Fix The Rng

[Tsunami23]

PLEASE fix the Random Number Generator. In the past few games themselves I have lost 3 warriors to 1 warrior on plains, a numid to a warrior on plains, and stuff like that. I play MP, in a league so that just makes it that much worse. PLEASE fix this, I am pretty much fed up with it.

EDIT: And NO, not make it so that you know the outcome everytime!

 

[Arathorn]

What would you have them do? Make it not random, so that it's horribly exploitable? Or completely eliminate all randomness so that combat becomes like chess, completely predictable? The PRNG used in Civ3 is very nearly perfect -- it is far FAR better than any human would be at diplaying true random characteristics. The very fact that sometimes bad streaks happen is a 100% necessity in any good sytem.

Arathorn

 

[Jon Shafer]

You realize a Numidian attacking a Warrior is only 2:1.1 odds, right?

So roughly 36% of the time the Warrior should win.

 

[KabeDerlin]

CHanging the combat method would turn Civ3 into a tactical game, not strategical game. In this situation if a "luck factor" comes out of the RNG factor it can be attributed to the fact that the commander of the "disadvantaged troops" was better. Justloom throughout history of when numbers lost to knowledge. Alexlander the Great beat the Persians and he was outnumbered 5-1. Just accept that the RNG factor is the best method to keep Civ3 a strategic game

 

[Colonel]

Heres another thought Friendly fire, as for mountians think rock slides

 

[Aussie_Lurker]

Though I personally don't see the need to alter the current RNG, I definitely feel that there should be more methods for pushing combat results one way or another! These include

1) More Terrain Bonuses and Penalties.

2) Leader/Commander Bonuses.

3) Effects of being 'out of supply'.

4) Variable unit Sizes (for certain unit types, you should have variable sizes. Larger units have more hps, but cost more population and have greater support costs!)

5) Simple Tactics, like flanking, entrenchment, harrassing and charging and/or penalties/bonuses from using one unit against certain unit 'types'!

Many of the factors I have mentioned are a part of a number of 'dice-based' table top war games, and add very little to complexity of the game, and do not reduce the fun and overall 'unpredictability' of the dice-rolling system!

Yours,
Aussie_Lurker.

 

[KabeDerlin]

Simple Tactics

That's my only problem with your ideas. Otherwords the rest sound great! I esspecially like the Supply Line idea.

 

[Tsunami23]

What would you have them do? Make it not random, so that it's horribly exploitable? Or completely eliminate all randomness so that combat becomes like chess, completely predictable? The PRNG used in Civ3 is very nearly perfect -- it is far FAR better than any human would be at diplaying true random characteristics. The very fact that sometimes bad streaks happen is a 100% necessity in any good sytem.

Arathorn

No, simply make it so that 3 warriors can't beat 1 warrior every 5th time it happens, their is no point in that.

 

[The Omega]

Just imagine taht the warrior your fighting has a really good commander leading them, and the commanders leading your units are idiots.

 

[Aussie_Lurker]

Yeah, the simple tactics idea is 'conditional' on the kind of combat that they use! If its done on a turn-based 'mini-screen' thing, then it would work and I say it SHOULD be in (as it worked great in BotF!!) If, though, they stick with the current system, then I say tactics should simply be 'assumed' within the combat model!

Yours,
Aussie_Lurker.

 

[KabeDerlin]

Maybe you and I just differ on what we call tactics. I see flanking as a tactic. I see entrenchment (if we're talking about the kind used in WWI) as a strategy. But entrenchment is kinda the same as fortresses. or the same effect. CHarging I see as a tactic as well. So on that aspect I think SOME of those things would be cool. Just not all

But don't get me wrong. I love tactical games. In fact I consider myself much more of a tactician than a strategist. But Civ3 should clearly remain strategic. That's why I play it! Gotta get better at strategy...

 

[SwitchbladeNGC]

Superior training and superior weaponry have, when taken together, a geometric effect on overall militatry strength. Well trained, well equiped troops can stand up to many more times their lesser breathren than linear arithmatic would seem to indicate. -- Spartan Battle Manual

Taken from SMAC

 

[The Last Conformist]

No, simply make it so that 3 warriors can't beat 1 warrior every 5th time it happens, their is no point in that.

There's roughly a zillion better ways to achieve that that do not involve messing with the pRNG.

 

[Shirastro]

I think that the problem here is that some people tend to remeber only when thy loose 3 wariors to one, but forget when their swordman beat the rifelman

 

[Tsunami23]

There's roughly a zillion better ways to achieve that that do not involve messing with the pRNG.

And how is that?

 

[sir_schwick]

I figured out why the combat results in civ can be screwy on an individual fight basis. If you follow the rough arthmetic of A/(A+D) was the attackers chance to win a round, a Regular Tank should usually bet a Regular Infantry. At an infinite number of trials, or rounds of combat, the Tank would win exactly 8 of every 13 rounds. The less number of trials, the more results can deviat from the expected results. The more trials, the closer it comes to the predicted result.

In Civ 2 and SMAC, units had 10 -40 HP. This meant that combat usually lasted a minimum of around 20 rounds. In Civ 3, units have 3 to 5 HP. This means that combat is at most 9 rounds. This means that Civ 2 has at least 2 times as many combat rounds, probably a lot more. This also means that Civ 3 has greater deviation from expected results then Civ 2.

So, in conclusion, to make combat more credible, and lower the occurence of , you just need to increase the number of HP. I would mod that in, but I do not have the time.

 

[Gingerbread Man]

In Civ 2 and SMAC, units had 10 -40 HP. This meant that combat usually lasted a minimum of around 20 rounds. In Civ 3, units have 3 to 5 HP. This means that combat is at most 9 rounds. This means that Civ 2 has at least 2 times as many combat rounds, probably a lot more. This also means that Civ 3 has greater deviation from expected results then Civ 2.

So, in conclusion, to make combat more credible, and lower the occurence of , you just need to increase the number of HP. I would mod that in, but I do not have the time.

Hence making a more predicable combat system the easiest thing to modify into the game. just double/triple the hit points, or have a non-linear scale e.g.
Conscript - 2
Regular - 3
Veteran - 5
Elite - 8

this would have a similar effect of messing with the pseudo-RNG, but without actually un-randomising it.

 

[sir_schwick]

Hence making a more predicable combat system the easiest thing to modify into the game. just double/triple the hit points, or have a non-linear scale e.g.
Conscript - 2
Regular - 3
Veteran - 5
Elite - 8

this would have a similar effect of messing with the pseudo-RNG, but without actually un-randomising it.

I like your thinking. Combine both the non-linear and higher number principles. It would really increase the importance of Barracks and elites.

Conscript - 5
Regular - 10
Veteran - 20
Elite - 40

Of course for Civ 3 purposes, bombardment weapons would need to multiple their Fire Rates accordingly, and you would have to turn off combat animations for mod purposes(each unit would take 5 mins to fight).

 

[Babbler]

I figured out why the combat results in civ can be screwy on an individual fight basis. If you follow the rough arthmetic of A/(A+D) was the attackers chance to win a round, a Regular Tank should usually bet a Regular Infantry. At an infinite number of trials, or rounds of combat, the Tank would win exactly 8 of every 13 rounds. The less number of trials, the more results can deviat from the expected results. The more trials, the closer it comes to the predicted result.

In Civ 2 and SMAC, units had 10 -40 HP. This meant that combat usually lasted a minimum of around 20 rounds. In Civ 3, units have 3 to 5 HP. This means that combat is at most 9 rounds. This means that Civ 2 has at least 2 times as many combat rounds, probably a lot more. This also means that Civ 3 has greater deviation from expected results then Civ 2.

So, in conclusion, to make combat more credible, and lower the occurence of , you just need to increase the number of HP. I would mod that in, but I do not have the time.

Of course for Civ 3 purposes, bombardment weapons would need to multiple their Fire Rates accordingly, and you would have to turn off combat animations for mod purposes(each unit would take 5 mins to fight).

But your forgeting Armies, which can have HP of 6-20 HP unmodded. Even if combat animations were off, it would still long with a 160 HP army.

Personally, I have a fairly okay experence with the RNG with regular units, and found Armies a little underwheming.

 

[Gingerbread Man]

I like your thinking. Combine both the non-linear and higher number principles. It would really increase the importance of Barracks and elites.

Conscript - 5
Regular - 10
Veteran - 20
Elite - 40

Of course for Civ 3 purposes, bombardment weapons would need to multiple their Fire Rates accordingly, and you would have to turn off combat animations for mod purposes(each unit would take 5 mins to fight).

I was thinking more in the way of the fibbonachi number sequence e.g.
1, 1, 2, 3, 5, 8, 13, 21 etc
where the next number is the sum of the previous two numbers. Therefore a conscript and a regular equals a veteran, a regular and a veteran equals an elite, etc. If you want a less random result, simply put everything up one notch in the fibbonachi sequence.

That way you could actually have a sliding scale of less random to more random, without seriously messing up the combat system. For once everybody could be happy!