Currently Online 
🇷🇺🇺🇦 Since Russia invaded Ukraine we looked for a framework to forecast interstate war duration
— Alex Isakov (@x1skv) January 18, 2024
Bennett & Stam's work provides exactly this - empirical benchmark, which avoids polarized 'brief vs forever' calls
But (B&S,1996) is mute on likely outcome
We address this now... pic.twitter.com/jib1kA1411
Based on CoW data, we estimate that over the 200 two centuries, most wars lasting more than 2years result in a loss or stalemate for the initiator
— Alex Isakov (@x1skv) January 18, 2024
The likelihood of loss rises from 33% in the first 0.5Y of war to 50% by mid-3Y — and grows further pic.twitter.com/aLv1bxFVGw
Still, I think, the core pattern of declining probability of victory for initiator of war declining with war is stable across history - and if anything is becoming even more pronounced over centuries
— Alex Isakov (@x1skv) January 18, 2024
Why is that? pic.twitter.com/e8dCxzF7c6
The longer a war continues, the higher the risk of a loss or stalemate for the initiating side
— Alex Isakov (@x1skv) January 18, 2024
In our sample of 96 interstate conflicts from 1816-2007, the initiator avoided defeat or stalemate in only 4 of 12 conflicts that lasted 2+ years. 2 of those 4 took place before 1900. pic.twitter.com/gV17hHZxBv
Essay Pro