HoMM3 Quick Combat - anybody know how it's calculated?
Varn
'quick combat' option since that's the only way to fairly resolve things by
e-mail. However, I would really like to know how this quick combat method
calculates losses in combat based on troop types etc. I assume this is the
same method which is used for the AI heroes when they attack wandering
monster stacks, but any info appreciated.
Thanks,
Varn
Gus Smedstad
No, it's not the same as the method the AI heroes use.
"Quick Combat" is exactly the same as running the combat with Auto
Combat on. The same code is executed, it's just run with the animations
off.
The AI heroes use something much simpler, for speed reasons.
- Gus
Varn
Gus Smedstad wrote in message <387A2B61...@nwcomputing.com>...
>Varn wrote:
>No, it's not the same as the method the AI heroes use.
>
>"Quick Combat" is exactly the same as running the combat with Auto
>Combat on. The same code is executed, it's just run with the animations
>off.
Ah, that sounds at least somewhat promising. Assuming I have AI spellcasting
disabled (to save mana for an importnant battle, e.g.) does this mean my
hero won't cast spells in quick combat? And if it's on, he will?
>The AI heroes use something much simpler, for speed reasons.
Hmmm. Always wondered how they manage to take out all those wandering archer
stacks without suffering noticeable losses.
Regards,
Varn
Gus Smedstad
>
> Ah, that sounds at least somewhat promising. Assuming I have AI spellcasting
> disabled (to save mana for an importnant battle, e.g.) does this mean my
> hero won't cast spells in quick combat? And if it's on, he will?
That would be reasonable. Unfortunately, to my chagrin, I discovered a
while back that I had coded it so that your heroes will cast spells in
quick combat, regardless of the AI setting. I don't know what I was
thinking :).
> >The AI heroes use something much simpler, for speed reasons.
>
> Hmmm. Always wondered how they manage to take out all those wandering archer
> stacks without suffering noticeable losses.
>
Actually, it attempts to simulate that. The AI method simulates 4
"normal" rounds, in which archers can shoot without retaliation. The
faster a melee creature is, the sooner it gets to participate in
combat. After that, the result is determined by a combination of the
ratio of the two forces and a random factor.
- Gus
Varn
Gus Smedstad wrote in message <387B6F94...@nwcomputing.com>...
>Varn wrote:
>>
>> Ah, that sounds at least somewhat promising. Assuming I have AI
spellcasting
>> disabled (to save mana for an importnant battle, e.g.) does this mean my
>> hero won't cast spells in quick combat? And if it's on, he will?
>
>That would be reasonable. Unfortunately, to my chagrin, I discovered a
>while back that I had coded it so that your heroes will cast spells in
>quick combat, regardless of the AI setting. I don't know what I was
>thinking :)
That's a shame. Any chance of it being patched? ;-)
>> >The AI heroes use something much simpler, for speed reasons.
>>
>> Hmmm. Always wondered how they manage to take out all those wandering
archer
>> stacks without suffering noticeable losses.
>>
>
>Actually, it attempts to simulate that. The AI method simulates 4
>"normal" rounds, in which archers can shoot without retaliation. The
>faster a melee creature is, the sooner it gets to participate in
>combat. After that, the result is determined by a combination of the
>ratio of the two forces and a random factor.
Does this really save much time compared with using the auto/quick combat
method? Doesn't seem to take too long to resolve those in quick combat mode,
even on my clunky old P266.
Regards,
Varn
Gus Smedstad
>
> >Actually, it attempts to simulate that. The AI method simulates 4
> >"normal" rounds, in which archers can shoot without retaliation. The
> >faster a melee creature is, the sooner it gets to participate in
> >combat. After that, the result is determined by a combination of the
> >ratio of the two forces and a random factor.
>
> Does this really save much time compared with using the auto/quick combat
> method? Doesn't seem to take too long to resolve those in quick combat mode,
> even on my clunky old P266.
>
I get asked this one often enough I ought to put it in a FAQ somewhere.
The AI uses the same method to estimate combats as it uses to resolve
them. Unless a human is involved, of course.
It does this because winning the game because the AI suicided against a
wandering monster stack isn't much fun. Hence the AI needs to be very
accurate in estimating combats that don't involve the player.
Inaccuracy in estimating odds when attacking the player is acceptable,
and even increases enjoyment at lower difficulties.
The AI needs to estimate a *lot* more combats than it actually fights.
Hundreds each turn, in fact. Even the current method should really be
tweaked for more speed, because the game can slow down when an AI hero
with a wide range of spells moves. The spellcasting portion of the
estimate takes too long.
- Gus
Geoffrey Tobin
...
> The AI needs to estimate a *lot* more combats than it actually fights.
> Hundreds each turn, in fact. Even the current method should really be
> tweaked for more speed, because the game can slow down when an AI hero
> with a wide range of spells moves. The spellcasting portion of the
> estimate takes too long.
I wish other designers of games i enjoy were as instructive as this.
Thanks, Gus!
--
Best wishes!
Geoffrey Tobin
Email: G.T...@latrobe.edu.au
WWW: http://www.ee.latrobe.edu.au/~gt/gt.html
Truthseeker
these exquisite 31 lines of text:
:===============:
>Varn wrote:
>>
>> >Actually, it attempts to simulate that. The AI method simulates 4
>> >"normal" rounds, in which archers can shoot without retaliation. The
>> >faster a melee creature is, the sooner it gets to participate in
>> >combat. After that, the result is determined by a combination of the
>> >ratio of the two forces and a random factor.
>>
>> Does this really save much time compared with using the auto/quick combat
>> method? Doesn't seem to take too long to resolve those in quick combat mode,
>> even on my clunky old P266.
>>
>
>I get asked this one often enough I ought to put it in a FAQ somewhere.
>
>The AI uses the same method to estimate combats as it uses to resolve
>them. Unless a human is involved, of course.
>
>It does this because winning the game because the AI suicided against a
>wandering monster stack isn't much fun. Hence the AI needs to be very
>accurate in estimating combats that don't involve the player.
>Inaccuracy in estimating odds when attacking the player is acceptable,
>and even increases enjoyment at lower difficulties.
>
>Hundreds each turn, in fact. Even the current method should really be
>tweaked for more speed, because the game can slow down when an AI hero
>with a wide range of spells moves. The spellcasting portion of the
>estimate takes too long.
I notice that AI heroes with the dimension door and fly spells take *very* long
to move, because they've got so many more possibilities, but that's okay,
because I like how intelligent they sometimes are (especially at higher
difficulty levels).
That intelligence made for an epic 10 month campaign in the Restoration scenario
where I had to defeat an enemy that had 5 Dungeons, and a super-powerful hero
with Expert Town Portal and a super army of 50 Black Dragons, 150 Scorpicores,
etc., while I had a Tower, Castle, and Rampart, with one super-hero with Expert
Town Portal and novice Dimension Door, and another super-hero with Expert
Dimension Door and novice Town Portal.
The AI super-hero would Town Portal to and attack any hero that got within one
move of any of the AI's owned Dungeons. Only my Dimension Dooring hero could
reach and take over the towns without getting stopped beforehand.
truthseeker