-
Notifications
You must be signed in to change notification settings - Fork 1
Home
We’re building an open, evidence-based platform to rank arguments, measure belief strength, and automate conflict resolution — bringing structure to online debate.
Google’s PageRank can be adapted to assess the strength of pro/con arguments.
Just as PageRank values a site based on the quality of links, ReasonRank values conclusions based on the strength of supporting arguments — factoring in logic, evidence, and relevance.
We integrate two powerful measures:
- Logical Fallacy Score — measures the degree and impact of fallacies in an argument.
- Evidence Verification Score — evaluates how well the claim is supported by independent, rigorous evidence.
Combined, these scores quantify argument credibility and encourage reasoning grounded in both logic and verifiable fact.
Our system measures the relevance and causal strength between evidence, intermediate arguments, and final conclusions.
The method:
- Weighs each link in the reasoning chain.
- Uses an adapted PageRank model to reflect collective judgment on relevance.
- Argument Importance Score — measures how much a belief depends on a specific argument.
- Belief Impact Score (BIS) — combines argument strength with cost-benefit analysis to assess the real-world consequences of accepting a belief.
These scores help identify the most influential arguments and the beliefs that most affect society.
We group equivalent or highly similar arguments so debates stay focused and concise.
- Equivalency Score — calculated using semantic similarity + argument performance.
- Displays “better ways of saying the same thing” to reduce clutter.
Our conflict-resolution engine:
- Focuses on interests, not positions.
- Evaluates solutions against objective criteria.
- Scores proposals on resolution likelihood, factoring in cost-benefit trade-offs and compromise requirements.
Measures how stable a belief’s score is over time based on:
- Depth of analysis.
- Unresolved sub-arguments.
- Stability across repeated evaluations.
A belief with thousands of hours of stable analysis carries more weight than one quickly assessed.
By combining crowdsourced reasoning with structured scoring systems, we aim to:
- Make arguments searchable, comparable, and ranked.
- Show the strongest and weakest points for any belief.
- Foster productive, respectful, evidence-based debate.
Join us in building the internet’s “truth infrastructure” — where better arguments rise, weaker ones fall, and decisions are made with clarity.