听起来你只对基于里程碑日期 * 之后的球员行（表的奇怪名称，因为它似乎代表比赛，而不是球员）进行分类感兴趣。
这是非常简单的（未经测试）：

select m.player_id, m.milestone_date,
    case
        when min(p.date_played) <= m.milestone_date + interval 60 day then 'Active player'
        else '60 day churner'
    end churn_status,
    case
        when min(p.date_played) is not null then 'returner'
        else 'churner'
    end player_status
from milestone m
left join player p on p.player_id=m.player_id and p.date_played > m.milestone_date
group by m.player_id

这将查找每个玩家在里程碑日期之后的玩家行的最小游戏日期（如果有的话），并使用它来确定状态。

使用左连接来匹配60天内和60天后的战术。

SELECT m.player_id, m.milestone_date, 
    IF(MAX(c.player_id IS NULL), '60 day churner', 'Active player') AS churn_status, 
    IF(MAX(r.player_id IS NULL), 'churner', 'returner') AS player_status
FROM milestone AS m
LEFT JOIN player AS c ON c.player_id = m.player_id AND c.date_played > m.milestone_date AND c.date_played <= DATE_ADD(m.milestone_date, INTERVAL 60 DAY)
LEFT JOIN player AS r ON c.player_id = m.player_id AND c.date_played > DATE_ADD(m.milestone_date, INTERVAL 60 DAY)
GROUP BY m.player_id

这与player连接了两次。第一次连接获取了里程碑日期和60天后的所有行。第二次连接获取了60天后的所有行。
如果两个日期关系都不匹配，则该连接将返回一行，其中player中的所有列都是NULL。IF()条件测试每个连接是否包含这些空值，并在该列中生成相应的结果。