"Imagine a fight between a boxer and a street brawler. The boxer puts in his mouthguard, assumes his stance, and starts boxing. For which he immediately gets sand in his eyes, a knee between the legs, and a bottle to the head. That is, just as he was starting the fight, his opponent was already finishing it.
It's the same with the Muscovites — while we were building up our castle, gathering resources for construction and upgrading the 7th dwelling, they were quickly upgrading and buying lower-tier troops and using them to break open conservatories, hives, and then Dragon Utopias. As a result, by the 3rd week, when in our opinion the game was just entering its acute phase, for them, with control and parameters, it was already ending."
Essentially, the ideology of competitive Heroes is exactly that — to do everything needed as quickly as possible and earlier than your opponent. If you can break open a conservatory-hive in the 1st week, you do that. If you can take a Dragon Utopia on the 2nd week without critical losses, you do that. And so on. There is no standard tempo here; it's a fierce rat race.
Prioritization when moving across the map is very important. What to fight, what not to fight, and what to fight but not right now, etc. And while people have generally learned how to fight, when it comes to priorities, there's a complete disaster in some places.
Here, you need to calculate development several days ahead, imagine where a given hero will run, how they will set up and rearrange chains. Often you have to give up on many important things, say, awkwardly placed conservatories, hives, or Dragon Utopias. Because if there's somewhere to run and something to do, losing a day of tempo often doesn't pay off with an extra chicken coop taken.
And sometimes it's the opposite — you need to stand still like a dad and smoke for 2 days at one spot. Rarely, but it happens. You shouldn't be afraid of that either, as long as it's truly justified.
This can't be explained with words; it has to be shown in practice. Sometimes I offer people various puzzles on optimal pathing, and a minority manages them.