Dominance is defined by my best judgement of MAU's , or the best available metric. Note that the market will resolve to an ESTIMATED MARKET SHARE. In other words, it is NOT winner-take-all. Ie if X has 55% dominance, I want X to resolve to 55%. Friends have told me this is possible. If I find it is NOT possible to partially resolve a free response market, I will resolve it N/A.
This market concerns the user distribution at 2028-12-31 23:59:59. However, it resolves somewhat later to allow for data to become available.
My preference is to consider platforms, not protocols. Platforms are inherently mutually exclusive, so I would hope this all works out cleanly. I'm using the names of platforms and not protocols (ie Bluesky over AT protocol) in order to make the comparison to centralized services more apples to apples.
That being said, I need the answers to be mutually exclusive. If someone makes an ActivityPub choice, we would only count the usage that is ActivityPub but NOT Mastodon. Overlap will resolve like so:
First, in the direction of Platforms over Protocols
Second, in the direction of whatever was added earlier.
For example, consider the services added in the following order 1. Mastodon, 2. ActivityPub, and 3. Misskey. Mastodon and Misskey are based on the ActivityPub protocol. I would put users buckets in the order 1,3,2 -- preferring platforms over protocols and prioritizing history of adds. The history consideration is just to prevent a sort of exploitation if we have some kind of Russian nesting doll situation with multiple platforms and protocols.
Related market:
https://manifold.markets/jgyou/what-will-the-dominant-social-media