In the Rubio versus Biden mod, it is possible to end up triggering an event in which Trump enters the race should you choose the wrong answer at certain points. Seeing as an independent Trump here would completely go against simple partisan leans, I had to develop a few methods for calculation.
To begin, Trump's baseline is set as around 12%. However, this isn't a flat 12% across all states. The method of calculation varies depending on if I categorize a state as a blue, red, or battleground state. This post is going to be dealing with the method for the blue states, since those are the easiest.
For each state, I consult the 2012 election results for partisan distribution and then add/subtract Obama's national margin (3.9%) so I get their pure relative lean. For example, CT was won by Obama with 58.06%. Subtract 3.9 from that to get 54.16%. From there, I multiply 54.16 by 15 percent? Why 15 percent? Because, according to the data at the time of the primaries, about 15 percent of Democrats had a favorable opinion of Trump, as per 538. Obviously, Dem support for Trump probably varies by state but data is nonexistent and given how solidly blue the following states are, I don't think it matters.
The formula for the Dem-half is 13% baseline multiplied by the relative lean of a state multiplied by 15 percent (an exception is made for Delaware, Biden's home-state, which grants Trump only a 10% multiplier among Democrats).
However, that only gives us Obama-Trump voters. What about Romney-Trump voters? This was probably the roughest part. I took the national polling average of 35 percent from around when the 2016 primary was at its most competitive as a baseline for what GOP support for Trump is nationally. However, GOP support of Trump also varies by state. So, what I did was I consulted the 538 polling averages from late-2015 to early 2016, and approximated a support level. For states where there isn't enough primary data, I just maxed out the support level to 35% for the sake of simplicity and also to make Trump more impactful. To make sure I didn't overall skew too high or too low, I calculated the average of these states' numbers, and noted their deviation from the mean. We can then subtract the deviation from that base of 35 percent to get a more proper number of support per state.
The formula for the Republican-half is 13% baseline multiplied by the relative lean of a state multiplied by the ratio of the state-level GOP support over Trump's national average of 35%.
You combine these two halves, and there's the proper state-by-state baseline for what Trump's support should hover around. I'll have to vary up the methodology for red and battleground states since their dynamics are different I'd imagine, but I'll get to that later.
Here's another article I found in my research which I may take advantage of later.