SELECT t1.*
FROM (SELECT tag, category, COUNT(*) AS count
      FROM tags INNER JOIN stuff USING (id)
      GROUP BY tag, category) t1
LEFT OUTER JOIN 
     (SELECT tag, category, COUNT(*) AS count
      FROM tags INNER JOIN stuff USING (id)
      GROUP BY tag, category) t2
  ON (t1.tag = t2.tag AND (t1.count < t2.count 
      OR t1.count = t2.count AND t1.category < t2.category))
WHERE t2.tag IS NULL
ORDER BY t1.count DESC;

我同意这对于单个SQL查询来说实在太多了。GROUP BY子查询内部的任何使用都会使我畏缩。您可以通过使用视图使它 看起来 更简单：

CREATE VIEW count_per_category AS
    SELECT tag, category, COUNT(*) AS count
    FROM tags INNER JOIN stuff USING (id)
    GROUP BY tag, category;

SELECT t1.*
FROM count_per_category t1
LEFT OUTER JOIN count_per_category t2
  ON (t1.tag = t2.tag AND (t1.count < t2.count 
      OR t1.count = t2.count AND t1.category < t2.category))
WHERE t2.tag IS NULL
ORDER BY t1.count DESC;

但它基本上是在幕后做同样的工作。

您评论说，您可以在应用程序代码中轻松地执行类似的操作。那你为什么不那样做呢？进行更简单的查询以获取每个类别的计数：

SELECT tag, category, COUNT(*) AS count
FROM tags INNER JOIN stuff USING (id)
GROUP BY tag, category;

并在应用程序代码中对结果进行排序。