I agree with Zhukov – Japan is trying to triple its income over the first four turns, so you have to ask yourself what kind of return on investment you can expect from different attacks. If Russia abandons Siberia entirely, or only leaves 2 infantry there, then you can steamroll the infantry with forces in Manchuria plus one or two planes that weren’t really needed elsewhere, and pick up an extra 6 to 8 IPCs of income every turn – that’s a good return on investment. You’re committing about 25 IPCs of low-value troops and losing about 3 IPCs worth of troops to boost your income by 7 IPCs per turn. It pays for itself, quickly.
Same thing with attacking Hong Kong, Malaya, and Borneo – Hong Kong and Borneo are practically undefended, and Malaya only has four infantry defending it. So you bring down a few loaded transports to a central region of the board, in a sea zone where your capital ships can easily guard them. You commit about 50 IPCs’ worth of material, expect to lose about 6 IPCs worth of troops, and you boost your income by 10 IPCs per turn, plus you knock out UK Pacific and ANZACs’ national objectives. Very good return on investment.
But if you attack a well-defended Siberia, now you’re looking at committing 70 IPCs’ worth of high-value troops – you need enough infantry, tanks, and aircraft to kill 12 Soviet infantry plus 2 Soviet AAs. You won’t be able to gather all of that force all at once no matter what you do, so your progress will be slower, so your income only goes up by an average of 3 IPCs per turn in the crucial early turns. You can expect somewhat higher casualties, since you’re fighting a real defensive force. So now you’re committing 70 IPCs and losing 15 IPCs to gain 3 IPCs per turn. Very bad return on investment. Even if you technically make a profit of 1 or 2 IPCs over the course of the game, that’s not nearly enough to justify the attack. You need opportunities for big, fast, reliable profits, not opportunities for small, slow, unreliable profits.
Some people try to include the value of the Soviet defenders as part of this calculation, saying, oh, well, I killed off 50 IPCs’ worth of Russian infantry and AAAs, so it’s still a good trade. Wrong. Those defenders can’t be in Siberia and Moscow at the same time. If they stay to guard Siberia, you’ve removed them from Moscow just as effectively as if you killed them. If they retreat toward Moscow, then you invade Siberia, after the defenders are gone…because then you get a nice fat return on investment, i.e., you can sweep up the entire Soviet Far East with just two or three units. There’s never any reason to assault a stack of 14 Russian defenders in a Siberian territory. You will always have a juicier target somewhere else – or, if you don’t, that means someone has already won the game.