Quote:
Originally Posted by luckymann
Hey bud, I suspect there might be a bug with the BoSox historical minors timeline. I'm still getting the Oakland Oaks as an affiliate in 1953.
They are listed on the csv as coming in for 1938 but I can't find any record at all of them ever even being a Sox affiliate, and certainly not still one into the 1950s.
They were briefly a Yankees AA affiliate in the 1930s, but never part of the Boston system according to BBRef.
Might be worth further investigation.
Cheers
G |
Two issues here:
1a)
Why is Oakland included as an affiliate of the Red Sox in 1938, as per the milbteams.csv file? No clue, but that's certainly what's in that csv file, as you also determined. I found an milbteams.csv file from OOTP 21 and it includes Oak as a 1938 Boston affiliate. I can't get to my files from earlier versions right now, but I'll bet that if/when I find them they would show the same thing. So it's a legacy issue either from an inaccurate source or simply a typo in the milbteams.csv file.
1b)
Is there any chance that some source counted Oakland as a Red Sox affiliate? Highly unlikely. Why? There is no single - and more important - definitive source on MLB > minor league affiliates. The two best sources (that I know of), baseball-reference.com and statscrew.com, don't always agree*. But they do agree in that Oakland was never a Red Sox affiliate. Back in those days it was no uncommon for an MLB and Milb team to have an unofficial agreement that would not be counted by b-ref or statscrew as an affiliate relationship. However, those can be apparent simply by looking at player movement for that season. The clincher for this issue? There were exactly zero players who played for both the Oaks and the Red Sox in 1938.
Conclusion? Typo or some other error that is easily fixable. Whoever is maintaining the milbteams.csv file can make that simple change, and then Lukas/Erik would need to make sure that update is compiled into the program that initializes a new game (milbteams.csv, milbleagues.csv, and era_stats_minors.txt - perhaps other files - are not referenced by OOTP at game creation; those files are only used for subsequent seasons).
2)
Why is Oakland still shown in your game as an affiliate into the 1950's, when the milbteams.csv file only has them as an affil for 1938? This is a definite bug that I forgot about but *does* sound familiar. What happens is that OOTP - despite what is indicated in the milbteams.csv file, does not change milb teams from having an affiliate one year to IND status the next year.
For example, I started a test with the 1937 season and Oakland is affiliate of the Yankees, as expected & as per the csv file and the two primary sources. I moved the game forward to 1938 and Oakland changes to an affil of the Red Sox, as we would expect given the incorrect data in the csv. I then moved forward to 1939 and Oakland *remains* as a Red Sox affiliate, despite the csv indicating that Oakland should change to an independent. I then searched for other teams that have a similar scenario: Minneapolis in the AA and Syracuse in the IL were both affiliated in 1938 (with BOS and CIN, respectively) and then become independent in 1939 (as per history and the csv). However, once I advanced my test game from 1938 to 1939, both teams are still affiliated with those same MLB teams.
Again, I had forgotten about this one, including that I had actually reported this issue back in March of last year. I will first confirm that it is still an issue in OOTP 27 beta and then, assuming so, I'll post an update to the bug report to hopefully get it on the radar
.
*There is no single, definitive, official source for historical minor league affiliations, especially in the pre-farm-system era and early decades of organized affiliations. What exists today is a reconstruction drawn from guides, league records, newspapers, and later research, which is why different modern databases sometimes disagree on specific relationships.
The historical affiliation data on Baseball-Reference.com comes primarily from the Seamheads Minor League Database, a research-driven compilation built from sources like The Sporting News guides, Spalding/Reach guides, league publications, SABR research, and contemporary newspapers. By contrast, StatsCrew.com relies on Chadwick/Retrosheet data for player and statistical records (as does B-R.com) but its affiliation listings are compiled separately from a mix of printed guides (such as the Encyclopedia of Minor League Baseball and Sporting News registers) and its own editorial compilation.
Because these two sites draw affiliation histories from different underlying sources — and apply different editorial standards in interpreting early “working agreements” versus formal farm affiliations — their listings don’t always match, especially in the 1920s–40s when relationships between major and minor league clubs were often informal or loosely documented.