Balancing, or: What did we do wrong?

So first off:
We released Prototype Defense this past weekend and within 3 days, already had more than a 1.000 downloads. Huzzah!
Now onto the important part of this post: Balancing!
As you all probably know, balancing is one of the more important aspects of a game.
We at CodePoKE all agree, and recognize its importance.
However; dammit all if it isn’t one of the most boring things to do during the development phase.
Understand that we’re a very small team of part time developers. (All of us are full-time students at the University)
So we simple have no-one to do the grinding aspect of balancing for us.
Balancing consists of nothing more than continually replaying the same level over and over again until you can barely win it.
As you reread that sentence, you probably already see where we went wrong.
We balanced all levels in such a way the we, the developers, could barely win by the skin of our teeth.
The problem here is that we already know exactly how to play the game.
We know what combination of towers work, and which ones don’t.
We thought we adjusted for this by setting the default difficulty level at 80% of our playing level.
Then the Google Analytics results finally came in; And oh boy, were we wrong:
The upside? Some of you were such die-hard tower defense players that, even when you only succeeded once every twenty tries on a level, you kept on playing until you actually beat the level.
The down side’s pretty clear though: only 1 in 6 players actually beat the first level, even after trying multiple times to beat it.
What we learned in the end is pretty simple: Never ever let the guys doing the developing actually do the balancing of a game. Buy your friends a pint or a pizza to come play the game, and let them do it!
Also; don’t forget to include some form of analytics in your game/app to actually see how people are using it!
Tonight’s going to be a long night, as we’ll want to fix this a.s.a.p. for 1.0.4!