ChrissBBB
Cool.
If you want to really build a rigourous table, what you do is use the Senese and Vasquez 2008 Steps to War book, especially chapter 5.
He gives the average predicted probability of war incurring at least once in the history of two in the 1816-1945 period by types of issues (territorial, regime, policy) and the presence of war-inducing factors (alliances, rivalry, arms race, reccuring MIDs). It ranges from .201 (no fostering factors preent) to .931 if all the war fostering factors are there.
You could then produce predicted probabilities for specific dyads and compare them to the average.
An easier and less statistical path would be to build a table with the MIDs, and in each case note if the following conditions were present during and in the 5 year period before the MID
A) Territorial issues (you can get that from the Issue Correaltes of War dataset)
B) Outside Alliances (you can get that from either the COW alliance datasets or the Allaince Treaty Obligations project)
C) Arms races or mutual buildups (A bit more tricky as you have to code it, but Horn already has those coded for the period)
D) Rivalry (from either the Thompson list or the Diehl list)
E) Repeated Crises
Depending on the number of factors present you could then have x number of pluses to a base roll of 20% for MID or dyad to go to war.