Thursday, August 23, 2012

Notoriously Databased v0.30

Strategy tests

I feel confident that the turn processing is at a point where I can attempt to test individual strategies and see where they would lead in the main game.

The Digger

The first strategy will be just a simple turtle that digs as much as possible. This will be the baseline to judge all the other strategies against as it will have the least amount of risk. If it is also sustainable without going through a wipe, then we'll have a big problem attempting to get the other strategies to eek out an advantage, so the first test is to see how far I can go before a wipe occurs, and also how long to get to, say, 10 spaces.

Since the hero attacks are predictable from your current notoriety, the strategy will be to only leave enough minions behind to fend them off and dig with the rest. It may pan out to be better to drop the defense and go through a wipe rather than holding out on the defense right until the end. We'll see.

Run 1 - Always defend: 15 turns to reach 10 size. Not too bad, but there are a couple of places where treasure could be utilized either for elites or on fortifications.

Run 2 - Defend until defenders more than diggers: 13 turns to reach 10 size (10.8). Seems a little wasteful throwing 9 into digging when it was already on 9.9, but 10 size is just an indicator and I'd rather keep the tests consistent to the strategy for the long run. Start certainly seemed better, and no drawn-out sections of no digging.

Run 3 - 1/2 Defend using elites:  Elites look Ok on the first turn, but then break horribly. More debugging.