The Germans were obsessed with the fear that Soviet forces would evade them and escape into the nearly limitless expanse of the steppes. Their war plan from the start emphasized the need to encircle Soviet forces and prevent retreat.
The Germans called their dynamic method of warfare blitzkrieg (“lightning war”). It was characterized by penetration of a front at a chosen point (the “schwerpunkt”) with massive mobile forces, supported by airpower, and the rapid exploitation of that penetration by follow-on forces.
Blitzkrieg has significant weaknesses – principally, that the penetrating and exploiting forces are increasingly vulnerable to being cut off themselves if the defenders don’t panic, and attack the flanks of the exploiting columns. Even failure to defeat the exploiting forces can result in them being bogged down in battle on a new frontline. Historian John Keegan has pointed out that for blitzkrieg to succeed, it usually needed “cooperation” of some form (involuntary of course) from the opposing force. In Poland, that wasn’t really a factor because Poland was geographically vulnerable, hit by surprise, and invaded by the Soviets too. But in France that “cooperation” took the form of the Allied forces advancing into Belgium and Holland willy-nilly (intending to keep the battle as far from French soil as possible) and thus falling into the trap at Dunkirk. In Operation Barbarossa, that “cooperation” came in the form of Stalin trying to defend every inch of his newly-seized frontier and forbidding retreat.
By the time of the battle of Stalingrad, the Soviet high command (particularly Zhukov) had regained enough influence to modify Stalin’s influence. They still made a firm stand at Stalingrad, but Zhukov cold-bloodedly gave the defenders just enough support to hold on, and held back the rest to accumulate enough reserves for an enveloping counter-attack.