On the ground of “Rainforest Realism” philosophy of science and the distinction between parsimony and simplicity, this thesis conceptualized “model set” consist of overarching grand-theoretical assumptions, mid-level models, and purposed to capture various “real patterns” in international security politics. Via the overlooked bait-and-bleed concept, I developed a novel and parsimonious Neo-Offensive Realist “model set” that explains why some states deliberately choose to initiate or enter an expected attrition war over an ex-ante bargaining settlement. My theory posits that rational major powers choose bait-and-bleed, especially indirect bleeding, if possible when they assess dyadic, composite denial favorability because the breakthrough is more likely to be contained, and the favorable attrition rate in a contained offensive increases net bleeding gain in a war. Vise versa, major powers choose blackmail-blitzkrieg when they access dyadic, composite offense favorability because the costly breakthrough is likely to succeed quickly, and the profitable exploitation is made plausible. Crucially, baiting is a rational but inefficient strategic behavior. The bait-and-bleed strategy demands the ex-ante concealment of private information. Depending on the informational nature of the rival strategy, the ex-ante bargaining outcome of a revisionist with an initial strategy of bait-and-bleed varies from a certain peace, possible war, to a certain war. After setting up the theory, this thesis designed plausibility probe consists of two selected sets of Ecksteinian critical cases: Soviet and Chinese bait-and-bleed efforts in the unfolding of the Third Indochina War, 1976-1979, followed by the British and Soviet bait-and-bleed efforts in the unfolding of WWII in Europe, 1939-1940. The fine tracing of causal processes in empirical evidence confirms my theoretical propositions.