- I'm not sure if this qualifies as an error, but I noticed that
the "stint" column for managers is 0 if 1 manager lasted the entire
season. 1 is only used to imply the first of 2 or more. In the
batting, pitching, & fielding tables, however, 0s are not used. I
think it would be more useful if 0s were used in the player tables as
well (or at least they should be consistent across tables). Is there
a reason for the discrepancy?
- I don't currently have access to a good relational database. If I did, the solution for stints in Fielding.csv is easy. For all 2011 fielding stints for pitchers, update the stints to match stint in Pitching.csv for same player, year and team. For all 2011 fielding stints for non-pitchers, update the stints to match stint in Batting.csv for same player, year and team.This will work for future years except when a player has 2 separate stints with the same team in the same year. The fielding stints for those rare players will have to be handled manually.Clem Comly
- BDB or Lahman DB would be "backward compatible" with the proposed change,
or at least have that as a potential.
If we do this with a STINTS table that shows it the way we've been talking
about it, then you could join the Batting, Pitching, Fielding table to the
STINTS table, and get the stintID for virtually every record. There's no
reason that it has to be "backward compatible" out of the box. But, one
extra join on each table will make it backward compatible.
The only issue will be the Rob Ducey's of the world, who leave and come
back to the same team in the same year, while interrupted by a different
MLB team. Again, that can also be handled with a StintSplits table,
similar to have a splits table for Batting v Pitch Hand if we wanted to
eventually go there.
Anyway, this is really in Lahman's hands, as the keeper of the DB. He
says that the data quality issues was that he needed to change his script.
If that's the case, that we won't experience data quality issue again as
a result of stints, then, fine, the issue is mostly moot. Those of us who
have no need to the stintID field can do the appropriate matching and
I will say that once you incorporate minor league data, we're going to
revisit this all over again. As it currently stands, we're going to have
massive key changes when that happens. JJ Hardy getting sent down to the
minors and being called up will have stintID 1 and 3 in 2009, whereas
right now, he has only 1 record. And it'll get worse for guys that get
sent down and called up multiple times in the same year. But, we're not