Shelled IGC05 -> GDC06

a few blogs back I posted some screenshots of what shelled was looking like for GDC06. as a milestone version, I wanted to take a minute to explain what's changed with the game since its dismal showing at IGC05 and provide some insight on what was learned. the difference between the two was 5 months time yet the two games are worlds apart.

I wish I could post screens to illustrate better but am lacking in time. after shelled is released I'll make the IGC version available to any dev's interested in comparing.

comparison does not include all changes of course, but the big ones from a high level POV. some just have to do with art that was ready at time of build and not necessarily in the build, and some are just features completed, but most of these reflect a change in philosophy.

###

IGC
complex game types with focus on variety

GDC
one game type for one game

why: better to focus on one game type that's super fun then a dozen that are kinda

#

IGC
complex interface with focus on user options and control

GDC
simple interface with focus on getting player quickly into game

why: the whole point of a casual game is to pop in and play; too many user options were provided for a casual audience

#

IGC
turn based

GDC
realtime

why: huge change with benefits immediately reaped - the game was TOO slow before, so much so that you could go and play another casual game while you were waiting for your next turn!!! the game is STILL slow compared to a FPS fragfest, but that's the point (again turtles, get it?), but it doesn't feel restricting but rather liberating (you can actually think / no twitch needed to win)

#

IGC
complex controls requiring RTFM, tutorial, or explanations

GDC
simple intuitive controls with optional advanced keys

why: not so much an immediate design change but an ongoing carving of the design with awareness and intention towards making simple, simple simple; adding optional keys gives the tech savvy crowd some additional stuff to play with, or for casual gamers who've master the game and want to further hone their technique; the important thing to note is that the advanced keys are NEVER required to play the game effectively and are totally optional; control scheme is as simple as mouse to fire, arrow keys to fly

#

IGC
wind and alternate gravities

GDC
no wind and single gravity

why: too hard!!! game is challenging enough as it is. created too many variables for adjusting user difficulty.

#

IGC
educational material on physics and seperate 2D mini-game

GDC
no educational material, no mini-game

why: less fluff, more game

#

IGC
20 weapons

GDC
9 weapons

why: LESS IS MORE. as jeff t said you don't buy games by the pound, so what does it matter if you have x # of this or x # of that? cut the fat off and kept the most popular weapons (i.e. spreads, nuke, rocket)

#

IGC
dynamically created levels

GDC
canned levels

why: this was an experiment and it failed. more time/funds was needed if this would have been pulled off. time/funds I didn't have to dick with something so non-critical to core fun.

#

IGC
no tank movement & hidden

GDC
full tank movement & highly accessible & new co-focus

why: tank movement code simply wasn't finished yet and it now is; but on top of that, having finished it brought about the discovery of how much fun it is to fly around, prompting movement to be much more important of a design focus than previously, also pushing it to the forefront of new simplified controls (you previously had to right click, select rockets, then move -- you now just hit a key and start to move)

#

IGC
green tanks and sci-fi mech tanks

GDC
rainbow colored tanks

why: mech tanks were cut, too sci-fi techy for a casual toony game; bright rainbow-colored tanks far more appropriate

#

IGC
environments by tim aste, high detail

GDC
christophe canon & toontown

why: when tim first sent concept art for the environments, it was very marble blast esque. I said no, no, no, way too toony, this game is somewhere between casual and hardcore. oops. well post IGC of course that changed, as everything moved in the direction of casual, but it was too late, the environments were done (and gorgeously might I add). tim's schedule was full so he couldn't re-do them, so canon stepped in and re-did the levels in his signature toontown style, which fit perfect. happily tim's art will be re-used for another game.

#

IGC
turtle tank char & borgbot char

GDC
cut

why: borgbot guy no longer appropriate for new casual/toon style; standard toon turtle cut because of no budget/time to further animate and work into the game or game art, and not critical to tank theme

#

IGC
misc finished models: cow, bullseye, balloons, darts...

GDC
cut

why: there was a ton of art to support aforementioned various game modes, all of which was cut to focus on just one game mode

#

IGC
both fun fonts and sci-fi techy fonts

GDC
fun fonts

why: casual / toon

#

IGC
magnus's sci-fi HUD

GDC
simple nearly no HUD

why: casual / toon

#

IGC
limited AI & no networking

GDC
fully functional AI & working playable networking

why: features that are now simply complete (or nearly)

###

so what was the lesson learned? well for starters the game at IGC was trying to sack both casual and hardcore players, that was big mistake number one. but a mistake that couldn't be learned until it was tried. the next big mistake was simply making everything too complicated in every place in the game in every way. it was TRYING to be simple, but trying is not doing. also, it was both reaching too high in depth, and reaching too far in breadth. start simple, start small. and finally, over-developing ahead of schedule, ie models done for game modes whose prototypes weren't even done to see if the game mode was even fun or not. too much forward thrust.

you can definitely "over-think" game design in that you can find solutions to problems that you anticipate but haven't actually occured yet. sometimes this works and you hit paydirt, and when that happens it's usually nothing spectacular, it pretends its all cool and knew all along that such and such thing would work. but when it doesn't work it's much more spectacular - a spectacular wreck. I'm proud of the IGC version I showed, it was a long road just to get to that point. and it was definitely a pre-pre-pre-prealpha and not even a dedicated build that was created. but it was wrought with unnecessary complexity, inconsistency, and user frustration. it took some hard, gut wrenching decisions (and of course lots of solid faithful work from lead programmer gary preston and programmer brian richardson) to get it back on track.

but back on track it is. some initial feedback is positive, but more importantly confirms the decisions made ("it was easy to get in and play"). and the game is set for you all to play and check next month -- but you gotta sign up for the newsletter on redthumbgames.com to get on the beta!

cheers.