A THREAD FOR NEWCOMERS!

and, in essence, is not a full-fledged participant in the game.

Because it's the dumbest bot, not a DeepMind or OpenAI neural network. And no matter how friendly you become, you won't become a musician. No matter how many heroes you give it for ransom, nothing will change globally. Because the AI can't conduct battles, and the AI can't build chains.

They still won't be suitable for musicians, that's true. Having several AIs at 200% is certainly not a good thing. But the possibility of at least conditionally making the AI a full-fledged participant in the action. By the way, I assume you haven't seen how the computer plays for a while, starting with a small respawn. It's quite adequate. Especially when each computer has several heroes – they quickly clear the respawn and go to the neighbor. But what we see in streams, say, on 6LM10a, is indeed a pathetic sight. One hero wanders around the respawn. In general, valiant online players put the already not-so-smart AI in a terrible position.

(by the way, here's an idea to throw in for new features in HotA: expand the pool of AI heroes, even just for online play. Or ND+)

No, it's simple here. One number in memory changes for this, and that's it. I have quite a few of these "mini-plugins" ready. At some point, I'll put them all together in a release. I hope so, of course. I'm not counting on HotA in terms of improving the Heroes, because I have a much more conservative approach, but unfortunately, much smaller opportunities. It's quite suitable for HD+, though.

You definitely know the "idea" better than anyone else. And maybe even better than the creators themselves. I don't even doubt it)

Of course, that's the beauty of it.

Seriously, the game is largely untapped for one reason – weak AI. Theoretically, this situation can be changed in the future.

Added 26 minutes ago

What is this idea?

I'll try to answer this question approximately. It's roughly what Gelimer wrote about.

In it, players essentially compete as they do offline, only arranging a "Final" at the very end, after which they go their separate ways. I miss the confrontation, with city battles, sieges, positional warfare, destruction of chains, gradually pushing the opponent off the map, etc.

That is, it's roughly the kind of game that existed when we were all just starting out in Heroes. The AI was a full-fledged player back then, because we all played poorly, you can't deny that? Well, the point is not in the AI, of course, but in a sufficiently high density of full-fledged players. Not like sardines in a can, of course, but not as in online play.

As in Vysotsky's song:

And the cold past will speak
Of campaigns, battles, and victories.

About campaigns and battles, not endless beating of defenseless monsters.

I'm really surprised by the players – they complain about a number of things in the game (and write suggestions for its "improvement"), not realizing that a huge number of these things are determined by the format of the game they play. You can list them endlessly – too weak skills, too strong skills, etc. They want to replace the Angels in the reserves with who knows what, or some other "wonderful" wish.
Returning, again, to the original idea of the game, I believe that the developers meant a balance between the creatures from the player's castles and the creatures from the outside. And such a balance exists with a normal game format (simply increasing the ratio of the number of castles on the map to the number of outposts and reserves – to the level of balance, it also becomes advantageous to develop castles).

Sometimes thoughtful and understanding Heroes players formulate this as "Heroes were designed for M maps." In fact, it's not for M maps, of course, but simply for a sufficiently high density of full-fledged players (on XL – 6-8).

You've gone off on a tangent. What are you talking about? Luck, approaches, online players, tournament stages. Everything is mixed together.

Leader, don't panic. Everything is under control.