I have to be honest when I say: This is not a "strategy" game. Throughout the entire game, I saw absolutely no reason not to spam the highest-grade melee unit I could(after all, any ranged unit becomes a melee unit, when they're attacked at close range - and their improvements cost more, anyways) and click wildly at any secondary concern (such as: make the hero attack, throw a spell, or so on). I didn't have to think about what I was doing. I just had to follow a pattern and click.
This brought me to the concept of tactical combat. Why should combat be tactical, if you have unlimited resources? You reward players for the death of their soldiers, depending on how close to the enemy's stronghold they die at. In true strategy, you don't gain a damn thing when you needlessly send troops to their deaths. You've trained them, clothed them, fed them - and when they die, you lose every resource that you've put into them, as well as all the experience they've gained. If you were a goblin, is that what you'd want out of your Clan Leader?
A strategy player should concern themselves with every troop, not with every gold piece. Try taking away the rewards for a unit dying anywhere, anyhow, and trying this game again. Then, we might have something.
2.5 stars for the effort put into production: The music, the graphics, the back-story, the code, the art, and the thoughtfullness put into this. If you want more stars from me, however, there are two lessons to be learned here: 1.) Those whom play your game for some kind of emotional appeal or intellectual challenge are going to want, surprisingly enough, the emotional appeal of a war situation, or the intellectual challenge of having to develop an actual strategy, rather than a system to follow. 2.) Those whom play your game for the sheer self-gratification of dismantling a particular system to optimize their victory are going to become bored fairly quickly, as the sheer math involved is too predicatable. A more random result might force such a person to choose between, say, a unit with the best probablitiy of a moderate result, versus a unit with a high probability of both extreme failure and extreme success. Giving some units a bonus against specific other units would also give a player pause, and add to the concept of "strategy"
I'm not writing this to be a dick. I'm writing this as a sincere critical review. If I didn't care, I wouldn't post.