Posts from A program for calculating the strength of the guard on random maps in Heroes of Might and Magic III, called H3CalcGuards. Auto-translated

Yes, I understand everything correctly and that it is usually used during the opponent's turn. But what prevents you from remembering 2-3 guards and the objects behind them, and then, during the battle, see what's what. Often, you simply enter a battle and spend 2-3 minutes thinking about what chains to build afterward.

Updated the program: http://yadi.sk/d/Lg48jhvZ2BoJg (the new version will also be available for download via the link in the first post of the thread). Added the ability to switch to Russian and save the program's state after closing and reopening it. For those who edited the text file: the format of the main text file has changed: now it обязательно requires a Language section, and in all sections, the element name has been moved to the end of its line, and each element must have as many names as there are elements in the Language section. Also, you can now switch between different text files through the program's interface: to do this, in the H3CG_Settings.txt file, in the Version section, specify the names of the text files (if any of them do not exist, there will be no error; the program will automatically exclude them from the list). If the program detects more than 1 existing text file among those specified, an option to switch between text files will appear in its interface. When switching between text files, the program's state is saved as much as the similarity of the text files allows.

Wouldn't it be better to simply display the probability in this format instead of using checkboxes for "simple/extended view"?

Fairy: 42-68
Average: 55
28% 20-49
72% 50-99

Who really needs the exact probability anyway?

My job is to provide. :) I considered shortening the list to these groups, but then I thought that the probabilities are unlikely to be useful to anyone (it's simply clear that the closer the value is to the average, the more often it appears), and I didn't bother, leaving it as is. If needed, I can do it, but I would like to hear the requirements more precisely.

We were playing Jebus with weak guards, difficulty 160. Last time, when playing Necropolis, we got Pandora with 15 Vampire Lords for 14 Horned Demons, and it matched the table. This time, there were 19 demons, which is outside the allowed range. There's only one Necropolis town on the map. Is something else not being taken into account?

I entered it into the program – it seems like 19 is a normal number, while 14 is impossible: screenshot
The opposite happens if you don't specify information about the types of towns.
That is, on the map you sent, everything is fine, but where there were 14, it's possible that it's not working.

Tested the program. Using the Diamond template. In the same zone, 2 guards: - Gogs guarding the Dungeon and 2 groups of Dragon Stones (9k) - Griffin Pegasi guarding the Hive (9k) The program correctly identifies the guards, but with different monster strength settings for the first point (monster strength of the zone). For the Dungeon, option 3 (avg.avg+avg.weak) is suitable, and for the Hive, option 4 (strong.avg) is suitable. Generator settings - strong. For all Hives, option 4 is the most suitable, and there are no Gogs with 9k strength in it. Why is there such a discrepancy for guards in the same zone? If needed, I can provide a save file.

The thing is, besides the usual resource piles, the generator can also place "Random Resource" objects. If you open the map in the editor, you'll see that some piles are placed as is, but there are also many "Random Resource" objects. At the start of the map, all "Random Resource" objects are replaced with a randomly selected resource pile. However, during generation, the value of a random resource is 1500, while the values of specific piles are 750 for gold, 1400 for wood/ore, and 2000 for the rest. Accordingly, you need to consider that the guard protecting seemingly identical resource piles may be different. It is impossible to distinguish these piles in the game; "Random Resources" make up approximately half of all the piles.

What guard index should be used for the starting zone of a skirmish map when using Stronghold settings? I still don't understand whether it should be 5 or 4.

4. The strength of monsters in the area is average. However, in the treasure area, it's 5.

An object with a value of 6466, strength 4, in the list of guardians – mages, group of mages, group of djinns (common ones are omitted); I set the strength to 5, and the common ones appear in the same quantity as the upgraded ones.
Is this explainable?

Yes, that makes sense. The AI Value for Djinns and Master Djinns is not significantly different (884 and 942), so with small numbers, they might appear in equal quantities. However, the number of regular Djinns on the map is 8-16 (average 12), while for Master Djinns it's 8-12 (average 10). If the average number in a guard is less than the average number on the map, then that monster cannot be generated. Apparently, the average number of Djinns is slightly less than 12 (rounded to 12), which is why they are not placed. But Master Djinns have approximately the same average, and their minimum average number in a guard is 10 – and they are placed.

))) that's what I call an insight!
thank you

Why are the list boxes for monster type and other options empty when I launch the program?

It's normal for the monster type to be empty; after creating a group of objects, a list of valid monsters will appear there. The group of objects should also be empty; the user forms it by adding objects. However, if the selection fields are empty, for example, the object type, that's not normal and likely indicates that the text file wasn't loaded.