hey guys, I'm toying with an idea of doing a reverse Tower Defense Game, essentially the idea is to select a group of guys to send through the gautlent.

essentially, when selecting the wave groups, your given x points, based on current wave, and number of succesful hit's on the target in the last wave, which can be used to construct the next wave(so fliers are more expensive, etc).

the hardest component imo is building the ai which can sufficiently construct tower's to defend. perhaps an only multiplayer tower-D, so one person selects tower construction, the other sends waves. and see who can win.


anyway...thoughts on such an idea?

Players will quickly learn what to survive and what they won't survive in terms of the defence etc., so this type of game play could easily turn into a stalemate where the attacking player doesn't even feel like attacking at all. Unless, of course, you make it so that there's a good enough return for using proper tactics etc. Maybe you could add traps and similar that the CPU can place out that you need to avoid, etc.

There's actually an interesting mission in Total Annihilation where you need to zig zag your units between obstacles towards a base that has a Big Bertha artillery cannon (I think it was). Out in the open, the Bertha hits them and destroys your army. But behind the obstacles, you're safe.

the hardest component imo is building the ai which can sufficiently construct tower's to defend. perhaps an only multiplayer tower-D, so one person selects tower construction, the other sends waves. and see who can win.

Maybe you're doing it too complex. Most TD games I've played have fixed setup.

• Enemy wave is designed and constant for a level
  
• Some TD games also restrict the placement of towers (those typically only allow critters to follow a path)
  
• Some TD games keep looping the surviving critters
  
  So I think there's no need, at least for an initial iteration, to have fully unknown starting point.
  But supposing you really want to start with a completely unknown initial configuration: who starts first? The AI building the towers? Is the player allowed to see them? It seems you're inclined to allow the AI see the monsters in advance. I'm not really sure this would work.
  
  There's a serious problem IMHO with this design. If the AI lost, it has no benefit. By contrast, the player is now more powerful so next level should theoretically be even easier. I think the only good way to solve this issue is to have predefined paths with pre-designed towers and let the player move between successive scenarios.

[quote name='slicer4ever' timestamp='1344639120' post='4968231']the hardest component imo is building the ai which can sufficiently construct tower's to defend. perhaps an only multiplayer tower-D, so one person selects tower construction, the other sends waves. and see who can win.

Maybe you're doing it too complex. Most TD games I've played have fixed setup.

• Enemy wave is designed and constant for a level
  
• Some TD games also restrict the placement of towers (those typically only allow critters to follow a path)
  
• Some TD games keep looping the surviving critters
  
  [/quote]
  perhaps, the only thing is that the Enemy wave is going to be unknown, I'd probably allow the AI to know what is potentially available to be sent(but wouldn't know exactly what is coming), maybe ever allowing the ai to construct a probably wave, and build against what it believes will come(or even evaluates what enemy's got the furthest on the prior wave, and try to construct towers to better handle the type of guys that got further(e.g. if fast but weaker guys got through further, it'll put up more rapid fire towers, but if heavy health/slow speed guys got through, then it'll build stronger but slower towers).
  
  

  
  So I think there's no need, at least for an initial iteration, to have fully unknown starting point.
  But supposing you really want to start with a completely unknown initial configuration: who starts first? The AI building the towers? Is the player allowed to see them? It seems you're inclined to allow the AI see the monsters in advance. I'm not really sure this would work.
  

  
  I believe i'd do it such a way, that the AI can see what the player has a possibility to send, but not what he actually sends(this gives it some fixed point to start building against), as for the player, I'd probably do a fog of war type system between rounds, so he can never see what's been built between waves.
  
  

  
  There's a serious problem IMHO with this design. If the AI lost, it has no benefit. By contrast, the player is now more powerful so next level should theoretically be even easier. I think the only good way to solve this issue is to have predefined paths with pre-designed towers and let the player move between successive scenarios.
  

  
  
  Too be honest, isn't this the same problem with a regular tower-D, if you let a guy get though, that's less money you have to create better towers. so you go into a slow down-ward spiral if you can't fix the problem.

There are several reverse tower defense or tower offense flash games on flash portals. Googling "tower offense" would bring up a few. Also, I found a full fledged 3D tower offense videogame: http://en.wikipedia.org/wiki/Anomaly:_Warzone_Earth.

Thanks mate, i didn't think of tower offense being the term for such gameplay, i'll try some out and see if they are enjoyable.

Thanks mate, i didn't think of tower offense being the term for such gameplay, i'll try some out and see if they are enjoyable.

You are welcome. Try googling "reverse tower defense" too. In particular, I remember Villainous being quite fun: http://www.kongregate.com/games/Rete/villainous.

[quote name='slicer4ever' timestamp='1344695211' post='4968415']
Thanks mate, i didn't think of tower offense being the term for such gameplay, i'll try some out and see if they are enjoyable.

You are welcome. Try googling "reverse tower defense" too. In particular, I remember Villainous being quite fun: http://www.kongregat...ete/villainous.
[/quote]

I actually came across that yesterday, It is quite fun. I do believe i'm going to attempt something for tower offense in my game, as i enjoy the concept that some of these games have done.

Bit old of a topic, but i recently re-visited this idea, and made a multiplayer tower defense game, where you both do defense, and offense and try to outlast your opponent:

[media]

[/media]

https://www.facebook.com/LiquidGames

Wouldn't castle siege be the reverse of tower defense?

Also, there's I, Zombie - the minigame in the Plants vs. Zombies tower defense game where you get to play the attackers:

">