co-endorsement
Strategy pairs people actually hold together.
Which pairs of strategies are most commonly endorsed by the same people. Behavioural data, not what the catalogue declares compatible. The two views can disagree in instructive ways.
Pairs with fewer than three shared endorsers are not shown: too noisy to read as signal. The corpus is sparse on cross-strategy endorsement: most people argue for one or two positions, not many. Even three shared voices is meaningful at this density. Compare against the declared relations matrix: where the data and the framework disagree, one of them is out of date.
Most-shared endorsers · raw count
Tightest pairs · jaccard
Same people hold both, ranked by overlap as a share of either set. Surfaces small but tightly-coupled positions a raw count would miss.