Apolyton University Mod (PTW version)

[Dominae]

Nathan's right to point out the AI's use of Regulars in the early-game: a Vet. Spearmen defending an Archer stack does too well against the AI's favorite defense, the Archer counter-attack. The reason is that the AI does not take into account the possible effect of bombardment on its chances when deciding whether or not to attack. So it may attack thinking it has an acceptable success/loss ratio, when in fact would not do so if it realised its unit will be one HP short when fighting.

The proposed solution (remove zero-bombard from Archers only) is a good one, since the effect described above is washed out by other factors when Longbowmen and Guerilla are available. Perhaps we could leave the Bowmen with zero-range bombard, as an interesting way to balance that unit (most people think its kind of weak).

Nathan, I never disband my "leftover" Archers. But the fact is that I never really produce all that many of them (typically less than 10), so I only have maybe 4 left when my Archer rush is over (due to losses). I always prefer Horsemen when these become available. Actually, playing AU206 I was wondering if producing more Archers might be a good strategy, but I forgot about that pretty quickly because Gallics would soon become available. Something to think about in the next AU game, assuming Archers still have the bombard ability.


Dominae

[alexman]

Guys, I think you overestimate the effect of zero-range bombard in an early archer rush.

A regular archer in stock Civ3 has a 30% chance of defeating a veteran spearman in the archer stack on plains.

In the AU mod it has a 20% chance to do the same. Not a big deal, and probably worth keeping the zero-range bombard for the added usefulness of Archers once swordsmen are available, IMHO.

[Dominae]

If those are the numbers (which I'm sure they are), then I withdraw my conclusion. It felt a lot different in my AU game. However, there is still the fact that the human player can use zero-range bombard a lot better than the AI, but what else is new?


Dominae

[nbarclay]

Not a big deal? The AI's chance of taking out a spearman is reduced by a third! For every three defending spearmen a player would lose in stock Civ 3, he'd lose only two in the AU mod. Maybe not earth-shattering, given the relatively limited number of archers an AI can typically strike with, but not completely trivial either.

[nbarclay]

By the way, in the open field, every hit point an AI takes off a spearman is a turn the spearman has to wait and heal before being able to rejoin the attack if he wants to be healthy for his next fight. So reducing the AI's ability to take hit points off spearmen reduces it's ability to slow down archer rushes even when it loses battles.

[Dominae]

I doubt such a difference imbalances the Archer in the AU mod. As you pointed out, you'll not likely be attacked by that many Archers (you would need to be attacked with 10 Archers to see a whole-number difference on average), so its not easy to "abuse" this advantage. Usually you'll just want to strike with your own Archers and be done with it. Without doing the math, I was under the impression that the Spearmen got a 20-30% increase in survival, which obviously would be quite significant.


Dominae

[Dominae]

I suggest a couple of us try a full-blown AU mod Archer rush in the next AU game (I know I'll try it). If Archer rushes suddenly become more than they're supposed to be, a change would be in order.


Dominae

[nbarclay]

An average of one in ten means one in five about half the time, or one in three a little less than a third of the time. It's not just a question of averages; it's also a question of, "How many spearmen do I need for an adequate safety margin?" And to whatever extent AI counterattacks can delay a rush even when losing battles, the AI has more time to build additional counterattacking and defending units, thereby making the war longer and potentially more expensive.

I don't think it's a game-breaker, but given the principle that the game should be changed only when the advantages clearly warrant it, I don't see how zero-range bombardment for archers can be justified. It tilts the balance more in favor of early archer rushes (already a fairly popular tactic in the stock game), not just because it improves archer stacks' ability to defend themselves but also because left-over archers can be used to augment the defenses of sword stacks in the next phase of a warmongering strategy. (That aspect, of course, did not come into play significantly in AU 206 because of Gallic Swordsmen's higher movement rate.)

Nathan

[nbarclay]

Originally posted by Dominae
I suggest a couple of us try a full-blown AU mod Archer rush in the next AU game (I know I'll try it). If Archer rushes suddenly become more than they're supposed to be, a change would be in order.

Not a sufficient sample size to determine statistical validity.

In any case, I thought the idea of the AU mod was that we don't make changes to the stock game unless the changes provide significant advantages. I can see serious advantages to giving zero-range bombardment to longbowmen and (as unrealistic as I find it) guerillas. But given the likelihood that my swordsman stacks will have a couple left-over archers in them anyhow (whether from an archer rush or from barb hunting), it's hard to see how giving the ability to archers either gives AIs a better chance against humans or imposes significantly more difficult strategic choices on human players. And in the absence of one or both of those, providing archers with that ability contradicts the AU mod's principles.

[alexman]

Nathan, I think you're arguing just for the sake of arguing.

The AI does not attack with enough archers for zero-bombard to make a difference in how many spearmen you bring along with your archers. You need to bring enough to garrison the city after you move on anyway. The difference will likely be the number of HP lost by a spearman, and in that case you would just leave him behind as garisson.

[nbarclay]

Originally posted by alexman
Nathan, I think you're arguing just for the sake of arguing.

The AI does not attack with enough archers for zero-bombard to make a difference in how many spearmen you bring along with your archers. You need to bring enough to garrison the city after you move on anyway. The difference will likely be the number of HP lost by a spearman, and in that case you would just leave him behind as garisson.

You're assuming a style of play that garrisons cities. I usually like to keep as many of my units as possible on the front lines while fighting a war, so it's not rare for my units to move on as soon as they heal, leaving a captured city ungarrisoned. Further, if a spearman only has one hit point of damage and he's on the road network, he heals faster waiting where he is than he does moving into a city and then waiting another turn to heal.

Also note that the difference is not just in terms of what forces a player would bring along but also in the risk involved in not bringing enough if the RNG turns against the player. And in any case, what advantage does zero-range bombardment for archers offer that makes it worth changing the game from stock rules for, given that human players may get more than their share of benefit from it?

[alexman]

Originally posted by nbarclay

what advantage does zero-range bombardment for archers offer that makes it worth changing the game from stock rules for, given that human players may get more than their share of benefit from it?

I stated them above, but I will do so again:

1) The AI benefits when it can't build better offensive units. The human usually doesn't have this problem.
2) The AI benefits because it actually builds archers even though it can build better units.
3) Players have a reason to use archers, after swordsmen/horsemen are available.

[nbarclay]

(1) and (2) are relevant only if the AI gets more advantage than its human counterpart does, or if it contributes to creating "killer AIs". That's arguable at best. Keep in mind that the same arguments for why the benefit to human players is negligible tend to also apply in the other direction. Further, situations where one AI has resources and another doesn't can help contribute to the existence of a "killer AI." Giving the AI without resources an advantage would tend to undercut that.

In regard to (3), how much difference does it really make in players' build priorities? Archers can't keep up with horsemen well enough to be of much value as a defensive escort, and AIs rarely counterattack with enough units to make more than two or three archers in an offensive stack useful. So the player is probably more likely to get a free added bonus for archers he would have built in the early game anyhow than to seriously alter his build priorities. Edit: Unless, of course, the bonus for archers encourages more aggressive early warmongering (since the left-over archers have more value). But I don't view giving early warmongering a boost over building as a positive development. The dark side already exerts too much influence.

It can be argued that giving archers zero-range bombardment has advantages, but I view the advantages as being far too small and questionable to justify the size of the deviation from the default rules. (As I said, I'm speaking specifically of archers here, not of longbowmen or guerillas.)

[Dominae]

Let me see if I'm getting this straight:

1. We're not giving Archers zero-range bombard to help the AI on defense, because that would cause long wars of attrition, which is bad.

2. Instead, we're adding an ability to Archers because the AI still builds them when other, better units are around. The logic is that the AI will be more potent of its weak units are made a little stronger.

3. The zero-range bombard ability is not useful enough for humans to exploit. In fact, it's almost insignificant.

4. But, zero-range bombard does gives the player an incentive to use Archers, even when Swordsmen and Horsemen are available.

Now, no offense, but that's quite a tangle of arguments. Some seem to be in direct opposition to one another. What is clear however is that the change 1) does not significantly help the AI (it just ends up attacking with the Archers anyway), 2) does not give the human player significantly more options (Archers are just better when you're Archer rushing; you would never use them alongside Horsemen and Swordsmen unless you've got a few leftover from an Archer rush), and 3) changes the rules from stock Civ3.

I admit I like the Archer modification, but I also believe that, given my understanding of its justification, it does not fit into the AU mod philosophy.


Dominae

[alexman]

OK, there is a difference between exploiting something, and just using it. The former breaks the game but the latter enhances it. Archers can be used as part of a combined arms force. They put the third dimension of Civ3 warfare, the bombard ability element, back into the ancient age, since catapults are not very popular. It might entice the player to bring his Archers along for the offensive, instead of just leaving them behind for MP. Can the human exploit this new ability? Not really. Does it change his strategy? Not really. Does it make the game more enjoyable because we have this new option of including Archers in a mid-to-late ancient-age attack? I think so.

Also, Civ3 is supposed to be based on realism, and it is difficult to justify zero-bombard for Longbowmen when you don't give it to Archers.

I still think this is a big fuss over a small matter.