Is the combat predictor mechanism off...?

[bcaiko]

Hey folks,

I've been noticing a lot in the past few days that the combat predictions the game is giving me when I attack are, more often than not, wrong. Especially in multiplayer games. I know there's an element of randomness to it, but I must be having a TON of bad dice rolls if that's the case. In almost every instance where I've been told by the predictor that a "DECISIVE VICTORY" was coming, it was, errr, not so much.

Anyone else been noticing this? Am I just having a rediculous amount of bad luck? I don't play with any mods, and I'm beginning to doubt anything the combat predictor is telling me.

 

[headcase]

Sounds like bad luck, unfortunately this game doesn't give a bell-curve or distribution of probabilities like the popular Civ 4 mods do.

Which units were involved? Was it a matter of the enemy still having a couple of HP after the attack, or was the enemy actually doing more damage each fight?

 

[vexing]

check out the war academy's mechanics of combat

 

[TPQ]

I was under the impression that the combat outcome prediction mechanism was a little bugged so it showed you what your odds were from where your unit currently is as you mouse-over, rather than where it will be when it attacks, which can lead to misleading odds being shown?

 

[bcaiko]

I do realize that the prediction can swing by a couple hp either way, but EVERY time the game was telling me I was going to finish off a wounded unit, it would survive and then use the insta-heal promotion (and I was Japan, so they didn't have Bushido). My swordsman (for which I had plenty of iron) couldn't even finish off archers.

I was quite irritated.

 

[gunnigolfer]

It's probably selective memory. You tend to remember unlikely events like losing a battle with >90% odds of winning but when you win such a battle it is soon forgotten. Poker players experience this a lot. Many poker player consider themselves so unlucky, always getting ''bad beat'' when in reality they they get their fair share of wins and losses. This is probably even more pronounced in CIV because the human plays better than the AI and rarely does combat against all odds.

 

[vexing]

I do realize that the prediction can swing by a couple hp either way...

it goes from approximately 2/3rds the displayed damage to 4/3rds. in a swordsman vs archer situation it might say 10 and still be possible to do as little as 6

 

[KrikkitTwo]

THe problem with things like "Decisive Victory" is they are trying to capture a few factors

1. Amount of damage We do to the enemy
2. Amount of damage the enemy will do to us
3. Chance we have of dying
4. Chance the enemy has of dying

So.. what if the enemy will definitely die, but I will take major damage... or the enemy won't die just take major damage, but I will only take minor damage. Or we will both only take minor damage.


I'm not sure what they use .... but I think it may only be #1+#2... ie if I will probably do more to them then they will to me then that is a Victory... even if it doesn't kill them.

 

[timtofly]

If you are used to 4, then don't trust V's predictor. It is not a guaranteed prediction. I have recently lost some highly promoted Stealth Bombers to cities, and the predictor showed everything A-ok. I would say that I have also won more stalemates than I have lost. It is a gamble that does go your way more often than not.

 

[hung_h]

My battles never result in much difference than the predictor. Once in a while a unit may take on more damage, but nothing like an archer beating a swordsmen.

 

[Venger]

I was under the impression that the combat outcome prediction mechanism was a little bugged so it showed you what your odds were from where your unit currently is as you mouse-over, rather than where it will be when it attacks, which can lead to misleading odds being shown?

This is where I've seen most of the problems with the predictor - not properly calculating percentages especially Great General and attacking over river. A decisive victory can turn into a decisive loss VERY quickly due to this bug. And it's only been a year now to fix it...

 

[spfun]

I've had problems with this ever since they patched in the added randomness many patches ago, it seems off to me. Maybe a bug in there somewhere, or I'm just consistently unlucky .

 

[MadDjinn]

one thing to pay attention to is your own health, wrt the potential damage you're going to take.

It may be a 'decisive victory', but if you're going to be down to 1-2 health, it could flip into the opposite direction. do to the randomness of the damage taken/given.

 

[bcaiko]

one thing to pay attention to is your own health, wrt the potential damage you're going to take.

It may be a 'decisive victory', but if you're going to be down to 1-2 health, it could flip into the opposite direction. do to the randomness of the damage taken/given.

I was playing as Japan, so I wasn't really worried about health effects. I just got tired of almost finishing off a goshdarned archer (!) and then it insta-heal promotes back to full health.

I may have just been having a very unlucky streak, but it did begin to feel like the game was feeding me some bad info.

 

[PieceOfMind]

When we have source for the dll I might look at modding this.

 

[KrikkitTwo]

I do think they need

1. A message that only appears when the enemy will definitely be killed (100% chance)
2. A message that only appears when your unit has no chance of dying

Those would help.