Settle
Sign in

For collecting societies

Pay the works that were played, in the sets nobody monitored.

Only a few percent of licensed club and festival sets are monitored. For the rest, a society estimates. Settle estimates per work, from public tracklists, and shows why every share goes where it goes.

Sign in

Settle, predicted blind

18%

Of the fee to the works that were played

The season's chart alone

1,5%

What a society has without a tracklist

Settle against the chart

11,7×

More of the fee to the right work

Scored blind on 646 sets by artists with public history, each hidden in turn: at least three other live sets within six months of the set. A first measurement on community tracklists, not yet a basis for a distribution decision: validation on monitored sets comes first.

Adam Beyer, 23 October 2025. Nobody heard this set.

Settle predicted it from other public tracklists alone, without its own. The ten most likely works, then what the published tracklist says was played.

Show what was played
# Work Probability

Of the 30 works played, the prediction named 11.Every row opens to the sets behind it.

The whole case

Public tracklists, graded

Tracklists from several sources are matched to recordings and works. Each source carries a grade: monitored, released, published or community.

A probability per work

For an artist and a date, the model weighs what the artist played in the months around that date and the season's chart.

A fee split you can explain

The fee goes per work to the most likely works, writers and labels receiving what their works receive. Every share carries its probability, the sets behind it and their grade.

What it is not

Settle never listens to a set: audio recognition and player data at the venue do that, and a monitored set is the strongest evidence it takes. It is statistics a person can follow line by line, not a black box.

Settle

Royalties for the sets nobody monitored.