Historical database issues go here!

[actionjackson]

Two major things to report:

1) In the Fielding.csv file, 2021 Catchers have SB in the WP column, and CS in the SB column. Shifting the numbers one column to the right will correct what will look very strange when looking at a players career stats, and may affect each catcher's abilities in game. Every catcher will have 0 CS, much less SB than normal, and a metric bleep-tonne of WP, which Garlon has said will only be going on the pitcher's record this year. Just move the two values one column to the right, and...Problem solved.

2a) Some pitchers in 2019, and all pitchers from 2020 forward have no pitch ratings in the Master.csv file. This is very problematic because it means that any SP in those years will import (I've tested it repeatedly) with just 2 pitches randomly selected by the game. This means they can only ever be relievers. Not cool. Usually, there are numbers from 1 to 7 (most amount of pitches I can remember a pitcher having), representing the frequency that a pitcher throws his Fastball, Changeup, Curveball, Slider, Sinker, Splitter, Forkball, Cutter, Knuckleball, Knuckle Curveball, Circle Change, and Screwball. That's trouble from a reality standpoint.

2b) Also, a number of pitchers seem to have these ratings running according to the perceived quality of the pitch (seems to be running from 5 to 1, for excellent to poor, or something). So, a really good pitcher could have four "5" pitches. If all pitches are given the same rating, like four "5" pitches, they will all be given the same quality in the game. That's not how it's supposed to be done, according to Garlon. When all this was redone for OOTPX, "The Neyer/James Pitchers Guide to Pitchers" was used to rank pitches from 1 to 7ish (or more) by frequency, not by quality.

[M's rule]

Originally Posted by actionjackson

Two major things to report:

1) In the Fielding.csv file, 2021 Catchers have SB in the WP column, and CS in the SB column. Shifting the numbers one column to the right will correct what will look very strange when looking at a players career stats, and may affect each catcher's abilities in game. Every catcher will have 0 CS, much less SB than normal, and a metric bleep-tonne of WP, which Garlon has said will only be going on the pitcher's record this year. Just move the two values one column to the right, and...Problem solved.

2a) Some pitchers in 2019, and all pitchers from 2020 forward have no pitch ratings in the Master.csv file. This is very problematic because it means that any SP in those years will import (I've tested it repeatedly) with just 2 pitches randomly selected by the game. This means they can only ever be relievers.

AJ,

I did a bit of testing with a couple of 2021 replay leagues. In league #1, using the current historical db, the season totals for catchers in SBA and RTO are in line with their actual stats.

Before creating league #2, I modified the Fielding CSV as you recommended and generated a new historical db. The season totals are not remarkably different from league #1, and again generally in accordance with how the catchers performed IRL.

To find out more about problem 2a, I looked at a couple of starting pitchers who made their debuts in 2021, Alek Manoah and Reid Detmers. I don't see the same results you're getting. Both of them import as SPs.

Manoah has four pitches: fastball 152/107 (current/potential), cutter 168/118, forkball 107/75, and changeup 18/13.

Detmers has four pitches: fastball 156/55, forkball 156/55, slider 151/53, and changeup 140/49.

The weirdest thing to me about this output is the vast difference between current and potential, plus the fact that potential is the smaller value. This is not the case with all pitchers. Others look pretty realistic, such as Kershaw. On the other hand, Corey Kluber has four pitches that all have the same values, which should be impossible. Are you seeing anything like this in your 2021 leagues?

[actionjackson]

Originally Posted by M's rule

AJ,
I did a bit of testing with a couple of 2021 replay leagues. In league #1, using the current historical db, the season totals for catchers in SBA and RTO are in line with their actual stats.

Before creating league #2, I modified the Fielding CSV as you recommended and generated a new historical db. The season totals are not remarkably different from league #1, and again generally in accordance with how the catchers performed IRL.

To find out more about problem 2a, I looked at a couple of starting pitchers who made their debuts in 2021, Alek Manoah and Reid Detmers. I don't see the same results you're getting. Both of them import as SPs.

Manoah has four pitches: fastball 152/107 (current/potential), cutter 168/118, forkball 107/75, and changeup 18/13.

Detmers has four pitches: fastball 156/55, forkball 156/55, slider 151/53, and changeup 140/49.

The weirdest thing to me about this output is the vast difference between current and potential, plus the fact that potential is the smaller value. This is not the case with all pitchers. Others look pretty realistic, such as Kershaw. On the other hand, Corey Kluber has four pitches that all have the same values, which should be impossible. Are you seeing anything like this in your 2021 leagues?

There's a reason Kluber has four pitches with the same value. The values for individual pitches are contained in the Master.csv file. Kluber gets a "4" for each of fastball (4-seam), changeup, curveball, and slider. Therefore, he will show up in game, with these pitches having the exact same value. If his were done right, he would have: 1 Sinker (or 2-seam fastball), 2 Cutter, 3 Slider, 4 Fastball, and 5 Changeup. These rankings are based on how often he's thrown these pitches across his career from most often to least often, excluding pitches that are below 5%. This is how it used to be done, and somewhere along the way, somebody decided to rank them by quality (or perhaps perceived quality). In recent RL years, Kluber's ditched the 4-seamer, and kept the other four going.

As for Manoah, and Detmers importing as SP, I think (don't quote me on this), Markus set some minimum innings or GS thresholds that pitchers had to reach IRL, in order to be considered starters. Could be totally wrong on that though. The Master.csv file does not have individual pitch ratings for guys who debuted in 2020 or 2021. Between that, and not distinguishing between the different pitches on some of the elder statesmen, it's a wee bit borked.

I'm using an OOTP21 Random Debut game (that I'm very attached to) to import Manoah, and others, and the 2020 and 2021 debut guys come in with two pitches, seemingly picked at random by the game. I'm updating all the historical database.odb stuff now to get 2020, and 2021 debut guys in, but there's gonna be some work with them, as well as other historical pitchers, due to the individual pitch rating snafu that we've discussed here. I've been wrestling with the Master.csv file, and imported Logan Gilbert last night, after I'd fixed him up.

Can't remember the exact pitch ratings, as I deleted him, so that he can show up in the random ammy draft when he shows up. I ranked the pitches in the order that I got from Pitch Info Pitch Type on fangraphs, which would've given me 1 Fastball, 2 Slider, 3 Curveball, 4 Changeup. He was a monster, but I think that's because my game uses 1984 for its stats output. Back in that season, the MLB WHIP was 1.34, H/9 was 8.9, HR/9 was 0.8, BB/9 was 3.2, K/9 was 5.4, and K/BB was 1.69. Gilbert's 2021 numbers were 1.17, 8.4, 1.3, 2.1, 9.7, and 4.57 in those categories, so he came in beastly, but the environment will tamp down his abilities, when he comes in through the ammy draft.

As an example of the tamping down effect, Verlander's put up a 256-150, 2.86 line in my game, with a 1.09 WHIP, 7.5 H/9, 0.6 HR/9, 2.4 BB/9, 7.3 K/9, 3.1 K/BB, 141 ERA+, and a 3.01 FIP across 516 GS, and 3955 IP. He's running out of gas, heading into his age 39 season though, as his RL numbers ended with his age 36 season, and the game is not kind (to say the least) to older pitchers.

RL Verlander:

236-132, 3.29, a 1.12 WHIP, 7.6 H/9, 0.9 HR/9, 2.5 BB/9, 9.1 K/9, 3.6 K/BB, 130 ERA+, and a 3.41 FIP across 469 GS, and 3085.1 IP. The lower R/G environment of 1984 brings his ERA down, that season's far lower HR/9 environment pulls his HR/9 down, and the far lower K/9 rate of that year, pulls his K/9 down. Both great either way. He's also gone deeper into games in my game, averaging 7.66 IP/GS, vs 6.58 IP/GS IRL. A real HoF horse in either version.

[actionjackson]

I also didn't get the far lower potential than current in the pitch ratings for Gilbert. He had higher potential than current ratings for all four of his pitches.

[actionjackson]

I'm fine with Manoah having very low stamina (currently 1 out of 250), despite being a huge fanboy. He'd only pitched 111.2 innings through 2021. After importing 2022 stats from OOTP24, though? Yeah, yer darn right I'll manipulate the Stamina setting like a do for every SP with at least 200 innings under his belt. I think Markus' thresholds are higher than mine. Mine are 200 innings, and a third or more of his appearances as a starter.

Eck sneaks in under the starter appearances wire (33.71% versus 33.33%), and I think he should, but I'd give him a reasonable Stamina rating. My method would give him 68 out of 250 for Stamina as a starting point, which I think is more than fair. People forget how good he was, as a starter, over the first 12 years of his career, because they only remember his sheer dominance as a closer, over his final twelve years. Just brought him in as a 19 year old, and the game gave him a 47 Stamina rating, so I don't think my 68 is outlandish.

[actionjackson]

After my Master.csv file manipulations, my Manoah has a 180/187 Fastball, a 153/159 Slider, a 116/121 Sinker, and a 72/75 Changeup. Not sure where the powers that be came up with a Cutter, and a FK for him, because he hasn't ever thrown either of those pitches, but that aspect is a cosmetic one that shouldn't affect results. The main thing is that I've given him four pitches, which is what he should have.

The four pitches that I've given him are Fastball (37.4% usage IRL), Slider (28.1%), Sinker (25.1%), and Changeup (9.4%). He's never thrown anything else, so that's what he gets from me. I never give a guy a pitch when he throws it less than 5% of the time, because if he's a starter, he'll only throw, at most 4 or 5 a game, and he probably doesn't have much confidence in it. If he's a reliever, he might not throw it at all. Less than 5% also seems to be the cutoff for getting into a guy's repertoire over at fangraphs, so I must be doing something right.

Remember too, that this is a Historical Random Debut game, based in OOTP21, that started in 1901, and is now through 1935, using 1984 as a base for league stats output. I don't know if that's what's creating the difference in what we're seeing, but it might be. As I said above, I've copy/pasted all the Master.csv file stuff, and Batting.csv, Fielding.csv, and Pitching.csv stats from OOTP22 (through 2020 season), and OOTP23 (through 2021 season) into said files. All in all, the players look really good, when I import them individually, so that's a pat on the back for me.

Still have to fix up pitching repertoires, like Kluber's (and many others) in the Master.csv, and check out 19th century biographical features (particularly birthdate, Bats, Throws, Height, and Weight), which have been dug up in recent years, but not (all) updated in the Master file. Also (against the advice of Garlon and many others), adding in all the Fed League guys. A little bit anal, but I think it'll look really good when it's done, if I do say so myself.

Doing this also means that I won't be left high and dry if Markus decides he's had enough with this game, and its demanding customer base. That's always a plus. To be able to continue a game with whatever the final version of the game winds up being with up to date historical players will be a great insurance policy. And then to be able to start another Historical Random Debut once I've gotten to 2050ish? Wicked awesome.

Slightly morbid, but it'd be like when your favourite artist/band etc leaves this plain of existence. It really sucks to know that you won't be getting any new material (or new game versions), but you'll still have their music (the basic framework of the final version of the game) to keep you going.

[actionjackson]

Detmers comes in as a 20 year old (always bring them in a year early because that's what the game does in an RD). He has a 170/174 Fastball, a 144/148 Curveball, a 111/114 Slider, and a 68/70 Changeup. Not bad for a dude who got shelled, and went 1-3, 7.40 in his first go around, in 2021. 132/135 Stuff, 74/84 Movement, and 80/87 Control for him.

You know Manoah's Pitch Ratings from above. He comes in as a 22 year old, and he has 188/194 Stuff, 152/157 Movement, and 95/98 Control. So, he shows up quite a bit better than Detmers, but that's because of their vastly different 2021 seasons. Hopefully, Manoah doesn't show up before OOTP24 comes out, because he'll stay a reliever due to his Stamina Rating, if he does. Unfortunately, Jordan Romano has already shown up, meaning his stats only run through 2019, meaning he sucks for the for the rest of his time in this game. Win some/Lose some. None of the other Blue Jays' cornerstones have shown up yet, so that's good.

[LansdowneSt]

I think this player's name should be Osmer Morales, not "Osmel"

https://www.baseball-reference.com/p...oralos01.shtml

https://www.mlb.com/player/osmer-morales-591672

[RGTML]

Like the title says and attachment shows, the draft pool is empty as soon as it hits the 2021 season. Could someone give me a fix? Spent quite a bit of time on this file.

[LansdowneSt]

Originally Posted by RGTML

Like the title says and attachment shows, the draft pool is empty as soon as it hits the 2021 season. Could someone give me a fix? Spent quite a bit of time on this file.

Can you post a screenshot of this screen in your game? Note the right-hand side which has the draft date and the date the pool is revealed.

[RGTML]

Here you go

Edit: Also, the drafts have been weird so far. They've been by which players played their first MLB games that season rather than which year they were historically drafted. I guess that helps with international free agents but I think it's what cause this issue.

[LansdowneSt]

Originally Posted by RGTML

Here you go

Edit: Also, the drafts have been weird so far. They've been by which players played their first MLB games that season rather than which year they were historically drafted. I guess that helps with international free agents but I think it's what cause this issue.

Go back to August 17, the date of the Draft Pool release. Was there a blank email or did it not come? It may be that you deleted it and we can't check that though.

Try moving the pool reveal date closer in so it is in front of you and see if it gets a chance to regenerate. You can reveal as close in as 5 days out from the draft but whatever the next milestone is for reveal is what I'd try.

Usually, my rookies appear at or about their debut age. Depends on when in the year my draft is. But maybe I just always have some option toggled that I don't pay mind to.

[David Watts]

Originally Posted by LansdowneSt

Go back to August 17, the date of the Draft Pool release. Was there a blank email or did it not come? It may be that you deleted it and we can't check that though.

Try moving the pool reveal date closer in so it is in front of you and see if it gets a chance to regenerate. You can reveal as close in as 5 days out from the draft but whatever the next milestone is for reveal is what I'd try.

Usually, my rookies appear at or about their debut age. Depends on when in the year my draft is. But maybe I just always have some option toggled that I don't pay mind to.

I'm really confused. If one is playing historical wouldn't things historical end following the 21 season? That's the final season that's actually been played and finished. I would think the only way one would get players is to go to the historical menu and unselect the option to import real players/rookies. At this point the game would start creating fictional players for the draft.

[LansdowneSt]

Originally Posted by David Watts

I'm really confused. If one is playing historical wouldn't things historical end following the 21 season? That's the final season that's actually been played and finished. I would think the only way one would get players is to go to the historical menu and unselect the option to import real players/rookies. At this point the game would start creating fictional players for the draft.

You, sir. Are exactly, 100% correct. That's the issue. [Slaps hand to forehead]

But do also set the reveal date in front of you as it now needs to generate those players.

[RGTML]

I tried that and the draft pool is still showing up empty

[David Watts]

Originally Posted by RGTML

I tried that and the draft pool is still showing up empty

Did you uncheck the box for real rookies and then go to the rules menu and set the number of rounds you want and also the how many rounds worth of players you want the game to create?

[LansdowneSt]

Jack Kruger was in the majors for a single game in 2021 but OOTP doesn't show it. Was one game, no PA.

https://www.baseball-reference.com/p...rugeja01.shtml

[RGTML]

I mean, wouldn't real rookies be the logical real step? Guys that were drafted in 2021 and then 2022? Or guys that were drafted earlier but haven't played an MLB game yet?

[LansdowneSt]

Originally Posted by RGTML

I mean, wouldn't real rookies be the logical real step? Guys that were drafted in 2021 and then 2022? Or guys that were drafted earlier but haven't played an MLB game yet?

I reckon that the historic database ends when history does. From the game's vantage point on release, everything after 1 April-ish is yet to be written. Next year 2022 will be pulled into the historic database. Maintaining an up-to-date database with draft eligible college and high-school players would be a monumental task and there isn't nearly time in the day to make that possible.

[weeqat]

Originally Posted by weeqat

Not sure if this is a database issue per se, but it relates to csv player import so I'll post here. I've got a list of 206 Negro League players which seems to import into different test leagues that I've set up with hugely different results. In some import scenarios all of the pitchers will come in as 20-20 RPs, because they only have a 40-grade fastball and no other pitches. Other times, they will import with a full arsenal of pitches and starter stamina, etc. Same list, same import year, nothing changes. Here are some of the variables that I've tried to isolate, but I can't seem to find a consistent process.

2 leagues are set up, both custom/fictional, but with the same 28 real teams created. In one, I set up a random debut league, deleted all the players and then imported full majors and minors systems for all those specific teams (eg 1969 Orioles, 2004 Red Sox) via csv. In the other, I deleted all the minor league levels and imported the same teams (which gave me 30-50 "real" players spread between active roster and reserve pool). I then added minor league levels and imported all of the "missing" players from the above-mentioned lists.

(As a side note, the first setup method above created some EXTREMELY weird, low ratings for a lot of players - very good teams might have their whole real life bullpen filled with 20-30 POT players. This was using normalized stats so I'm not sure what happened there. The 2nd league was set up in order to fix this specific problem, and seemed to have provided a better range of ratings).

Back to the NeL import. First phase was to import about 1200 different player-years in order to find the ones that the game "liked" best. I used normalized stats, with adjust/weaken set to 200/40 for hitters and 30/10 for pitchers (this will be consistent for all import attempts). I narrowed this down to 206 players.

First league: I first imported the list into the FA pool with the above settings, and all of the pitchers were 20/20, with a single pitch. I deleted them and then, in the same game session, I imported them into the draft pool instead - and I got solid ratings and results.

In the second league, I imported them into the draft pool but had the 20/20 issue. I deleted them and then tried importing them into the FA pool instead, and the results were good this time!

If I try to import any of the players singly, using their id and year, they come in as 20/20. Hitters are generally better but their ratings are often much lower than I get when importing the entire list.

I'm curious for any insight as to why there is so much variation in what seems to be an easily repeatable process? It seems like there is some underlying coding error causing this.

I wanted to bump this message since only a couple of people commented and I'm not sure it got seen by the people who might look into it. Short version, the historical player import function seems buggy. I've seen other posts with a similar situation, where player import of high-quality pitchers sometimes randomly creates a "default" player with 1or 2 40-grade pitches only.