Costing/Economics

I haven't designed a game like this, but I have a couple thoughts from real economics.

First, the "expected value" of something that can end up in N different states is

Value in state 1 * probability of being in state 1 +
Value in state 2 * probability of being in state 2 +
...
Value in state N * probability of being in state N

where the probabilities ought to add up to one. This can help you do a first-round balance of more-vs-less risky investments. For example, production costs "4" to get "4"-worth of stuff whereas raiding costs only "2" to get "6"-worth of stuff, but only one-third of the time. Each "1" cost has an expected "1" payout. That "6"-worth might be a personal gain of "3" coupled with "3" damage to an opponent.

Since people tend to be risk-averse, you'll probably find you need to give a slight buff on the risky option from what is "fair" to get people to try it. So maybe a raid costing "2" gets you a 40% chance at "6"-worth.

** Note that economists have formulas to calculate the buff appropriate for specific assumptions about behavior. The underlying models allow for extremely accurate predictions that have almost no relationship with how people actually make decisions. **

Second, keeping the value of the resources nearly uniform is a bit trickier. The most direct method is to work out the rate of substitution between them. That is, what would be needed on average to make up for a shortage of 1 in a resource?

Suppose I can make up for a lack of 1 Food with an extra 2 Materials... a lack of 1 Materials with an extra 2 Money... and a lack of 1 Money with an extra 2 Food... that would hint at nominally equal value across the three. With nominally equal values, any serious imbalance in practice would cause your playtesters to exchange all of the useless resource for the valuable ones.