No ratings yet.

In the past, when artists released a new single or album, listeners would have to go into a physical store to get a CD to play at home. In store, customers would pay a set price, usually a similar price per album or artist based on market forces. Now, artists can release their tracks to thousands of different digital streaming platforms with the press of a button. How much money an artist sees from their streams depends heavily on which platform the listener consumes the music.

Part of Spotify’s success can be accounted for by their algorithm. Users can discover new music very easily, by listening to other users’ playlists or the personalized playlists based on previous listening behavior. On the other hand, there is Apple Music, where the core value lies in collecting music, the same as how people collected their CD’s and later songs on iTunes.

Streaming platforms’ payout structures are quite complex and dependent on various variables, but in general they use the same key features. The revenue of all the users is collected in one large pool, usually split per country to account for varying purchasing powers and related premiums users pay. The streaming services keep a percentage of the revenue to cover their operational costs, e.g. for Spotify this is capped at 30%. Then all the streams from a country are added together, and the total payout is divided by the total streams. This usually entails that each stream is worth a fraction of a cent.

For artists, this has a couple of consequences. Firstly, whereas in the past it did not matter if a fan wanted to listen to a single song only once or have the album on repeat for weeks, the revenue they received would be a one-time payout. In present day, the number of times someone streams a song determines the payout an artist receives. One stream is almost negligible but having the song or album on repeat can generate higher revenues than the sale of a CD. Additionally, it is no longer the absolute number of streams but the relative number of streams that matter for the revenues. The more people listen to music on a platform, the less each stream is valued because it is pooled. For this reason, songs are becoming shorter and shorter to accommodate this change that the digitalization of the industry has brought forward. With each track being 2,5 minutes instead of 3, artists can gain one stream extra per 15 minutes that their music is consumed. What is more, Spotify almost incentivizes users to listen to more different songs with their algorithm, and thus artists receive a lower payout as their relative share decreases. That’s partly why Apple Music can pay higher revenues to most artists as their discovery mode is more limited and fans listen to the same artists repeatedly.

Some people have expressed their concerns, fearing that the artistic value behind the music is undermined by this business model. Besides, when large artists release a new album, such as Bad Bunny generating 183 million streams on his release day, they take away a large part of the revenue from other artists that are unable to come even close to those numbers, and see their relative share grow smaller as depicted in Figure 1.

To circumvent this issue a new model has been proposed. Here, the relative share per user would be used to pay as revenue to an artist rather than the pooled model. The more a user listens to an artist’s repertoire, the larger the share of the premium of that specific user the artist would receive. For example, if a user only listens to 1 artist, that artist would receive the full premium, no matter if they have listened to a single song or to an album on repeat. However, the largest music distributors in the industry, of which 3 hold around 70% of the market share, would potentially see their revenues decline as they benefit from the pooled distribution and are naturally hesitant to adopt such a new payout structure.

Currently, not only a user’s own listening behavior, but also that of others determines who gets which share of the pie. Should users reclaim this autonomy over which artists they want to support, like they were able with CDs?

References

Mack, Z. (2019). How streaming affects the lengths of songs. Retrieved from https://www.theverge.com/2019/5/28/18642978/music-streaming-spotify-song-length-distribution-production-switched-on-pop-vergecast-interview

Musically. (2022) Majors slightly increased their market share in 2021. Retrieved from https://musically.com/2022/04/06/majors-market-share-in-2021/

Spotify. (n.d.) Audience. Retrieved from https://artists.spotify.com/artists

Spotify. (n.d.) Royalties. Retrieved from https://artists.spotify.com/help/article/royalties

Velardo, V. (2019). Spotify’s Discover Weekly explained – breaking from your music bubble or, maybe not? Retrieved from https://medium.com/the-sound-of-ai/spotifys-discover-weekly-explained-breaking-from-your-music-bubble-or-maybe-not-b506da144123

No ratings yet.

Computers have been beating humans for the past two decades when it comes to games of chess. Chess engines analyze chess variations and generate (a list of) the best possible moves. They are still widely common even though no human player can compete with them, and they keep getting stronger as computing power increases and more games are played. This resulted in additional variables being added into the engines to limit their strength, so that human players could practice against the engines. The engines can mimic human behavior by not continually selecting the optimal move, weakening their game. However, this makes the engine prone to abuse.

Recently, Hans Niemann rose to the top competing against the world’s best players. When playing against world champion and grand master Magnus Carlsen his opponent forfeited his game against the newcomer, hinting at the fact that Niemann might be cheating.  

In the most recent investigations, it has been discovered that Niemann has likely cheated over 100 times during online games between 2015 and 2020. Earlier, the player had already been caught after cheating-detection software had marked his online games with suspicious play. Algorithms consider various variables that constitute to the level of play, but also behavioral factors and datapoints about the player themselves. Moreover, the player has admitted to using an electronic device that could give him signals to find the best moves during games.

During a recent upwhirl, new accusations of cheating against Niemann came forward, which have yet to be proved. For physical (high-level) games, players are commonly screened and checked with a metal detector to establish if they are not using devices placed on their body. Therefore, in some cases, players hide phones in the bathroom to consult during a break. Nonetheless, other forms of cheating can still happen without the assistance of technology, such as signals from coaches.

Other forms of sport are also highly regulated when it comes cheating. For highly physical intensive sports, drug test can be taken at random to confirm if athletes are not cheating by taking illegal substances. Checking of technology to cheat therefore seems nothing out of the ordinary, yet there is quite a lot of uproar.

Winning by cheating takes the fun away from the sport, yet players are incentivized by money prizes and fame. Now that these ways of cheating are being uncovered, how long will it take before a new technology can help these players rise to the top?

References

Chappell, B. (2022). Chess world champion Magnus Carlsen accuses Hans Niemann of cheating. Retrieved from https://www.npr.org/2022/09/27/1125316142/chess-magnus-carlsen-hans-niemann-cheating#:~:text=As%20speculation%20swirled%2C%20Niemann%20admitted,to%20find%20the%20best%20moves

Chess.com (n.d.). Chess Engine. Retrieved from https://www.chess.com/terms/chess-engine

Chess.com (2022). About Online Chess Cheating. Retrieved from https://www.chess.com/article/view/online-chess-cheating

Doggers, P. (2022). Chess.com: ‘Niemann Has Likely Cheated In More Than 100 Online Chess Games’ retrieved from https://www.chess.com/news/view/chesscom-hans-niemann-report-cheating