometimes even a 20 could provide very good service.

[Ted]

REGARDING BABIP/CON/AVK

I don't really expect CON to be a skill. It's just annoying that it's listed with things that are a skill. The reasons I point things out holding CON constant is to make the point that low AVK, high bapib players are much more babip dependent in driving their batting average from year to year. With BABIP being more variable than AVK, the more babip dependent guys are just going to vary more year to year. We're just going to have to disagree about BABIP being a skill. I think it's a result of things like speed, swing type, power, etc. I hate having it as in input on the same level as power and eye and avk (which is really con). I get that it's unavoidable with how the game engine works, but I don't like it.

REGARDING my comments about the 66/33 and 70/30 weighting of RHB versus LHB and not changing the overall rating dependent on batter handedness.
- This isn't really a critique so much has me trying to point out that it's important to know this, and it makes the overall rating much less meaningful with big platoon splits. It would be exceptionally difficult to the engine to factor these things in, but I think there's a LOT to be gained from know that a low overall rating player with huge splits could be a LOT better based on their handedness.

And yup #'s 3 and 4 are me just trying to quantify #2 a bit.

I really don't have a better idea. But a lot of people say, "the star ratings are junk". Or the 20-80 stuff are junk. They definitely have problems, and I think it's helpful to know what's going on so maybe you can know a bit better how to interpret them.

[RonCo]

I'll add a little more later. But bottom line I'd that this is really nice work that to me shows that the overall rating is essentially base completely on the idea of showing or projecting how the game expects the player would be as a full time player.

Which to me is a pretty major point.

[Ted]

I'll add a little more later. But bottom line I'd that this is really nice work that to me shows that the overall rating is essentially base completely on the idea of showing or projecting how the game expects the player would be as a full time player.

Which to me is a pretty major point.

Yup, I think that's the big thing. That and the degree to which it differs from platooning.

[RonCo]

Regarding BABIP as a "Skill":

We've been through these discussions a few times. If we were to be in a conference room with a white board together we could have a great time discussing how to design a baseball game, and if we could manage to create algorithms that worked on a per pitch basis, I would 100% agree that BABIP has a direct relationship to speed. Several other factors affect BABIP, too, such as line drive percentage (or exit velocity and launch angle, which weren't even around when OOTP was designed).

As we've discussed before, though, the game works on a plate appearance basis. This is actually a profound statement, and if you get your brain fully around it, you'll at least understand why we can say BABIP, in OOTP speak, is a skill in the same way we can say POWER is a skill.

In real life, we might say POWER was related to several things...bat speed, or swing plane, which again leads to exit velocity and whatnot. We could have a great time together designing that kind of thing, and we would come to agreement. But in OOTP, because it resolves events on the basis of plate appearances, POWER is "merely" HR/AB...which we can call a skill because that ratio is provable to follow players from year to year. In the same way, BABIP is a definable skill because it, too, follows players from year to year in a statistically significant fashion.

In OOTP, however, I should note that BABIP _is_ correlated to Speed. That correlation happens in the player creation process, though. Players with higher speed ratings tend to have higher BABIP ratings. Just like players with speed tend to have higher range values. This is no guarantee. Speed in real baseball does not guarantee you a high BABIP guy. There are actually several quite elegant correlations between the OOTP ratings that are done during player creation. All of these are done because statistically significant relationships exist between real baseball player's skills. There are, for example, ties between HR-Rate and K-Rate for hitters because--on the whole--there exists a correlation between these "skills" in real baseball players. Again, this is not a guarantee. You can find high speed, low BABIP players in real life and in OOTP. You can find high power, low K players in OOTP and in real life. But taken in the whole, there are correlations between these "skills" (stat outputs) that mirror real baseball players.

So, yes, I agree with your supposition that BABIP is related to speed. Outfield range is also related to speed. OOTP does this, too, but does it in player creation rather than as part of its results engine.

So, yes, BABIP, when looked at on a Plate Appearance basis definitely needs to be considered a "skill" (defined as something with a stand alone rating).

Aside, as Stephen Shaw can attest to, back in the day I actually spent several months trying to write my own baseball sim that resolved results on a pitch-by-pitch basis rather than a plate appearance basis. It is a massively complex undertaking...many orders of magnitude bigger than doing it on a per plate appearance basis--which is complex in itself.

When I finally gave up, I had created a workable pitch process that rated individual pitches and used heat diagrams to locate pitch results by count and several other factors. I had also designed how defense might work and begun to develop a state engine to create the results of a ball in play. I then realized there were not enough hours in the day for one person to do this and gave up. You may be smarter than me, but I just couldn't do it.

Playing with the pitcher was fun, though. I probably spent ten hours with a couple test pitchers and gathering ball/strike data.

[RonCo]

REGARDING my comments about the 66/33 and 70/30 weighting of RHB versus LHB and not changing the overall rating dependent on batter handedness.
- This isn't really a critique so much has me trying to point out that it's important to know this, and it makes the overall rating much less meaningful with big platoon splits. It would be exceptionally difficult to the engine to factor these things in, but I think there's a LOT to be gained from know that a low overall rating player with huge splits could be a LOT better based on their handedness.

Yes. I just disagreed with the assessment that OOTP doesn't bake handedness into the overall rating. It does it in Player Creation. So in that sense, overall rating seems to be good there.

I wouldn't say a LHP with splits is any better than a RHB with splits as long as they are used in their proper platoon fashion...though a RHB in the platoon is less valuable than the LHB just due to playing time.Overall rating isn't, to me, about defining "value" so much as it is skill level. An 80 1B is probably less valuable than an 80 CF, though are both obviously rated 80 on a skill basis. Maybe I'm splitting hairs there. I'm pretty sure we agree on most of this in the end.

[RonCo]

The question of what an overall rating is supposed to do seems like it should be quite simple to answer until you're the guy who has to write the code. Over the years I've watched a lot of people try to tackle it with what I'll call variable levels of success. Ultimately, I suppose it's great to have them as some kind of shorthand, but the final answer to me always seems to be that the only way to get a real read on what a player can really do is to ignore the overall ratings and focus on the components and their splits.

[RonCo]

REGARDING BABIP/CON/AVK

The reasons I point things out holding CON constant is to make the point that low AVK, high bapib players are much more babip dependent in driving their batting average from year to year. With BABIP being more variable than AVK, the more babip dependent guys are just going to vary more year to year.

Going back over this, I should have commented here.

This isn't really right. You're better off thinking that, unlike a pitcher's BABIP, a hitter's BABIP is fairly consistent year-to-year, with the exception that the aging curve will eventually hit it. This is essentially true in real life and in OOTP.

[Ted]

REGARDING BABIP/CON/AVK

The reasons I point things out holding CON constant is to make the point that low AVK, high bapib players are much more babip dependent in driving their batting average from year to year. With BABIP being more variable than AVK, the more babip dependent guys are just going to vary more year to year.

Going back over this, I should have commented here.

This isn't really right. You're better off thinking that, unlike a pitcher's BABIP, a hitter's BABIP is fairly consistent year-to-year, with the exception that the aging curve will eventually hit it. This is essentially true in real life and in OOTP.

I'm not sure why we are seeing this differently. Every bit of data I can find shows that the year to year BABIP correlation for MLB players is quite poor. It's actually worse than batting average in some studies, and worse than ERA in most (just to pick another highly variable stat). It is nowhere close to as consistent as things like K%, BB%, swing rate, etc.

Giving that OOTP is using it as an input, I suppose it could be less variable here, but my experience is that BABIP also varies in OOTP.

Where we may be differences in our interpretation of variable. It is the case that some players run higher or lower than average BABIPs based on their skill set and will continue to do so (when not impacted by aging injury, etc), but they still vary considerably from year to year in their own personal range.

This intuitively makes sense as well, as BABIP is essentially batted ball "luck". What that luck really is, is the combination of park factors, weather, different defenders, etc that takes a sample size literally in the thousands of plate appearances to stabilize.

So then the point I'm trying to make is that in OOTP, for a player who's CON (which is really a batting average projection) is being driven more by BABIP than AVK, the year to year variability of BABIP is greater than AVK, and therefore their batting average and performance will be. (This is outside your very good analysis of how CON predicts average very poorly for low AVK players in a high STUFF environment, which is another topic entirely, but a very good one and interesting one)

I will concede that I'm assuming OOTP is modeling BABIP in a reasonable way with a reasonable amount of year to year variability, and that AVK has a relatively stable performance (like swing rate and o-swing and z-swing, and contact % do in real life). I do have not actually studied this in OOTP to be certain, but it "feels" like that is the case.

[RonCo]

I do think we're looking at variability a bit differently. And it makes sense that BABIP as a result will vary more than K% or BB% as a result will for hitters, because there are more elements involved.

But when we use stats results as a direct replacement for hitter ratings, we're missing the idean behind what it means to model the interaction between hitter and pitcher.

This goes to your questions regarding the results engine, so if you think that through, you'll more easily see what I mean.

But...

- POWER is an OOTP skill for hitters because HR/AB has a statistically significant correlation from year to year.
- AVK also has a statistically significant correlation from year to year.
- BABIP also has a statistically significant correlation from year to year for hitters. The _strength_ of that correlation is less than others, probably because there are so many other influences. But it is still statistically significant.

[sjshaw]

This is good stuff

[Ted]

This is good stuff

I really don't know if there is much real utility. I decided I don't care enough to try to come up with reliable scaling, because I don't think it can be done. There's too much variance in the way we handle our minors. Also, controlling for all the effects that mess with the other overall ratings is tough as well.

But it is neat to see how a split affects the overall rating.

[RonCo]

Yes, to simplify, though, the major point is that overall ratings assume players as full-timers. This is valuable for every GM to know, and should be considered basic baseball 101 in OOTP speak.

I'll have to think about how this sits, but it seems to me there could be a scaling for players who have big splits. In other words, something like:

For RHB:
Normal Splits: assume he faces 70/30 RHP/LHP
Moderate Splits: assume he faces 50/50 RHP/LHP
Huge Splits: assume he faces 30/70 RHP/LHP

[RonCo]

Either way, though, us GMs need to be able to scale something. In the current method, we need to be able to scale overall ratings of platoon players up in our heads. If the overall ratings were so adjusted to show two players at, say, "65" the GM would need to be able adjust his expectations that the full-timer is still more valuable than the platoon guy.

There's no such thing as a free lunch here.

Of course, if overall ratings were just ignored, and everyone just looked at components and did their own assessments, that would be dandy, too.

[RonCo]

Since I stumbled across some actual data, I'll follow up here on my point on handedness/platoon advantage and the idea that left handed hitters are just "better" than right handed hitters. This is neither right nor wrong. In context of the game itself, it totally depends on how you use hitters as to how you will assess their values. This is because splits for RHB are much smaller than the splits for LHB (something Kyle and I discussed several times).

I noted this in my first post on this thread, but in going over olld stuff I came across 2034 BBA data I'd pulled. Bottom line, the best thing you can have (all things being equal) is a LHB v RHP. But that is heavily offset by the fact that the very worst thing you can have is a LHB v LHP (especially a LHB with big splits against a LHP without such).

I agree that GMs should know these things, but I'd say if they know these things they wouldn't make blanket statements that "LHB are better than RHB" (though Ted qualifies it correctly as the case if you use them in their platoon advantage, some will not really remember or even include that qualifier).

Here were various splits in the BBA during 2034:


AVERAGE:

LHBvRHP: .271
RHBvLHP: .264
RHBvRHP: .257
LHBvLHP: .242

HR/PA:

LHBvRHP: .035
RHBvLHP: .030
RHBvRHP: .029
LHBvLHP: .026

K/PA:

LHBvRHP: .167
RHBvLHP: .176
RHBvRHP: .193
LHBvLHP: .220

BB/PA:

LHBvRHP: .093
RHBvLHP: .091
RHBvRHP: .069
LHBvLHP: .069

[sjshaw]

Bottom line, the best thing you can have (all things being equal) is a LHB v LHP. But that is heavily offset by the fact that the very worst thing you can have is a LHB v LHP

You mean LHB v RHP for the former?

[RonCo]

Yep. I'l have to fix it.

[crobillard]

You two need to teach a OOTP baseball statistics education class for dumb dumbs like me. I like the green and blue numbers.