I recently spent some time looking at a bunch of tower defense games, and it saddened me to see how little innovation most people attempt in that genre, although I guess that is something pretty generally applicable to all genres, and most peoples first game. I might be lucky, in that I always have an urge to put some sort of ‘spin’ on any game I make, i am never happy to just do ‘a game in this genre’ and leave it at that. It’s worth remembering, as a game designers, that the player will always be asking themselves ‘why do I need to play this, when I can just go play the classic tower defense game X’.  You really need to have an answer to that question.

With regards to Gratuitous Tank Battles, which is my first ever tower defense game, I probably overdid it a bit in terms of trying to innovate. They say you should only innovate in one direction at a time, but I think life is too short for that. What I tried to do was question all of the ‘design assumptions’ of the genre.

The first assumption is that the player is the defender. Obviously the clue is in the title, but why wouldn’t a ‘tower attack game’ work? I think GTB shows that it can work. There was already a game released that did this, although it was essentially just escort missions, thankfully, because it was released just as I was half way through making GTB.

The second assumption is that the towers are invulnerable. I think this is arbitrary and crazy. It adds an extra level of excitement and gameplay to Gratuitous Tank Battles to know that you can’t just place a big gun somewhere and know you have that site covered. This seemed like a major change, and a change for the better to me.

The third assumption was that the towers are of fixed designer-decreed configuration, and can receive a linear upgrade path mid-battle. People really expect that in a TD game, and not including it does un-nerve people, but it adds a whole new layer to the game in terms of unit customization. Also, that ties in nicely with the design of Gratuitous Space Battles, where it was the major focus.

Another assumption was that the attackers come in a pre-set path, linked to a radar which gives you advance warning, also that they come in waves. All of this is totally arbitrary. There are not stone tablets decreeing the rules of tower defense design, you can do whatever you like. I broke all three of those assumptions about attacking waves, and personally I think that adaptive AI for the attackers (and defenders) totally changes the nature o the game and vastly expands the play-time available. This is one of the assumptions I was most proud to break.

There are other changes to GTB that make it a non conventional TD, such as the setting, the level editor and map/unit sharing online, but I think the key to making the game work and be interesting was that I looked at the ‘set-in-stone’ assumptions and basically kicked them all out. I find that this works amazingly well in game design. Some assumptions are there for a reason, but many are not, and when they are broken, once we get used it, we love it. The Sims can be turn based (see Kudos) A game can have no ‘game world’ at all (see Democracy) A space battle RTS can have no player control (see GSB) and there are much bigger examples too:

The object of an FPS can be not to kill (See Thief), health packs can be made redundant, with auto-regenerating health (wasn’t that Call of Duty 4?). Base building doesn’t have to be in every RTS game (again, not sure who started that).

The ‘classic’ design of a genre is a mere starting point. When you design a game, you need to question every aspect of it, and make sure you have a rock-solid defense of why it ‘has to be that way’.

11 Responses to “Tower defense game design”

  1. Will says:

    One thing that I personally dislike in TD games (a genre I normally adore) is if I’m forced to ‘babysit’ my towers/defenses. That means if they can be damaged, then I have to always worry about playing whack-a-mole putting them back rather than doing the fun bit – planning and building my system. Unfortunately, this means that I have never been able to extract any joy from my purchase of GTB. That’s fine, I’m not sad :-) Buying any game is a small gamble (like seeing any movie or doing just about ANYTHING), and I’m certain that this exact game-style is wonderful fun to other people.

  2. Szymon Kurek says:

    Heh, I don’t know what’s the next game you are working on, but two things are sure – you’ll break a few rules and you’ll add a lot of dials, parameters and plots ;) That’s a sort of signature for your games.

    There are other interesting ways to defy well-known game design schemas – for example, try cutting out the obvious bit. What’s an RTS without moving units? What’s an RPG without combat? You might get a distilled form, or something completely new :)

  3. cliffski says:

    In some ways an RPG without combat is the Sims, and that was pretty popular :D

  4. Lars Doucet says:

    Bravo, Cliffski!

    We broke a lot of classic Tower Defense rules in Defender’s Quest, too, so I support you in this. We also doubled-down on letting “towers” get hurt, and I think that really adds a lot to the game if you get it just right.

    There’s so many new strategic elements that particular feature adds – not only are you then able to design a whole panoply of enemies that attack in different ways, the player now has to really think about whether they want to invest in that specific tower without dedicating support to keep it alive. For longer-ranged towers, it also makes “vulnerable front line w/ max targets” vs. “safe back row w/ fewer targets” an interesting tactical trade-off.

  5. Werit says:

    So what’s next for you Cliff? :)

  6. Actually, the first auto health regen FPS I can think of is Medal of Honor.

  7. Trifler says:

    I would buy a DLC adding much more complex unit and turret design. While one of Cliffski’s blogs indicated that he thought the complexity of ship design was somewhat of a turnoff for some people in GSB, I personally love complex design and actively look for games that have it. Just my $0.02 cents.

  8. Muttala says:

    I like your post Cliff :)

    just my additional opinion, some rules are really good to follow even they seems to be common.

    I Quote the example of TA and MOO2 in RTS. TA (Total Annihilation) has a really strong fan’s love but when it comes to latest incarnation – Supcom2 it is heavily criticized for “dumb down” when supcom2 trying to make it more appealing to general public. MOO2 (Master of Orion 2) being the same as with the automation and oversimplification in MOO3.

    In my view, games today are moving in a direction to appeal “everyone” or more “accessible” to general public. That’s why remake become an alternative direction which is more highly welcomed. Example settler 2 10 years anniversary vs settler 7, Baldur Gate remake – enhanced edition vs upcoming Dragon age 3, and the fan’s cry for AOE 2 remake if you google , etc

    We certainly cannot make everyone happy with 1 single game design, I guess 2 rules of thumb we can consider are

    1) Like people ask Chris Taylor on the success of TA and he says well, at that time I just want to create TA as “RTS for myself”, something I myself are deeply excited to play for – Gamer for Gamer things.

    2) study on the fun factors in top rated game. If any fun factor suitable to add up in your games or you want to break the common pattern and create a better one.

    Just my 2 cents,
    I myself not very keen of the pace of GTB and common WW2 setting but that’s ok as Will point out buying games are small gambles…

    P/s: Damn where is our next GTB update….? lol

  9. Albert1 says:

    Autoregen in modern FPSs is one of the reasons I still mostly play Doom 2/Quake 1/2.

  10. Matt says:

    Interesting post!

    I think in some ways though genres are defined by what they DON’T do.

    If you had a tower defense game where you could attack or defend, had moving units, all units could take damage, and you could redirect the units at will (not just fixed paths), you would basically have created an RTS, wouldn’t you?

    If not, then what makes it still TD instead of RTS? Even Warcraft 2 had fixed defense towers.

    It IS good to question those assumptions though, because the genre is really arbitrary at the end of the day, what matters is whether or not the game is good.

  11. […] Tower Defense Game Design (Cliffski.com) “I recently spent some time looking at a bunch of tower defense games, and it saddened me to see how little innovation most people attempt in that genre, although I guess that is something pretty generally applicable to all genres, and most peoples first game. I might be lucky, in that I always have an urge to put some sort of ‘spin’ on any game I make, i am never happy to just do ‘a game in this genre’ and leave it at that. It’s worth remembering, as a game designers, that the player will always be asking themselves ‘why do I need to play this, when I can just go play the classic tower defense game X’.  You really need to have an answer to that question.” […]