Attack anomaly?

Tharak

Chieftain
Joined
Oct 4, 2003
Messages
84
Location
Surrey, British Columbia
Has anyone else noted that fairly often, regardless of the strength of the unit used, the results of any one battle often turns out the same? By way of clarification, I tried this the other day.

Lone infantry defending a city. Attacked with veteran tank, knocked off 1 hp, tank dies. Reloaded, attacked with calvalry - identical result. Reloaded, attacked with swordsman - identical result.

I understand the whole idea about preserving the random seed, but why would a 16 attack tank be equally effective to a 3 attack swordsman?

I've also noted that the progression of hp loss seems to run in similar patterns. If my unit gets the first hit in, I will generally reduce the defender to 1 hp, and then my attacker will lose down to 1 or 2 hp before destroying the defender. Conversely, my attacker will often be reduced to 1 hp, then the defender will take 1 pt of damage before killing my attacker.

Anyone else note these things, or just me (BTW -using latest patch version of PTW)
 

Atotarho

Scourge of Europe
Joined
Sep 3, 2003
Messages
121
Location
Right here right now
Try this one.

Swordsman attacks spearman - dies horribly

Horseman attacks warrior - wins easily

reload

Horseman attacks warrior - dies horribly

Swordsman attacks spearman - wins easily


I have also seen my Elite attacker bash a defender down to 1 hp and then loose every one afer that until he dies :cry: (of cause its good when its the other way around ;) )

Its just the strange and mysterious RNG at work.
 
Joined
Nov 3, 2003
Messages
404
you tend to remember unlikely events. I mean if you have 300 normal battles and one where the dice rolled against you what do you remember?

Not the 300 normal battles.
 

FenrysWulf

Evil Norse Wolf
Joined
Nov 17, 2001
Messages
373
Location
Naperville, IL
I've noticed the patterns in attacks too. The computer likes to crush my spirit by having my attacker get the defender down to red and then lose four straight "strikes" to be killed. That's killer on morale. If you have preserve random numbers turned on, then yes, the computer has decided that you will lose the next attack, and no matter who you attack with, you'll lose. That's frustrating, and I turned that off. Now if I attack and lose seven times in a row I reload it and hopefully get a better outcome.
 

The Economist

Warlord
Joined
Nov 8, 2003
Messages
102
Location
New York
I used to do that thing with the "no preserve random seed", but then I realized that it wasn't fun because I could win ANY attack if I reloaded enough. You could kill a 20/20 Army of Mechanized Infantry fortified and chillin' in a fort on a mountain accross a river, and backed up by a radar artillery and within the defensive range of a radar station with a conscript warrior you got on the first turn of the game if you reloaded enough times. Yes most people would lose patience, but the fact that you WILL win the fight if you reload enough times, that just takes all the fun out of the game. But of course if its what other people like to do when they play the game, hey, doesn't affect me so they should do whatever makes them get the most out of the game.
 

Tharak

Chieftain
Joined
Oct 4, 2003
Messages
84
Location
Surrey, British Columbia
Originally posted by jeremiahrounds
you tend to remember unlikely events. I mean if you have 300 normal battles and one where the dice rolled against you what do you remember?

While this is very true, I was paying particular attention to it in my last GoTM. Being a Domination game, there was ample combat to watch. Once I finished, I took an old save and tried it a few times (not enough to get any statistical result..was just curious). I had a range of units from attack strength 3 to 16, each attack the same target (reloaded for each time). EVERY time the result was the same...did it 3 times with each unit before I got bored, but it was always the same. My attacking unit would take off 1 hp, the defender would knock me down to 1 hp, I'd do 1 more point of damage, then he'd either kill me or I'd retreat - depending on the unit. Still unsure why this happens..I thought a random attack value was generated by the computer, and your attack value vs defence value was used to determine the result...this makes it look more like the result is determined independant of this....just wondering if anyone has looked further at this, or am I just paranoid? :)
 

theGrimm

Reaper Man
Joined
Sep 5, 2003
Messages
125
Location
South Africa
I must admit, I generally accept the results, but one particular event nearly made me throw my pc out the window...

My Cavalry army (14Hp) attacks the last musketman in a size 11 city. Reduce musketman to 1HP without losing a single HP myself...then lose 14 straight rounds to lose my cavalry army. It didn't even retreat!!!
 

Lyonesse

Emperor
Joined
Nov 9, 2003
Messages
1,200
Location
Utah, USA
OMG! I think I would have succeeded in throwing the PC out the window! Hubby would have hated me, but I'm sure it would have felt good :)
 

The Economist

Warlord
Joined
Nov 8, 2003
Messages
102
Location
New York
WOW... stupid musketmen... :::shakes fist:::

I've noticed that Armies never seem to retreat even when they're down to 1hp. I'm not sure what the implications would be, and/or if it'd be unfair, but I think it might be reasonable for armies to have a feature that they automatically withdraw when they're down to 1hp unless fighting another army or a unit with more than 1 movement point.

I just don't think that armies should be able to die in one turn unless they're down to 1hp or are facing a fast unit. But that's just me- I guess having it the way it is just adds some more unpredictability and risk to the game, and allows for "miracles" (or in your case) "catastrophies".
 

Ace4nyC

Warlord
Joined
Nov 10, 2003
Messages
141
Location
NYC !!! Duh
Blah try this one for size....Advanced Armour attacks Pike man deals no damage dies....

10 knights attack 1 spear man all slaughtered not even 1 damage to spear man

Worrior Elite attacking tank Tank looses ( i had spare worrior in isolated city so i kinda let him loose) Combat is bugged im sorry ;(
 

sealman

Hater of Babylonians
Joined
Jan 23, 2002
Messages
3,326
Location
Manassas, VA
Originally posted by jeremiahrounds
you tend to remember unlikely events. I mean if you have 300 normal battles and one where the dice rolled against you what do you remember?

Not the 300 normal battles.

I remember some of the lucky ones. Like when my 2 Regular Musket men, fortified on a hill, in a size 3 city, across from the river against a HOST of Babylonian attackes, for what seemed like 20 turns, until I was able to provide reinforcements and counter attacked those Babylonains back into the stone age.

My Cavalry army (14Hp) attacks the last musketman in a size 11 city. Reduce musketman to 1HP without losing a single HP myself...then lose 14 straight rounds to lose my cavalry army. It didn't even retreat!!!
Were you allied with the Babylonians and attempting to take my city :lol:
 

anarres

anarchist revolutionary
Joined
Apr 22, 2002
Messages
6,069
Location
www.civ3duelzone.com
Originally posted by Ace4nyC
Blah try this one for size....Advanced Armour attacks Pike man deals no damage dies....

10 knights attack 1 spear man all slaughtered not even 1 damage to spear man

Worrior Elite attacking tank Tank looses ( i had spare worrior in isolated city so i kinda let him loose) Combat is bugged im sorry ;(
There have been lots of studies done with the RNG and it does indeed provide random numbers, and the battle system works as stated.

Note that each round of battle (1HP loss) is calculated as:

A / (A + D)

So, with a warrior it has roughly a 1/10 chance of taking a HP from a Tank (defense 8 with 10% bonus for terrain = defense 8.8).

The chances of taking 3HP (for example) form a tank in 3 consecutive goes is roughly 1 in 1000. If you allow the warrior to lose 4HP during the battle the odds are massively increased (simple binomial calculation but I am too tired to show it).

Telling us that the system is bugged because you saw a "1 in 1000" chance happening isn't anywhere close to convincing me, and it is unlikely to convince anyone else I'm afraid.

If this happened regularly and you could post saves to prove it I will happily believe you.
 

Ace4nyC

Warlord
Joined
Nov 10, 2003
Messages
141
Location
NYC !!! Duh
Originally posted by anarres
There have been lots of studies done with the RNG and it does indeed provide random numbers, and the battle system works as stated.

Note that each round of battle (1HP loss) is calculated as:

A / (A + D)

So, with a warrior it has roughly a 1/10 chance of taking a HP from a Tank (defense 8 with 10% bonus for terrain = defense 8.8).

The chances of taking 3HP (for example) form a tank in 3 consecutive goes is roughly 1 in 1000. If you allow the warrior to lose 4HP during the battle the odds are massively increased (simple binomial calculation but I am too tired to show it).

Telling us that the system is bugged because you saw a "1 in 1000" chance happening isn't anywhere close to convincing me, and it is unlikely to convince anyone else I'm afraid.

If this happened regularly and you could post saves to prove it I will happily believe you.


Well damn it isnt a bug ... Still it pisses me off every time ... But on the other hand computer never has that bad luck eh ?
 
Top Bottom