Why Won't You Die

[mortalmadman]

My god. I was just playing a game. Tiny, 3 rivals, and 2 islands. Mid-ancient age. I deceided to go for a quick attack and take some workers fast and take countrol of my island. I pretty easyly take most of his cities withous too much troble. Amazingly easy. I got GREAT position. But he had only 1 city left. 1 pop in it. Should be easy. I have plenty of archers and he has spearmen and archers. I march my men to his city in the desert. I make sure not to attack as to give him the river advantage. It took me allthought the rest of the ancient age and the beggining of the middle ages. Why was it SOOOOO tough. I lost TONS of archers to a few stupid spearmen. Why. Is there something I might of done wrong. Dose this happen to you. It was bad this time.

 

[Tomoyo]

How many archers did you send? Was the city a capital (and thus, maybe hiding walls)? Was it on hills? Did you send all the archers at the same time? Did you cheat, angering the RNG gods?

 

[mortalmadman]

How many archers did you send? Was the city a capital (and thus, maybe hiding walls)? Was it on hills? Did you send all the archers at the same time? Did you cheat, angering the RNG gods?

I don't know? 10 atleast. But probaably 17. Of course it was a capitol, it was the last city. It wasen't on hills. It was on desert, plains, or grasslands. Plus, not at the same time. But I sent big groups. I sent 5 or 6 in one group.

 

[Tomoyo]

5 or 6? Not enough to assure victory.

My guess would be that there were (invisible) walls inside the city, and maybe you were attacking across a river.

 

[Bluemofia]

I think the spearmen had values to represent reality

 

[Prince David]

I've noticed this same phenomenon several times. It makes me wonder if there isn't some factor in the battle calc that kicks in occasionally to respresent the determination and despairation of the last vestage of a civilization. More likely there's some component of the RNG that determines an "overall success factor" for each player in a given turn. Each player is going to have a good, neutral or bad turn, with most being neutral. Very, very often I have turns where NOTHING goes right on any front - 10 trebuchets and only one defender damaged by one point at one city (or worse, 10 bombers), while elsewhere 3 vet knights wiped out attacking a 2 HP spearman traversing a hill. Or an entire stack that can not make a dent on a small unimportant city. So, you get a "bad" turn, they get a "good" turn, and you can't catch a break. I may be blowing smoke, but it happens too consistantly to be bad numbers from the RNG at each and every event in the turn.

 

[JoeBas]

But I sent big groups. I sent 5 or 6 in one group.

This is your mistake. It's a pop 1 city in the desert, before drafting. It's not going to be generating new defenders very quickly. It's faster to save up for a few turns and send 15 or 18 at once, than to send 5 or 6, lose them, let the AI heal, send 5 or 6, lose them, let the AI heal, send 5 or 6, lose them, let the AI heal... overwhelm them with the big group so they don't have time to heal, and voila.

 

[Offa]

6 vet archers v 3 reg spearmen fortified in a town on flatland gives an 82% chance of taking the town. Therefore chances of failure are not low.

If you attack with 8 archers chances increase to 97%.

 

[Crakie]

It seems that RNG-caused bad luck in Civ3 does not come as isolated incidents but as small clusters. I have lost far too many elite knights to reg spearmen (without any particular defensive advantanges) or experienced far too many turns in which a stack of artillery did not hit a single shot. Maybe it's a matter of programming, maybe it's a statistical phenomenon.

Interesting question: if there are stretches of bad luck, are there streches of good luck as well? I don't think experienced players would benefit from them because they would think twice before attacking 3 fortified spearmen with 4 archers, yet in their mini-RNG-golden age they could win.

 

[Pentium]

I once captured 4 cities, each fortified with one MI, with 2 MA. I think that's a good luck. Of course I didn't think twice, I just attacked.

Edit: that was in one turn, so no healing.

 

[JoeBas]

6 vet archers v 3 reg spearmen fortified in a town on flatland gives an 82% chance of taking the town. Therefore chances of failure are not low.

If you attack with 8 archers chances increase to 97%.

Yeah, but he didn't say anything about...

1) there only being 3 spearmen -- in fact, he specifically said that they have "Spearmen and archers"

2) They only being regular spearmen. In fact, judging by the fact that he's repeatedly attacked and failed, I'd be willing to bet they're elite by now.

3) The town being on flat land. It's entirely possible it's on a hill too.

The basic rule of common sense is what applies here... if you send 6 archers in, and they get slaughtered, and the enemy just gets promoted and heals, and you send 6 archers in, and they get slaughtered, and the enemy just gets promoted and heals, and you send 6 archers in, and they get slaughtered, and the enemy just gets promoted and heals... chances are that it's not just a RNG thing.

What word has a definition of making the same mistake over and over and over again and expecting a different result?...

 

[IglooDame]

What word has a definition of making the same mistake over and over and over again and expecting a different result?...

I think the word is "ChieftainLevel"

 

[Offa]

Yeah, but he didn't say anything about...

1) there only being 3 spearmen -- in fact, he specifically said that they have "Spearmen and archers"

2) They only being regular spearmen. In fact, judging by the fact that he's repeatedly attacked and failed, I'd be willing to bet they're elite by now.

3) The town being on flat land. It's entirely possible it's on a hill too.

My point was that that was the best hope he could have. He did say the town was on the flat but referred to "a few" spearmen, which I take to mean more than 2. The spearmen in practice would almost certainly have been regulars at the time of the first attack as the AI hardly ever build barracks early on. He didn't specify that the attackers were vets but I gave him the benefit of the doubt here.

Depending on the circumstance attacking an AI capital with 6 archers may often be justified, but just remember it may not work. He was unlucky to find 3 defenders, as early on in a last hideout town 2 is probably more likely. Having failed the first time, subsequent attacks clearly need more units as some of the defenders will probably be vets, so if at all possible you should try to take the town the first time.

 

[Bluemofia]

attack en mass. always works if you have enough. (20 impis swarmed and killed 3 cavalry in a size 12 city)

 

[Desertsnow]

I've noticed this same phenomenon several times. It makes me wonder if there isn't some factor in the battle calc that kicks in occasionally to respresent the determination and despairation of the last vestage of a civilization.

I wouldn't doubt this myself. In one recent game I played, in which my Egyptian civilization that had spead over most of a large continent was reduced to a single city (Alexandria) in a three-front war, that last city fought like a rabid hyena, staving off tanks, cavalry, and elephants with a few veteran to elite spears and a cavalry or two for several turns before it finally went down.

 

[allhailIndia]

I have noticed that phenomenon too...One conscript Rifleman in a city on grasslands took out an entire 4 Vet Cavalry Army and 2 tanks!!

However, in this situation, I would have tried to bring in swords and catapults to try and take them out..

 

[mortalmadman]

attack en mass. always works if you have enough. (20 impis swarmed and killed 3 cavalry in a size 12 city)

I LOVE your sig

 

[The Last Conformist]

It seems that RNG-caused bad luck in Civ3 does not come as isolated incidents but as small clusters. I have lost far too many elite knights to reg spearmen (without any particular defensive advantanges) or experienced far too many turns in which a stack of artillery did not hit a single shot. Maybe it's a matter of programming, maybe it's a statistical phenomenon.

Real randomness is noticeably more "streaky" than most people would think.

 

[FinnMcCool]

I've noticed this same phenomenon several times. It makes me wonder if there isn't some factor in the battle calc that kicks in occasionally to respresent the determination and despairation of the last vestage of a civilization. More likely there's some component of the RNG that determines an "overall success factor" for each player in a given turn. Each player is going to have a good, neutral or bad turn, with most being neutral. Very, very often I have turns where NOTHING goes right on any front - 10 trebuchets and only one defender damaged by one point at one city (or worse, 10 bombers), while elsewhere 3 vet knights wiped out attacking a 2 HP spearman traversing a hill. Or an entire stack that can not make a dent on a small unimportant city. So, you get a "bad" turn, they get a "good" turn, and you can't catch a break. I may be blowing smoke, but it happens too consistantly to be bad numbers from the RNG at each and every event in the turn.

And Offa offers:
Offa
1.1.2




Join Date: Feb 2002
Location: Mercia
Posts: 612 6 vet archers v 3 reg spearmen fortified in a town on flatland gives an 82% chance of taking the town. Therefore chances of failure are not low.

If you attack with 8 archers chances increase to 97%.


I am dying to find out how you know these numbers, Offa. !!

I'm a fanatical probability cruncher, and what Prince David writes looks like a real possibility, but I have my own theory. For any games I've created (FRP and such mostly) I always use probability CURVES, but I suspect Sid doesn't. I get this impression especially in the ******ed diplomacy from previous Civs, but also from combat. It seems to me that there is an equal chance of all possibilities happening in combat. If it were ONLY the terrain odds modifiers, you wouldn't see pikemen slaying Panzers nearly so often

Offa, WHERE IS THE ODDS CALCULATED?

 

[Tomoyo]

The stats of the units...