Your squad

Read your squad, compare ability and season performance, and export the results.

Open the save

import fmsave

with fmsave.open("career.fm") as career_save:
    managed = career_save.managed_clubs()
    if not managed:
        raise ValueError("No managed club: choose a club from career_save.clubs()")
    my_club = managed[0]
    squad = career_save.players().where(club_uid=my_club.club_uid)

fmsave.open is a context manager. Call the readers you need inside the with block; asking a closed save for another table raises SaveClosedError. The records and tables themselves keep working after the save closes, so you can carry squad out of the block and use it for the rest of the program.

players() returns the players fmsave reads from the save. where(club_uid=...) cuts that to one club, including its youth and B teams. team_id and team_slot tell its teams apart.

Sort it

best = squad.sorted_by(lambda player: player.ability.current, reverse=True)

for player in best[:5]:
    print(player.name, player.age, player.ability.current, player.ability.potential)
    # "Alex Example" 24 148 165

A slice of a Table is a Table, and tables are immutable, so sorted_by hands back a new table and leaves the original alone.

Read a player

player = best[0]

player.attributes.finishing  # 16
player.attributes.determination  # 17
player.positions.stc  # 20
player.natural_positions  # ("STC", "AMC")
player.personality.professionalism
if player.contract is not None:
    print(player.contract.wage, player.contract.end)

attributes holds 52 attributes on the 1 to 20 display scale, from crossing and passing to handling and reflexes for a goalkeeper. positions rates the 15 position slots on the same scale, and natural_positions is the shorthand for the ones rated 18 or better. ability carries current and potential, reputation carries 4 figures, and personality carries the 8 personality attributes.

Wages and transfer values come back in the unit the save stores them in, which is not the currency the game displays. fmsave does not convert them.

Some fields are coded values: they carry both the number the save holds and the label fmsave reads it as.

if player.contract is not None:
    status = player.contract.squad_status
    if status is not None:
        print(status.label, status.raw)

A value fmsave cannot read is None rather than a guess. Before you lean on a field, check whether it is verified. See Trusting a number.

Narrow it down

squad.where(on_loan=True)
squad.filter(
    lambda player: (
        player.age is not None
        and player.age <= 21
        and player.ability.potential is not None
        and player.ability.potential >= 150
    )
)
squad.find(name="Alex Example")

• where(**fields) matches top-level fields for equality. Flat column names such as contract_wage are not field names; use filter for anything nested.

• filter(predicate) takes any function of a record.

• find(name=...) returns a table of all exact name matches, ignoring case and surrounding whitespace. It can be empty or contain several people.

Passing an enum label to a coded-value field matches every record carrying that label:

with fmsave.open("career.fm") as career_save:
    starters = career_save.contracts().where(squad_status=fmsave.SquadStatus.STAR_PLAYER)

Compare season performance

with fmsave.open("career.fm") as career_save:
    season = career_save.player_season_stats().where(
        club_uid=my_club.club_uid, kind=fmsave.SeasonStatsKind.OVERALL
    )

regulars = season.filter(lambda row: row.minutes >= 900)
ranked = regulars.sorted_by(lambda row: row.expected_goals_per_90, reverse=True)
for row in ranked[:5]:
    print(row.player_name, row.minutes, row.expected_goals_per_90, row.pass_completion_percent)

Choose one kind before comparing players: league, cup and overall rows overlap. OVERALL is the club season total; rows can also cover another team a player played for this season. Join a row’s player_uid to Player.uid when you need attributes or contract details.

Rates are calculated from the counts and included in every export. They are None when the count does not apply or the denominator is zero. The minutes filter above excludes zero-minute rows and limits comparisons to players with some playing time.

Out to pandas, CSV or JSON

frame = squad.to_pandas()  # needs fmsave[pandas]
frame[["name", "age", "ability_current", "contract_wage"]].head()

squad.write_csv("squad.csv")

The DataFrame and the CSV use flat column names: nested groups become ability_current, contract_wage, attributes_finishing. Getting data out covers the rest of the formats, the flattening rules, and doing the same job from the command line.

The rest of the club

The squad is 1 of 27 tables. The same club uid opens the others:

with fmsave.open("career.fm") as career_save:
    club_uid = career_save.managed_clubs()[0].club_uid

    staff = career_save.staff().where(club_uid=club_uid)
    finances = career_save.finances().where(club_uid=club_uid)
    injuries = career_save.injuries().where(club_uid=club_uid)

Use uid to join records within a save. To match players, staff or clubs across careers started from the same database, use unique_id when available. The game can reuse a deleted person’s ID for a new person, so check identity before treating it as a lasting match.

What a save holds lists the readers and their limits.