How to design a rubric committees trust
Criteria, descriptors, weights and anchoring, and why a 0 to 10 scale with written descriptors produces a ranking a committee can defend.
Read the article
Blog
Notes on running better calls
Rubrics, rounds, evaluator workload and the administrative detail that decides whether a call finishes on time. Written for the people who run programmes, by the team that builds the platform they run them on.
Running a multi-round seed-fund call
A timeline, three rounds, promotion rules written before the scores exist, and the workload arithmetic that decides how many evaluators you need.
What grounded AI evaluation means
The difference between an assistant that writes a plausible paragraph and one that researches and cites, and where an AI co-evaluator belongs in scoring.
Resubmissions, duplicates and lineage
Teams resubmit for good reasons and bad ones. How to tell the difference fairly, keep a lineage chain, and give evaluators one version to score.
See IdeaScore run a call with your own rubric
A 30-minute walkthrough with a founder, using your programme's form and criteria.