Civ III: Conquests Patch Notice

[eliliang]

Originally posted by swat-spas2
Right, it would appear the principal problem is that in order to kill 'streaks', the probability of an nearly equal combat is skewed heavily toward the 'better' unit.

Why is it so necessary to kill streaks? If the job at hand was to flip a coin multiple times, would one want to fix it so that the coin couldn't 'streak'? In this case, the coin would no longer be 'fair' either.

I think that the only reasonable interpretation of a streak is that we don't want 'misbegotten' ones. And those only come from poor PRNGs. Let streaks happen. The universe has ordained that they should exist.

[nbarclay]

What's really "screwed" is the AI's ability to mount an offensive in the time between infantry and tanks. If Alexman's calculator is on target, the most likely outcome of twelve veteran infantry attacking a single conscript infantry fortified in a city (or walled town) is probably twelve dead attackers and an elite defender withabout two hit points left. (And that's if the city is not on a hill.) Cavalry would have a little bit better chance because if they retreat, they don't provide an automatic promotion every other battle until the conscript is elite. But even so, if Alexman's calculator is anywhere near on target, attacks against cities with fortified infantry will be difficult with artillery bombardment and essentially hopless without.

On the other hand, what I'm hoping is that the description Alexman's modified combat calculator is based on was significantly off target.

[Garth Vader]

Something that would have helped combat a lot for me would have been to have a "display odds" option like in SMAC.

I never complained about the combat from SMAC, but yet I frequently complain about Civ. Part of that is the lower hitpoints too.

[ayronis]

Let us not forget that this is only a beta-patch. If it turns out to be a grand failure, I'm sure they would change it back. I don't perceive this 'test' as a bad idea, since we can still undo the changes if we don't like it.

[nbarclay]

Originally posted by ayronis
Let us not forget that this is only a beta-patch. If it turns out to be a grand failure, I'm sure they would change it back. I don't perceive this 'test' as a bad idea, since we can still undo the changes if we don't like it.

Agreed. And if there is a problem, I'd much rather we catch it sooner rather than later.

By the way, in case I forgot to say it: Thanks, Firaxis, for releasing the beta version for us to use while we're waiting for the final one!

[Alex]

Originally posted by ayronis
Let us not forget that this is only a beta-patch. If it turns out to be a grand failure, I'm sure they would change it back. I don't perceive this 'test' as a bad idea, since we can still undo the changes if we don't like it.

wise words. and let us not forget that all we can do now is speculate. we have yet to play with the beta patch in order to draw conclusions about combat results.

[eliliang]

Originally posted by Alex


wise words. and let us not forget that all we can do now is speculate. we have yet to play with the beta patch in order to draw conclusions about combat results.

This is hardly speculation if we have understood Mike correctly. These are more logical conclusions based on our understanding, since the analysis is so simply done. What we can speculate on is how this might affect grand strategy in the game, as those types of effects are not really computable.

[TheArsenal]

The Great Wall's strategic value has just gone up.

[Theseus]

Walls' strategic value has gone up.

[SirOsis]

The Greeks defensive capabilities in the Ancient Age are certainly impressive now. Catapaults will be required even more.

[Jeem]

Originally posted by JesseSmith
Combat System "Less Streaky"

The Change was having the random number generator (rng) calculate the Results 4 times instead of 1.

Possible Outcomes: AttackerLoses, AttackerWins

If there is a tie, it retries.


This forces the combat system to be more accurate and less "streaky"

If I'm reading this right (EDIT - I wasn't, ignore this and see below), it seems like overfixing Jesse. Although I frequently get frustrated by unlikely combat results, this will make them incredibly rare.

Furthermore, those 'close' calls (say 6 attack vs 9 defence) are now much more likely to fall in favour of the highest strength. MUCH more likely - the 6 attack would only have a 33% chance before, but with this change it'll now be down to around 4%! (33% chance needing to win 3 combats?)

Checking twice will be more than sufficient, and will put spearmen beating tanks into the realm of 'incredible luck'.

On the SEED:

The SEED allows you to type 'CIVILIZATION' and it generates a map. Anytime you use the SEED: CIVILIZATION you will get the same map (only type in the SEED, ignore the other settings).

This is a welcome addition. Are civ starting positions still randomised with this? For example, if the exact same races were chosen, would they start in the exact same spot?

[Jeem]

Re: Combat

Originally posted by Mike Breitkreutz FIRAXIS
Think of each round of combat as the attacker rolling dice to see if they hit the defender. If the number rolled is greater than or equal to the defense value, the attacker hits; otherwise, the defender hits. The change we made was to the way the attacker rolls the dice. The attacker now rolls multiple times and the result is the average of all the rolls. This makes the combat results be more in line with what you'd expect them to be. It will reduce the luck factor because it eliminates most wild, "unnatural" runs. However, the luck is still present and it even makes lucky wins more meaningful since they are more rare. A single spearman can still beat a single tank but it's highly unlikely that a single spearman will be able to stand up to a stack of tanks (which is as it should be).

OK, bad me for not reading this before responding to Jesse's post.

Take the 6 attack vs 9 defence. That's a 33% chance of 'hitting' each round under the old system right? With 4 attempts each round, is this not simply adding to the chance of a single unlikely roll making the result a bit skewed?

For example, the rolls could be 10 (fail under old system), 10 (fail), 100 (hit) and 20 (fail). This assumes 33 needed to hit. The average roll of those four rolls is 35 so a hit is scored overall. Seems like the chances of hitting haven't really been changed much, but are still dependend on a lucky roll - the main difference being one extremely high roll (which is much more likely given 4 attempts) will usually mean a hit regardless.

[Nor Me]

Edit: Forget it.

[Jeem]

Originally posted by Nor Me
The trouble is that the number of rolls, 4 is even. The effect is then going to be more than squaring the attack and defence.

If the number was odd and high enough, we wouldn't even notice. We have to hope that 4 is wrong.

Problem is, the more attempts you have, the more likely the 'true' result is had. You might have a 1-10,000,000 chance of winning the lottery, but if you put on 10,000,000 tickets it's a sure thing. In order to keep 'randomness' to be a factor, you need to strike a balance. I reckon 4 attempts is way too high, and twice would be enough - at least to prevent those really unlikely results.

We've all seen it ourselves - the time when the tank loses to that spearman. How many of us see it happening twice in a row? By simply calculating the value twice then we get to the 'highly unlikely' stage. Four times makes it almost an impossibility.

[WarpStorm]

Personally, doubling up rather than quadrupling would have given a better feel. It would give results closer to the mean, but with a fair amount of variance.