Tan Ho

Hmm. It's a bit of a mix of things so not a straightforward question to answer. Is the hope to replicate it or to cite it?

Closing this as not planned. fwiw in an anti-points league I'd probably award positive points for every snap and negative points for every stat accumulated, that would up the challenge...

column naming or data types? column naming is difficult because we'd need to either support backcompat or do some sort of warn/deprecation process which wouldn't be fast. I'd broadly be...

join_coalesce is just much better with dplyr, should probably move to a different nflverse-utils pkg tbh. https://github.com/tanho63/tantastic/blob/main/R/coalesce_join.R

> > add trimws() to FTN and rebuild > > Has this been implemented? Or is it actually necessary? I don't see any difference > > ```r > p >...

Data docs is an ongoing battle! Along with dictionaries missing above, here are some missing fields for existing dictionaries: ``` R> load_schedules() |> +++ dict_check(dictionary_schedules) old | new - "alt_game_id"...

offhand I'm thinking of: - sportradar_id (legacy, but still impt for fantasy fb) - nextgen stats (i.e. gsis_it_id, I think?) - college football reference id - spotrac (if we ever...

IDs shouldn’t be too tricky aside from the usual name join BS. Contract data definitely a bit more involved. Is OTC better than spotrac to store?

100% useful!

Update: after some discussion about naming it's now actually here lol https://github.com/nflverse/nflverse-data/releases/contracts Still TODO: gsis/nflverse id -> otc id