This is something that's been bugging me for a while, but I never put my finger on exactly what the problem was until last night when I ended up spending about an hour on a CTF rotation server.
Simply put, TF2's Capture the Flag mode feels like Valve never got around to really finishing it. Think about it: Every other mode has its own special entity setup. You've got your tf_gamerules and your train watchers and your control point masters... CTF just has the flags themselves and the capture zones. The "timer" just shows how much time you have until the round changes, and even the goal is controlled by a server setting. And when the timer runs out, is it a stalemate? No, whichever side has the most points (which I assume means the total of all the players' points on the scoreboard, so basically whoever has the most frags — see, it really is Deathmatch!) automatically wins.
Not only is this a shoddy and unprofessional way to design a gamemode, it's a lousy testing environment to boot. If you look at, say, the stats recorded by our feedback system, the RED/BLU win ratio tells you absolutely zippo because there's no way to tell if a given "win" happened because they actually captured all the flags, or because time ran out and they had a higher score. Nor how much time the round went because maybe out of a 30-minute test, it took them three minutes to win one time, twenty-one the next, and then won by default when the timer ran out.
It's kind of no wonder there are so few good CTF maps; nobody has ever been able to gather good stats on how playtests turned out! As someone who currently has two CTF maps in the works, this is rather disconcerting.
So what should be done about this? Personally, I'd like to see the same standards used for 5CP applied to CTF. Start with about five minutes on the clock, add some time whenever there's a successful capture, end with a stalemate if the timer runs out, no matter what. Hard-code a three-capture win condition into the map. If a round ends with less than five minutes left on the map clock, we're done; if the map clock runs out before the round is over, keep it rolling until it is.
There are other ideas I have for improving the mode, which I mentioned in the 2fort thread (have the flag constantly heal whoever's carrying it like the Payload card, implement the respawn time manipulation mechanic that was removed from Asteroid) but that's stuff to be experimented with down the line. For now, the stuff I described above seems like a good place to start, and given how malleable stock CTF is (I don't think SD uses any entities that weren't already in the game, for example), it should be possible for the map creator to hard code them into their map without demanding any work from Valve. What do you all think?
Simply put, TF2's Capture the Flag mode feels like Valve never got around to really finishing it. Think about it: Every other mode has its own special entity setup. You've got your tf_gamerules and your train watchers and your control point masters... CTF just has the flags themselves and the capture zones. The "timer" just shows how much time you have until the round changes, and even the goal is controlled by a server setting. And when the timer runs out, is it a stalemate? No, whichever side has the most points (which I assume means the total of all the players' points on the scoreboard, so basically whoever has the most frags — see, it really is Deathmatch!) automatically wins.
Not only is this a shoddy and unprofessional way to design a gamemode, it's a lousy testing environment to boot. If you look at, say, the stats recorded by our feedback system, the RED/BLU win ratio tells you absolutely zippo because there's no way to tell if a given "win" happened because they actually captured all the flags, or because time ran out and they had a higher score. Nor how much time the round went because maybe out of a 30-minute test, it took them three minutes to win one time, twenty-one the next, and then won by default when the timer ran out.
It's kind of no wonder there are so few good CTF maps; nobody has ever been able to gather good stats on how playtests turned out! As someone who currently has two CTF maps in the works, this is rather disconcerting.
So what should be done about this? Personally, I'd like to see the same standards used for 5CP applied to CTF. Start with about five minutes on the clock, add some time whenever there's a successful capture, end with a stalemate if the timer runs out, no matter what. Hard-code a three-capture win condition into the map. If a round ends with less than five minutes left on the map clock, we're done; if the map clock runs out before the round is over, keep it rolling until it is.
There are other ideas I have for improving the mode, which I mentioned in the 2fort thread (have the flag constantly heal whoever's carrying it like the Payload card, implement the respawn time manipulation mechanic that was removed from Asteroid) but that's stuff to be experimented with down the line. For now, the stuff I described above seems like a good place to start, and given how malleable stock CTF is (I don't think SD uses any entities that weren't already in the game, for example), it should be possible for the map creator to hard code them into their map without demanding any work from Valve. What do you all think?
