A tax policy looks like it helps cities survive, but is it really better?

Players with strong cities may pick one option more often, which makes it look like the cause of their success. This entry shows how to compare the choices fairly inside groups of similar cities, and why the plain totals can point the wrong way.

Why this works

Strong players pick differently from weak ones, so a policy can look good only because good players like it. Comparing cities of similar standing, and only then combining the gaps, takes the skill out of the comparison.

When not to use it

It cannot save you when the groups differ in ways you did not record, or when few cities fall into each band. Then the safe move is to give the policy to some cities by lottery and watch what happens.

In the full entry

• What to build, step by step
• Starting values and the reason for each
• 3 common mistakes and the fix
• 26 lines of example code with 3 checks that pass
• How your AI shows you that it works

Say this to your AI

Use the Spielerando entry "A tax policy looks like it helps cities survive, but is it really better?" (read_entry strat-policy-choice-simpson-trap) in my game. Build it step by step, and show me how you checked that it works.

Or skip this and connect the MCP: your AI then finds all of this by itself.Connect the MCP