5.58 Brainstorming thread

[merlinp]

But if everyone would have fewer units why can't the AI still overwhelm the player with superior unit numbers?

[Nelphine]

I consider sprites borderline problematic anyway - they are super weak, so the only thing they can do is treasure hunt, which makes them a very niche unit that ends up feeling overpowered when it works, and awful otherwise (see all the strategic combat issues). If we broke sprite hunting I don't really think that would be a big gameplay loss. But in also willing to accept my opinion is wrong on this and it's a bad idea overall.

I completely agree with seravy about length of turns. They really aren't that bad, and I'm already doing games where I have less than 20 offensive units, often less than 12. Reducing number of units any more would mean I use less than one stack, and that defeats a lot of the game design.

[Ozymandyus]

3. Ignore it. Sapher plays much better than the average and in the hands of a "normal" player, this snowballing never happens anyway.
Might be a good idea actually, he has been beating the game with pretty much anything he ever tried. So it's not trivial to say this or that is specifically overpowered.

I think you have it there, Seravy.  You could chase this particular issue around in circles for a long time, but the law of unintended consequences says that either it would cause unforeseen problems, or else the top players would just adapt to a new strategy.

(Side note - thank you Sapher for sharing the videos.  It was interesting to see some ideas I would not have considered.)

[Seravy]

Did anyone observe the game display a different number of population being killed after conquering a city than the real amount?
I see a bug here that should be doing that - the variable storing the amount of people killed is also used as a temporal variable for clearing "production interrupted" records. Should be easy to fix by making the game store the amount to the proper location earlier.

Meanwhile, implemented the following changes :

5.58
-Damage to buildings and population is now applied even if the city had no defenders. (Unfortunately, it was far too easy to abuse the system by killing all defenders except combat summoned or confused ones, pass until the end of battle and then conquer the city after they disappear. Aside from this abuse, undefended cities should be extremely rare in the first place.)
-Each treasure budget point is now worth ½ gold or mana instead of 1. This doesn't apply to the 3 “starting” lairs placed next to the human player's capital.
-Treasure gold and mana amount is set to 0 if less than 50 was generated.
-Reduced the amount of very rare spell treasure generated from 2-5 to 1-4 per map.
-Towers that would have no spell otherwise will now always have a Common spell instead of either a Common or Uncommon.
-Adjusted Magic Power map setting (and renamed options) :
Min : 0.5
Low : 0.75
Fair : 1
High : 1.25
Max : 1.5
(this does not affect the amount of monsters guarding the nodes or the treasure budget.)

Ultimately, the engine for snowballing is raw gold, mana and power, and turning it into even more of the same. Reducing the benefit of conquest across the board should reduce snowballing problems significantly, without affecting "normal" gameplay much. Closing the hole on building destruction should make conquest snowballing less effective, while reducing raw gold and mana in treasure should make treasure snowballing less effective.
As this kind of snowballing is exponential, the nerf should have exponential effectiveness as well.

While the player still has an advantage at lair hunting efficiency, the AI is supposed to also have an advantage, better availability of resources/faster growth into higher tier units. This does make it a fair race under "normal" conditions.
If the player has to make a decision between fighting enemy wizards or hunting treasure due to the mana costs involved, the AI has a better chance to use their advantage as they don't get eliminated early. I'm hoping for that to be enough and won't change how lairs work at the moment.

Overall I'm unsure if this makes Chaneller stronger or weaker - on one side less resources means every mana crystal saved is all the more valuable, on the other hand less mana to use means less total saved by having the retort. I'm hoping for the latter to "win", as less available mana and gold means a higher ratio of it needs to be spent on activities other than combat spellcasting. You still need to buy buildings, raise skill, use mana for overland spells, pay maintenance etc. - speaking of which maybe the halved maintenance costs on Chaneller is excessive? Is there really need for the retort to even do that, ultimately leaving more unspent mana for combat spells? Considering the mana upkeep of all those undead fantastic creatures, the fact the retort helps making them by lowering the cost of doing so, and that the creatures themselves can be used to initiate more battles thus use Channeller even more sounds a bit too much of a positive feedback loop, though the upkeep reduction is actually the smallest part of it.

Also, is there anything else we wanted to do for this update or was this all of it?

[Sapher]

Also, is there anything else we wanted to do for this update or was this all of it?

In Apprentice menu flame strike says 30 strength fire attack. In combat info it shows 32 strength