from AnonomousWolf@lemm.ee to fediverse@lemmy.world on 18 Apr 09:37
https://lemm.ee/post/61681988
🧮 Decentralization Scoring System (v1.0)
This scoring system evaluates how decentralized and self-hostable a platform is, based on four core metrics.
📊 Scoring Metrics (Total: 100 Points)
Top Provider User Share (30 points): Measures how many users are on the largest instance. Full points if <10%; 0 if >80%.
Top Provider Content Share (30 points): Measures how much content is hosted by the largest instance. Full points if <10%; 0 if >80%.
Ease of Self-Hosting: Server (20 points): Technical ease of running your own backend. Full points for Docker/simple setup with good docs.
Ease of Self-Hosting: User Interface (20 points): Availability and usability of clients. Full points for accessible, FOSS, multi-platform clients.
📋 Example Breakdown (Estimates)
📧 Email (2025)
- Top Provider User Share: Apple ≈ 53.67% → Score: 4.5/30
- Top Provider Content Share: Apple likely handles >50% of mail → Score: 4.5/30
- Self-Hosting: Server: Easy (Leverage email hosting services) → Score: 18/20
- Self-Hosting: Client: Easy (Thunderbird, K-9, etc.) → Score: 18/20
Total: 45/100
🐹 Lemmy (2025)
- Top Provider User Share: lemmy.world ≈ 37.17% → Score: 12/30
- Top Provider Content Share: lemmy.world likely hosts ~37% content → Score: 12/30
- Self-Hosting: Server: Easy (Docker, low resource) → Score: 18/20
- Self-Hosting: Client: Good FOSS apps, web UI → Score: 18/20
Total: 60/100
🐘 Mastodon (2025)
- Top Provider User Share: mastodon.social ≈ 42.7% → Score: 11/30
- Top Provider Content Share: mastodon.social ≈ 45–50% content → Score: 10/30
- Self-Hosting: Server: Docker setup, moderate difficulty → Score: 15/20
- Self-Hosting: Client: Strong ecosystem (Tusky, web, etc.) → Score: 19/20
Total: 55/100
🔵 Bluesky (2025)
- Top Provider User Share: bsky.social ≈ ~90%+ (very centralized) → Score: 0/30
- Top Provider Content Share: Nearly all content on bsky.social → Score: 0/30
- Self-Hosting: Server: PDS hosting possible but very niche → Score: 4/20
- Self-Hosting: Client: Mostly official client; some 3rd party → Score: 10/20
Total: 14/100
🟥 Reddit (2025)
- Top Provider User Share: Reddit ≈ 48.4% → Score: 0/30
- Top Provider Content Share: Reddit hosts a significant portion of user-generated content → Score: 0/30
- Self-Hosting: Server: Not self-hostable (proprietary platform) → Score: 0/20
- Self-Hosting: Client: Some unofficial clients available → Score: 3/20
Total: 3/100
How Scores are Calculated
🧑🤝🧑 How User/Content Share Scores Work
This measures how many users are on the largest provider (or instance).
- 100% (one provider): If one provider has all the users, it gets 0 points.
- No provider > 10%: If no provider has more than 10%, it gets full 30 points.
- Between 10% and 80%: Anything in between is scored on a linear scale.
- > 80%: If a provider has more than 80%, it gets 0 points.
📊 Formula:
Score = 30 × (1 - (TopProviderShare - 10%) / 70%)
…but only if TopProviderShare is between 10% and 80%.
If below 10%, full 30. If above 80%, zero.
📌 Example:
If one provider has 40% of all users:
→ Score = 30 × (1 - (40 - 10) / 70) = 30 × (1 - 0.43) = 17.1 points
🖥️ How Ease of Self-Hosting Scores Work
These scores measure how easy it is for individuals or communities to run their own servers or use clients.
This looks at how technically easy it is to run your own backend (e.g., email server, Mastodon server) or User Interface (e.g., web-interface or mobile-app)
- Very Easy: One-command Docker, low resources, great documentation → 18–20 points
- Moderate: Docker or manual setup, some config, active community support → 13–17 points
- Hard: Complex setup, needs regular updates or custom config (e.g. DNS, spam) → 6–12 points
- Very Hard or Proprietary: Little to no self-hosting support, undocumented → 0–5 points
PS.
This is Version 1.0 so there are likely flaws and mistakes in it, feel free to help create the best version we can I’ve put it on https://github.com/NoBadDays/decentralization-score
threaded - newest
The main reason for me creating this is to shine light on services like Bluesky claiming to be Decentralised, but they are in fact not.
If we have an unbiased scoring system, then we can make things like this clear to people, and we can work towards building truly robust decentralised services.
This can even be used to help move users or content to smaller instances to help strengthen decentralisation
Nice :D
I agree with your message, but if you want credibility you need sources for this data.
Agreed, we need to find a way to take as much if not all subjectivity out of the calculation. And get reliable data.
This is just a starting point, it can be improved a lot.
Are you saying Apple is a larger email provider than Google? I’d find that very surprising.
Based on my brief searches yes, but I haven’t looked into the example data in great detail.
If you have a good data point for me I can update the examples.
@AnonomousWolf great avatar btw HOhOHOhoH
@fediverse @anothermember
No data I’m afraid but it just doesn’t ring true to me, unless there are vast regional differences. It sticks out to me as much as if you’d said that Bing is the largest search engine; I’ve barely heard of Apple email but almost everyone I know uses Gmail except me, including Apple users I know.
You are 100% correct yes, As I said I didn’t put much effort into making sure the data in the examples are correct. This post is more about getting visibility on the idea. I will put more effort into better data and more examples.
Same here. Do you think this data is from Email Apps/Clients rather than the addresses? iPhones are extremely popular and so I can imagine loads of people open their mail on their phones. But I have a hard time seeing more users being on Apple mail adresses compared to the huge widespread gmail.com users.
Or Microsoft/outlook/exchange. Apple wouldn’t even be in my top 20 for this question.
This is a very interesting approach. Putting this on a website with graphs, sources and additional information would elevate it to a fledgling standard.
I’d love to do that, there are existing Fediverse stats websites that should be able to implement this without too much difficulty.
What is the 48.4% for Reddit, since it isn’t decentralized at all?
That’s a mistake, I’ll update the examples in the next version
Fixed in version 1.1 on GitHub
No way that’s true.
You’re right, I’ll update the example data in the next version
That’s very wrong yes. I fixed that in version 1.1 on GitHub.
Still not perfect, but I’ll work on it further after easter weekend
One metric you might want to add is the network effect: how much of a difference does it make to the user experience to join a large instance (or the same instance most of your friends are on) compared to a small or self-hosted one? (Or in other words—does the nature of the platform software potentially incentivize consolidation?)
Thanks, I need to think about how to score that, but good idea.
Maybe what % of the network is federated to the median size instance?
piefed? mbin?
Very interesting idea, thank you :)
So for my feedback, i love your project and i can't wait to see where you will go. Great job :3
For me lemmy is still centered on lemmy.world and theqbiggest community (in any instance) doesn't help users to move to another community as they will post where there are lot users to reach them.
For myself, I would add two layers for decentralization scoring, maybe a separate scoring :
Ability to migrate
- account and its followers
- community : miror content and move it to another instance with its followers
- export a backup of a server in case of absent admins
- change software. The problem is the golden cage, we thould be able to migrate between lemmy > iceshrimp, mastodon > piefed...mostly between social network. I think nomadic identity can offer us an interesting solution.
Culture diversities
For lemmy, i think content is still too western centric :
- If you reduce lemmy instances by spoken language as french, we only have 1-3 servers, so we have a single point of faillure. On mastodon there are several french instance.
- If i count english content versus indian, chinese, arab, spanish...we are pretty far from being decentralized. And yes, we also need to note their countries regulation as our to get a clear overview.
Thanks this is a great idea.
I like the simplicity of that current scoring system (only 4 metrics)
Perhaps the same metric can be used, but applied for different cultures and languages.
Eg Lemmy would score high in English, but not in French.
I’ll keep refining it and listening to input.
Yes culture and language should be another metric.
But migrate option, i think they could be integrated because they are a key part of a decentralized system. I you can't move easily from facebook, reddit, you will stay. And that's why people stay, they shared life with people, they have build together a community...
There is an article about nomadic identity that explain well those part. :)
Ability to migrate is definitely very important, In a newer version I’ll add it in.
What weight do you think it should have?
I’ll shrink the others so the score stays 100
Or maybe a separate scoring system ? So i will be easier to manage it ?
I think the biggest difficulty would be able to figure out a scoring system that works for all platforms.
Eg. What does “being able to migrate users” mean?
Does it include followers? Content etc.?
And what about “sub-reddits” it’s important for it to be able to be migrated for Reddit/Lemmy but PeerTube doesn’t have something like that
You can limit it to social network. I class peertube, pixelfed as a media plateform rather a social network one.
For migration, can i migrate my account with :
- My followers and giving them my new adress ? Do it update itself or do we have to do it ?
- my contents : communities, bookmark, posts, comments...
- can i delete my account ? Can i delete all my content ? and how easy it is ? There we will see Meta doing an amazing score.
- bonus : can i export settings as colors scheme ? (Accessibility for visually impaired)
- can i give community to new mod ? Can i fork it ?
And let's add accessibility. :)
We don't talk much about it and we are doing a very bad job. So the scoring system should also raise this issue : how accessible is the decentralized software.
Isn't ActivityPub extremely network intensive though? If all you wanted was a single user subscribing to a handful of communities then Lemmy would be inexpensive but to pull from a lot of communities I thought people have said that it can cost a bit of money, time, etc. Also defending against attacks such as CSAM.
Maybe make a distinction then between running a "tiny personal instance with only a few niche community subscriptions" vs. a small instance, either with multiple users or even just one person subscribing to many communities, if that cost would start to become more prohibitive?
I don’t understand why reddit would even be on this list.
For comparison