You think its hard beating The Eye?  You should have seen what it took to balance that level to the razor thin margin that it has.  The idea was that you had to be perfect to win it (or find some way I didn't think of).  Actually, it's the "find some way I didn't think of" part that's the real problem. 

When working on Swarm, I really leaned on my friend Skippy to help me build levels.  We have worked together for years, so I knew we were on the same wave-length as far as this sort of design work.  But the funny thing was how one of us or the other would build a level and hone it to a razor fine balance and then give to the other to test and they'd beat it instantly with no problem.  Why?  Just because they'd use an approach that was different.

In fact, in Swarm... this problem is a lot worse than it is in GD.  The reason is that GD uses fixed paths... so one had to test lots of ways people could put towers around the paths.  It's a lot of variables to test for!  But nothing compared to the open-field style of Swarm!  You can approach the same level is a lot of different ways.  Even the more constrained of the levels still have vastly more variables than any of the geoDefense ones with its fixed paths.

Having two sets of eyes on the balance issue should make those medium and hard levels quite challenging and fun to beat.  The worst thing you can do is totally blow it and make the game too easy... then there's no challenge and no fun.  And it's easy to miss something.

I think I'll turn my thoughts on this into a blog post...

-- David

I was able to get 17,460.

I'm really looking forward to the online scoreboards - I'm never in the top ranks, but I love using the ridiculously high scores I see as motivation.   :)