I’ve been making steady progress for the last few days, but I’m stumped by a tricky little problem involving the calculation of uncertainty in response to a direct action by Actor2 on Actor1. Let’s say that Actor2 responds to a bit of gossip from Actor1 by saying merely “I see…”. Actor1 is justified to suspect that Actor2 is not being forthright with him; hence Actor1 should decrease his PHonest value for Actor2. I’ve got that covered. But how should his UHonest value for Actor2 change? On the one hand, we could say that this is a concrete act that clearly reveals something about Actor2’s Honest value; that means that our uncertainty about our result should decrease. On the other hand, what if the resulting change in P2Honest is really large? Should that not imply an increase in uncertainty?

Another problem arises from the initialization of P2 values. I have to start with the actual value of a virtue (good, honest, powerful) and then randomly assign a nearby value as a P2 value for others. I must also establish a U2 value. One of the first lessons I learned in physics was, “If in doubt about a problem, draw a picture of it.” So here it is in terms of the number line:

This really was quite easy to figure out once I drew the number line.