Start with random actions for each range and assume a static strategy. Calculate the total expected value of each action of each combo at every node. Use an algorithm (like CFR) to nudge the strategy of each combo towards the most profitable action, working backwards from the leaf-nodes. Iterate until Nash.
Or you could just use the free version of literally any solver.
https://www.piosolver.com/products/piosolver-free
Last edited by tombos21; 05-28-2020 at 04:16 PM.