Seasons and eras

resolve(), palette(), team_colors(), logo_url(), logo_image() and marks() all take a season.

Season conventions

Seasons follow the SportsDataverse conventions, the same as the logo archive:

• NHL, NBA, men's and women's college basketball (nhl, nba, mbb, wbb): the ending year. The 1994-95 NHL season is 1995.
• Every other league: the single or starting year. The 2010 NFL season is 2010.

For example, the Quebec Nordiques' last season was 1994-95 and the Colorado Avalanche's first was 1995-96:

import sdvplot

sdvplot.logo_url("COL", "nhl", season=1995)   # the Quebec Nordiques mark (archived for 1980-1995)
sdvplot.logo_url("COL", "nhl", season=1996)   # the first Avalanche mark (1996-1999)

A season is a year: 2010, 2010.0 or "2010". None or NaN means no season. Anything else, such as "2019-20", raises ValueError. Pass one season for every value, or a list with one season per value:

sdvplot.resolve(["OAK", "LV"], "nfl", season=[2010, 2024])   # ['13', '13']

Reused codes

Aliases can carry a season range. When you pass a season, resolve() first looks only at aliases whose range covers it, so a code that two teams used in different eras would go to the team that held it that season. Then it looks at every alias, so a code still resolves when the season is outside its range, such as a modern abbreviation on old data.

In the bundled index today, no identifier names different teams in different seasons, so season never changes which team a value resolves to. For example, nflverse's "LA" is always the Rams. Where a season does matter is the logo.

Relocations

A relocated franchise keeps one team_id. Its old codes are aliases with the seasons they were used, and its old marks are dated the same way:

League	Code	Franchise	Old mark's seasons
NFL	OAK	Raiders ("13", now LV)	1960-2019
NFL	SD	Chargers ("24", now LAC)	1961-2016
NFL	STL	Rams ("14", now LAR)	1995-2015
WNBA	SAS	Aces ("17", now LV), in San Antonio	2003-2013
WNBA	SA	Aces ("17"), in San Antonio	2014-2017
WNBA	DET	Wings ("3", now DAL), in Detroit	1998-2009
WNBA	TUL	Wings ("3"), in Tulsa	2010-2015

sdvplot.resolve(["OAK", "SD", "STL"], "nfl")              # ['13', '24', '14']
sdvplot.resolve(["SAS", "SA", "DET", "TUL"], "wnba")      # ['17', '17', '3', '3']

Today's mark, or that era's

Without a season, logo_url() and logo_image() return the team's current mark. With a season, they return the mark in use that season. The season picks the mark, not the code you passed:

sdvplot.logo_url("LV", "nfl")                # the Las Vegas mark
sdvplot.logo_url("LV", "nfl", season=2010)   # the Oakland mark
sdvplot.logo_url("OAK", "nfl", season=2010)  # the same Oakland mark
sdvplot.logo_url("OAK", "nfl", season=2024)  # the Las Vegas mark again

With a season, a mark whose dated range covers it wins over an undated (current) mark. If nothing covers the season, you get the best mark of any era. Logos has the full selection order, and the logos tutorial shows the eras side by side.