The IJA as a whole was fielding something around 2M troops by 1939, and at least until the Allied navies start sinking transport ships and stuff would be able to supply them.
The Red Army committed about 1.5M to August Storm in 1945 - considering the supply difficulties faced by that operation within a couple of weeks of it beginning this is probably close to the 'maximum effort' that the USSR can manage... everything has to come across the Trans-Siberian Railroad, which is too long to be given additional tracks along a significant length during the period this war would take place.
So quantitatively the two forces, committing everything they realistically can, are roughly equal.
Qualitatively, the advantage lies with Japan: at both the Changfukeng battle in 1938 and Nomonhan in 1939, Japanese forces were outnumbered by quite a significant margin (3:1 in the first case, 2:1 in the second) and still inflicted disproportionate casualties on the Soviets. Outside of those two battles, the Japanese have also been performing quite well in China (at least as far as pitched battles are concerned - the occupation itself was a bloody mess): for this we can look both to the string of victories in 1937-8 and the battle of Nanchang in early 1939 (where a 5:3 Japanese numerical inferiority is probably comparable to what they may face against a Russian spearhead). I will concede the Japanese did lose the Battle of Changsha that September, but the 5:2 inferiority they faced there is unlikely to be repeated in Manchuria absent some shockingly poor strategy or intelligence by the Japanese - always a possibility but not really helpful in considering an "average" scenario.
On the other side, the Soviets gave an average performance against the demoralised Poles in September (4:3 advantage in numbers, non-captured losses are around 3:4 as well), and a pathetic performance against the Finns in November. These fronts do not indicate the Red Army would do very well against the Japanese, and if in this scenario those fronts are active combat zones as well then the Soviet command will have its attention divided which never helps things.
As for the war itself, much of Manchuria north of the big cities like Harbin, Mukden and Hsinking is basically deserts and hills with little infrastructure - if the Russians are the aggressor in this scenario, they will be able to grab this area with little problem as the Japanese can't support a 2M army there, while it is right next to the TSRR. If instead the Japanese are the aggressor, they may hold it if an army is sent further north to cut the TSRR near Chita, or they may concede it and instead focus their effort on Vladivostok and Khabarovsk instead. If Japan does strike at Vladivostok, they will get a bloody nose on the fortifications, although may be able to take the city itself eventually (most likely after a siege and bombardment from the IJN). Chita is probably a doomed effort - they can't push forward any more than they can maintain a large defence in northern Manchuria. Most likely there is some sort of stalemate along the middle of Manchuria: Japan can win the battles, but they can't push forward to take advantage of their victories.
Of course, that is the 'maximum commitment' scenario, where both sides are only interested in that war and nothing else. Which is quite obviously not the case.
Japan's main consideration is China - they've been stuck there for two years now and were still as determined as ever to defeat them even in 1944 when the war elsewhere had gone totally to hell. Unless you have a way out for the Japanese (which is somewhere between extremely unlikely and outright ASB as long as the war is following a near-OTL course), this will be an important consideration for Japan.
During August Storm Japan was maintaining about 700k men in Manchuria, after four years of war with America and eight with China had been pulling forces out of the region. In 1939 most of that obviously hasn't happened yet, and while I don't have a number of Manchurian forces on me at the moment I believe 1M is a fair estimate. At the same time, Wikipedia gives Japan's strength in China in 1939 at just over 1M as well - half their forces are there. If reinforcements for the Manchurian front can't be found in the Home Islands, they will need to come from China.
Meanwhile, late 1939 to early 1940 was a period of Chinese success in the 2SJW - first the battle of Changsha which was a successful defensive battle, followed by the various offensives in Guangdong that almost broke the Japanese position in southern China. If we "reroll the dice", these battles could go much worse for Japan, and they certainly may do so if anything has been pulled out of China to fight the Red Army.
At this stage I leave it to you to decide if Japan could maintain both fronts - if Japanese qualitative superiority can fill in for lesser numbers, then a stalemate may be reached on both fronts. If it cannot, and the Japanese maintain their commitment to China, the Soviets may take over Manchuria throughout 1940 (though probably not Korea, where the terrain is much better for defence and much farther from the TSRR and Soviet supply lines). If the Japanese 'abandon' China and can get a white peace or similar out of Chiang, the possibility exists for a significant Japanese victory.
The Soviets don't really have any other commitments at this time - Poland, Finland, the Baltic States and Romania are all fairly minor fronts and the Red Army is large enough to handle them at the same time as fighting Japan (they can't support any more than 1.5M in the Far East, which is just a fraction of the Red Army's full strength). If Europe stays quiet, then the Red Army has a much larger pool of reinforcements to draw from, giving them the advantage in a long war (but they already had this anyway, with much greater industrial production capabilities than Japan).
The only European power that can influence this at all is Germany (absent the Western Allies attacking Russia or some other drastic changes). Ignoring the possibility of Eastern Poland giving Hitler a bloody nose (France is a very uncertain business as it is), and assuming Germany follows a near-OTL course, the earliest that Barbarossa can take place is 1941. As it was, Germany barely managed to pull it off after being effectively given huge piles of resources by the Soviets, a near-bloodless victory in France and a huge haul of captured equipment taken from the French. Absent any of that, Barbarossa can't achieve anything close to what it did OTL (which was close to the best-German-case), and in 1940 none of that equipment is available to the Germans, and they are also lacking the better part of a year's production. Pz2s against KV-1s is also a stupid idea.
If near-OTL holds in Europe, Germany will find Barbarossa a much more difficult prospect anyway - much of its success came down to the Red Army being part way through a massive expansion and reorganisation just as the invasion began - half formed units without a proper allocation of commanders and resources in exposed positions were up against the elite Panzer divisions. If Stalin has been in an all-out fight in the Far East, he's either not going to see the need for such a drastic expansion (if the Army is performing well against Japan) - in which case the Soviets are better prepared to fight Germany, or he's going to start the expansion much earlier. If the Tripartite Pact is still signed, Germany is marking itself as a clear ally to Japan and a clear enemy to the USSR - Stalin is much less likely to dismiss British intelligence warnings and will start doing whatever he needs to to be ready for Hitler's attack. From September 1940 at the latest.
- BNC