gistfile1.cs

    private static LeaderboardName getLeaderboardName(
      Kind kind, LeaderboardKind lbKind
    ) {
      var kg = kind as Kind.Global;
      if (kg != null)
        return new LeaderboardName(
          string.Format("lb_{0}", leaderboardKindToString(lbKind)), lbKind
        );

      var kw = kind as Kind.World;
      if (kw != null)
        return new LeaderboardName(
          string.Format("lb_w{0}_{1}", kw.world, leaderboardKindToString(lbKind)),
          lbKind
        );

      var kl = kind as Kind.Level;
      if (kl != null)
        return new LeaderboardName(
          string.Format(
            "lb_w{0}_l{1}{2}_{3}", kl.world, kl.level, kl.bonus ? "b" : "",
            leaderboardKindToString(lbKind)
          ), lbKind
        );

      throw new Exception("Unknown kind " + kind);
    }

Once again, very strange code, clearly not in C#-way.

So, you want to get leaderboard name from some object that is of class Kind. But why the hell do you create a separate static method for that? A clear and pretty simple objected-oriented approach would be to create an abstract method in Kind and to define it in it's children.

Are there any downfalls to this traditional OOP way of doing things that I don't see?