The Japanese military had basically two philosophies for the end of the war. One was semi-rational: the United States, if it met with a lot of resistance for the Japanese homeland, would be driven to the negotiation table by the amount of blood spilled. The other was not rational: that somehow, through sheer force of will and honor, the Japanese would not only repel the United States but somehow, magically, win the entire war.

Both of these philosophies were based on the idea of Japan just taking on the United States. They also were based on the assumption that Japan would be able to some degree still be supplied from its holdings in Korea and Manchuria, which were essential for its war and domestic industries.

They were also both based on the idea of Soviet neutrality. Well before August 1945, the Japanese military had concluded that if the Soviets entered into the war against them, they would not be able to fight them off. Why not? Because a) the Soviets would likely seize Manchuria and Korea, and thus cut the Japanese off from much-needed supplies; b) the Soviet army was immense; and, importantly, c) Stalin was not going to be driven to the negotiation table by a sea of blood, because he was not the head of a fickle democracy.

Losing Manchuria by itself would have been a troublesome thing only because of the resource losses. But more important is what it symbolized: a full Soviet invasion would likely also eventually reach the Japanese home islands, and whatever tactics they thought might work against the United States would certainly not work against both the United States and the Soviet Union. To hold out was to prolong the inevitable, which was loss and occupation. And occupation by the Soviets was a more horrific thing to contemplate than occupation by the United States, as well — as the case of Poland had already shown.