Historically based batter splits

[Brad K]

I have an older version of the game that only provides random splits and have been editing historically based splits manually.

So how are the historically based splits in 19? Is it known what methodology is used?

I've found in doing my editing that some players even with adequate sample size aren't consistently better vs L or R in all categories, IOW, they might be better in contact vs R but better in power vs L. That sort of thing. Does 19 reflect that?

[Spritze]

they go back to 1913 and just are what they are. no methodology was used at all by OOTP.

[Brad K]

Are you saying that each year they plugged In The Raw data from that year? That would surprise me very much.

[Spritze]

then being surprised is the state you are in. I hope you were wearing pants, it is terrible to be surprised without pants on. usually anyway.

[eriqjaffe]

What about being surprised by pants?

[Brad K]

Here's an example of why I'd be surprised if there wasn't a method involving adjustments and raw data was plugged in.

Bobby Bonds

Year/vs Right/vs Left

2000 .340/.230
2001 .334/.312
2002 .363/.384
2003 .331/.363
2004 .395/.307

If they're doing some kind of smoothing, fine. Curious what it is. If they're using raw data, I see no reason to upgrade. I'll continue to do edits on my old version.

[Drstrangelove]

Originally Posted by Brad K

Here's an example of why I'd be surprised if there wasn't a method involving adjustments and raw data was plugged in.

Bobby Bonds

Year/vs Right/vs Left

2000 .340/.230
2001 .334/.312
2002 .363/.384
2003 .331/.363
2004 .395/.307

If they're doing some kind of smoothing, fine. Curious what it is. If they're using raw data, I see no reason to upgrade. I'll continue to do edits on my old version.

https://www.baseball-reference.com/p...&year=2000&t=b
https://www.baseball-reference.com/p...&year=2001&t=b
https://www.baseball-reference.com/p...&year=2002&t=b
https://www.baseball-reference.com/p...&year=2003&t=b
https://www.baseball-reference.com/p...&year=2004&t=b


First, I'm sure you mean Barry Bonds.

Second, those are the real numbers. Obviously, if you want to edit them, it's your game. But, people have been reconstructing batter splits and game log data since at least the 1980's by going through newspaper after newspaper, microfiche etc. across the country. Numerous numbers like Ty Cobb's hit total, Cy Young's win total, etc., etc., have been corrected by people tediously reconciling game reports, box scores and reported MLB totals. When I first found Retrosheet, it was stuck in the 1930's. Now it's deep in the deadball era.

This is highly tedious work that thousands of baseball fans have worked hard to get right, not people plugging in numbers to smooth data. The problem is that when people started looking at league, player and team data, in detail, they realized there were obvious errors. And not just errors, but missing data that people needed in order to help improve calculations of WAR, determine exactly how Ted Williams hit against lefties, how many home runs Babe Ruth would have hit in modern sized parks, etc. and other advanced metrics. (I'm certain it's been going on longer because I began using microfilm to find game log data back in the 1970's, and I wasn't any sort of pioneer in the matter.) I get it: you don't agree with the numbers.

https://www.retrosheet.org/ and https://sabr.org/ are fine sources for this process.

I can't vouch for all the data in every year for OOTP, but those Bonds numbers are exact and correct.

As for "plugging" in the data, my assumption is that OOTP is extracting data from something like a csv file, which can be easily acquired from sources like those above. Inputting is possible manually, of course, but it's not how I'd do it.

[Spritze]

It would not be historical if it was smoothed or made up.

we leave the made up stuff to the individual user. if you want Luis Aparicio to hit 80 hr's in 1958 you can do that.

[Brad K]

Originally Posted by Spritze

It would not be historical if it was smoothed or made up.

we leave the made up stuff to the individual user. if you want Luis Aparicio to hit 80 hr's in 1958 you can do that.

You're not going to get historical results by putting in data as I listed. Because of a small sample size of a single year the results can vary significantly from actual ability.

So the year Bonds hit 340 against left and 230 against right. Those are not the results you're likely to get if you plug that in as values in the game. Given that performers in the game vary significantly from the rating, in that year bonds might hit 190 against right-handers which is truly ludicrous.

It is what it is already doesn't apply to the game. Batters with low at-bats at pitchers with low Innings pitched are altered, and there are options to rate players on three years or five years with or without double weighting the current year.

So if ootp used unadjusted raw data for splits it would be contrary to what they do in other areas.

People here are arguing that ootp uses of raw data. Instead of arguing about it, could someone who uses 19 check? I don't have 19 which is why I asked the question. I would like to compare the 19 method to the method I use.

[Spritze]

Since I myself was tasked with this task I don't need to check as I know the task that was asked and what was done taskificially speaking.

feel free to alter anything you wish to..

[Spritze]

Originally Posted by Brad K

It is what it is already doesn't apply to the game. Batters with low at-bats at pitchers with low Innings pitched are altered, and there are options to rate players on three years or five years with or without double weighting the current year. these options all use unadjusted raw data as the base

So if ootp used unadjusted raw data for splits it would be contrary to what they do in other areas. all other areas also start with unadjusted raw data

I think you may have mixed up apples with rocks in your head. A rethink may be in order?

[Brad K]

So are you saying that when they say splits are historically based they are adjusted raw data?

I've never suggested making things up. (Like, for example, the prior system of randomly generated splits.)

[Brad K]

Originally Posted by Spritze

Since I myself was tasked with this task I don't need to check as I know the task that was asked and what was done taskificially speaking.

feel free to alter anything you wish to..

Feel free to let someone who might be willing to answer my question do so.

[Spritze]

Originally Posted by Brad K

So are you saying that when they say splits are historically based they are adjusted raw data? No

I've never suggested making things up. (Like, for example, the prior system of randomly generated splits.)

adjustments to the raw data (if any) are dependent on how you choose to play the game. ALL historical data is raw, major and minor leagues.

yes, someone else will need to further explain how the game choices work since you are not getting it near as I can tell.

regards

[Brad K]

Seem to be posting Barry Bonds 2000 ratings the settings based on a single season would explain a lot.