Pull requests, based on https://blog.jessfraz.com/post/analyzing-github-pull-request-data-with-big-query/

pull_requests.sql

SELECT
  repo.name AS repository,
  COUNT(*) AS pull_requests,
  EXACT_COUNT_DISTINCT(actor.login) AS contributors
FROM
  [githubarchive:year.2016]
WHERE
  type IN ( 'PullRequestEvent')
  AND JSON_EXTRACT(payload, '$.action') IN ('"opened"')
GROUP BY
  repository
HAVING
  contributors > 10
ORDER BY
  pull_requests DESC
LIMIT
  50

results.csv