Would be interesting if they split funding with anonymized data and the other half with just randomized funding year per year