2. In a 49 Dem - 50 GOP Senate, I give Orman a 50/50 probability of caucusing with either party.
3. The effect is that the distribution is skewed from 50/50 to 51/49.
Below are three histograms (this time displayed with lines rather than the usual bars) that show GOP seat distribution in my current projections. The first chart is the one I am using in my model with the Orman assumptions. The second only uses the "majority party" assumption. Therefore in the second chart, Orman always chooses the Democrats when he has the ability to decide control. The third chart always assigns Orman to the GOP.