Document hawk/dove convergence logic + collect model status (running/convergence)

[rlskoeser]

• include write up of convergence logic in hawk/dove readme
• include status (running/converged) in model data collection so we can make sure we're analyzing runs where the model stabilized
• add an option to limit total batch runs for convenience
• make agent data collection optional in batch run script

@LaraBuchak : please review the write up of the convergence logic

• is it clear enough?
• are you ok with these numbers (30 and 15 for rolling window and minimum window)?

here's a more readable version of the convergence logic in the hawk/dove model readme: https://github.com/Princeton-CDH/simulating-risk/tree/doc-convergence/simulatingrisk/hawkdove#convergence