AU 206: Gallic Glory - Stories and Strategy Tips

[Nor Me]

From that screenshot, I attacked China razed those cities, got a leader and made peace. That was a 4 turn war. I researched Mathematics, rushed the Pyramids and attacked India. I made peace with India , attacked Persia, made peace with the Persians and asked the Chinese to leave my teritory triggering war. From that war, I got 2 leaders, giving me the FP and a gallic swordsman army. Rome declared war on me and I paid Gandhi to defend me. the Persians also decided to join in. I left the Chinese on 2 cities for the usual tech.
This 150AD minimap hardly looks out of place:

[Nor Me]

Originally posted by alexman
A Veteran Gallic Swordsman attacking a fortified regular spearman wins 66% of the time, and retreats 18% of the time. The expected loss of shields is 0.16*50=8.

A Veteran Swordsman attacking a fortified regular spearman wins 70% of the time. The expected loss of shields is 0.3*30=9

So on average, you lose more shields by attacking with regular swordsmen than you do with Gallic Swordsmen, even if the latter cost 50 shields compared to 30!

But this results in Gallic Swordsmen doing less damage. So you need more of them. You would probably lose at least as many shields worth of Gallic Swordsmen taking the same city because of this.
It certainly means that you need a high shield/gold investment to do anything with the Gallic Swordsmen.

In this game, I was able to use overwhelming force most of the time. But given a worse start, ordinary swordsmen would be much better. This is why I disagree with Arrian and nbarclay about the Celts having good traits for the swordsmen. While the Iroquois might be improved by Militaristic, the celts would definately be better as Industrious or Expansionistic. As it is, they might be in trouble in harder games than this one. Of course, if either had the option of giving up Religious...

[Chemical Ollie]

I am about to destroy all civs in my part of the world even before the Amaricans and French make contact with them! The Russians are gone too, killed by Americans.

This was a close call however. Look at the American galley that almost reached Roman land before I conquered it. I hope they don't find the last Roman city before I take it (on the island to the south west).

Sorry for the messy screenshot. I'm not that used to edit things in MS Paint.

[alexman]

Originally posted by Nor Me
But this results in Gallic Swordsmen doing less damage. So you need more of them. You would probably lose at least as many shields worth of Gallic Swordsmen taking the same city because of this.
It certainly means that you need a high shield/gold investment to do anything with the Gallic Swordsmen.

While it is certainly true that GS do less damage than swordsmen because they retreat before their job is done, I still think they are more cost-efficient than swordsmen.

Look at the ratio of the expected shields lost, to the expected HP damage caused to the defender. This gives you a number for shields lost per damage caused, and should be a better indicator of the cost-efficiency of units:

(All veteran, versus fortified spearman on plains.
Values are expected shields lost / HP damage caused)
Warrior: 6.17
Archer: 4.67
Swordsman: 3.62
Horseman: 4.05
Gallic Sword: 3.42
Immortal: 2.00
Mounted Warrior: 2.05
Jaguar Warrior: 3.35
Gallic Sword (40 cost): 2.73

Obviously, the lower the loss ratio, the more cost-effective the unit. Persia and the Iroquois have the most amazing UUs, but they are not militaristic. Militaristic civs with ancient-age offensive UU like the Aztecs and the Romans hava a loss ratio of around 3.5. Making the GS cheaper would put its loss ratio lower than those militaristic civs, and thus would probably be too good of a unit.

In any case, Nor Me, you are right that GS are best used in overwhelming forces, just as horsemen, or an even Jaguar Warriors. I just still think that the current cost is fine for the GS, given that it belongs to a militaristic civ. Even if it is not fine, it's along the lines of the rest of the ancient units in the game.

[Chemical Ollie]

My plan to kill all neighbours before they made contact with the other continent worked!

Edit: I tried to change the attached picture but it didn't work and eventually disappeared. I have to post it again!

[Chemical Ollie]

[badams52]

Good job Olaf!!! I didn't finish off the Romans, Indians, Chinese and Persians as easily as you did so I ended up selling all contacts to the other civs for profit

[Chemical Ollie]

Originally posted by badams52
Good job Olaf!!! I didn't finish off the Romans, Indians, Chinese and Persians as easily as you did so I ended up selling all contacts to the other civs for profit

I started with 2 lost games on emperor level, so I'm kinda cheating by playing the same map on monarch level now. Doing better than usually however. I think the extremely early single-archer attack on China started the chain of positive events leading to the current position.

I have 2/3 of the land mass, all 8 luxuries (3 by trade), just installed democracy and France just declared war on America, which means they will both stall in the tech race in the early part of the industrial era.

[Chemical Ollie]

See here, the Americans had no contact with the Romans right before I killed them. My slate will be clean (first time ever I accomplished that!)

[Nor Me]

Originally posted by alexman
While it is certainly true that GS do less damage than swordsmen because they retreat before their job is done, I still think they are more cost-efficient than swordsmen.

Look at the ratio of the expected shields lost, to the expected HP damage caused to the defender. This gives you a number for shields lost per damage caused, and should be a better indicator of the cost-efficiency of units:

(All veteran, versus fortified spearman on plains.
Values are expected shields lost / HP damage caused)
Warrior: 6.17
Archer: 4.67
Swordsman: 3.62
Horseman: 4.05
Gallic Sword: 3.42
Immortal: 2.00
Mounted Warrior: 2.05
Jaguar Warrior: 3.35
Gallic Sword (40 cost): 2.73

Obviously, the lower the loss ratio, the more cost-effective the unit. Persia and the Iroquois have the most amazing UUs, but they are not militaristic. Militaristic civs with ancient-age offensive UU like the Aztecs and the Romans hava a loss ratio of around 3.5. Making the GS cheaper would put its loss ratio lower than those militaristic civs, and thus would probably be too good of a unit.

In any case, Nor Me, you are right that GS are best used in overwhelming forces, just as horsemen, or an even Jaguar Warriors. I just still think that the current cost is fine for the GS, given that it belongs to a militaristic civ. Even if it is not fine, it's along the lines of the rest of the ancient units in the game.

I looked at your earlier results again. I love the way a 0.7 shield advantage is rounded to 1.
The trouble is that against a 2hp fortified spearman, the swordsman has an expected loss of 4.7 shields wile the Gallic Swordsman has one of 5.6 shields. That's a marginally bigger difference in the opposite direction. Gallic Swordmen retreating are likely to leave a few 2hp spearmen around.
All this is academic, the big difference in losses in practice is from counterattack by freshly built units on roads.
The Celts being militaristic gives them more leaders but this is normally only relevant if they have that overwhelming force. It doesn't really help them get more GSs out early.
I have never advocated a cost reduction for the GS. I voted for leaving it alone. But it is a bad UU for a bad start.
In this game I had higher losses with the GS than I would with swordsmen. But being able to get the gains from war much faster meant warmongering was a far better option than for most civs. I won't argue with 3 leaders and 3 crippled civs.

[alexman]

Nor Me, there is no rounding in the sword/GS comparison, but there might be some error because I wasn't calculating the odds analytically. I was using Zachriel's calculator, which is really a simulator.

But even if there is some small error, I think we agree with the fact that Gallic Swordsmen are about as cost-effective as regular swordsmen, which is just fine for a militaristic civ.

BTW, Militaristic doesn't give you more leaders, it gives you more promotions, which means less losses from counterattacks because of the increased HP and the increased retreat odds.

[nbarclay]

Originally posted by Nor Me

[Edit: This crossposted with Nor Me's and Alex's next posts.]

But this results in Gallic Swordsmen doing less damage. So you need more of them. You would probably lose at least as many shields worth of Gallic Swordsmen taking the same city because of this.
It certainly means that you need a high shield/gold investment to do anything with the Gallic Swordsmen.

In a limited campaign, Gallic Swordsmen require a larger investment. To take one city, you need at least as many GS's as you would regular swordsmen, maybe more because of the retreat ability (although the fact that retreat costs the enemy possible promotions at least partially offsets the reduction in damage caused). So if you just want to trim a single enemy back a little, you can definitely do the job a whole lot cheaper with conventional swordsmen.

But in a sustained campaign, either against one opponent or oscillating among multiple opponents, the Gallic Swordsman has several advantages.

1) Lower replacement costs due to higher survivability.

2) A higher percentage of elites and more leaders generated, since elites don't automatically die if they lose.

3) Shorter times to capture the same amount of territory.

4) Quicker transit times to go from one front to another when changing opponents.

5) Less war weariness under Republic to conquer a given amount of territory, due both to shorter wars and to lower casualty rates.

6) [Edit: I forgot this one] Less time for the enemy to build forces, upgrade forces, and gain technology for better forces during a war.

Also, the higher up-front cost of Gallic Swordsmen is made up for in higher back-end value. Left-over swordsmen, even if upgraded to medieval infantry, have a hard time playing a useful role in a knight campaign and an almost impossible time playing a useful role in a cavalry campaign. (At least that's certainly true with the kind of blitzes with overwhelming force that I tend to use.) In contrast, left-over Gallic Swordsmen are just as fast as knights and can play the same "clean-up" and "attack the counterattackers" roles that I often use left-over elite knights for in cavalry operations playing other civs.

I certainly don't regard the Gallic Swordsman at 50 shields as one of the greatest UUs in the game. But why should he be? What inherent right does he have to be better than average, or even to be as good as average, as UUs go? As it is, he's fairly middle-of-the-road, not nearly as good as the Chinese Rider (in my opinion) but still a big step up from the Man-of-War, the F-15, or the standard-rules version of the Musketeer. Which tells me Firaxis did a fairly good job of balancing him.

[/quote]In this game, I was able to use overwhelming force most of the time. But given a worse start, ordinary swordsmen would be much better. This is why I disagree with Arrian and nbarclay about the Celts having good traits for the swordsmen. While the Iroquois might be improved by Militaristic, the celts would definately be better as Industrious or Expansionistic. As it is, they might be in trouble in harder games than this one. Of course, if either had the option of giving up Religious... [/QUOTE]

With the kinds of starts where the Celts have trouble now, barring great luck with an early settler from a hut, they'd do a lot worse with Expansionistic instead of Militaristic. If the Celts have plenty of room to expand early, they'll do fine without the Expansionistic trait (at least as long as Iron is available; otherwise, the whole UU issue is irrelevant). If they don't have plenty of room, cheap barracks and early archers can often let them deal with the problem without waiting for Gallic Swordsmen and without having to blow their GA under Despotism if they don't want to. I'd gladly trade either of the civ's traits for Industrious, but given that Firaxis isn't going to make all the civs in the game Industrious, I think the Celts have a good synergy.

[nbarclay]

Originally posted by alexman

BTW, Militaristic doesn't give you more leaders, it gives you more promotions, which means less losses from counterattacks because of the increased HP and the increased retreat odds.

More promotions means more battles involving elites instead of veterans, which means (on average) more leaders. The effect is indirect, but that doesn't make it any less real.

[Dominae]

Originally posted by nbarclay
I certainly don't regard the Gallic Swordsman at 50 shields as one of the greatest UUs in the game. But why should he be? What inherent right does he have to be better than average, or even to be as good as average, as UUs go?

This was my point when this whole debate started. I'm just glad to see I'm not the only one on my side...! The Gallic Swordsmen promised to be before PTW came out, and on paper continues to promise to be, a dominant UU. That is no reason to turn it into one.


Dominae

[Arrian]

Originally posted by Arrian
*snip* ...I'm vaguely concerned about America "over there" builing the Sistine - so I want Monotheism quickly too.
*snip*
-Arrian

Ok, I fired my game up (400AD) and continued my methodical destruction of the Romans. I researched Feudalism first, and rush Sun Tzu immediately. I then researched Engineering & Invention and rushed Leo's immediately. I then traded Engineering to Persia for Monotheism and started on Theology.

One turn into that research, I was notified that "The Americans have completed the Sistine Chapel in Washington." Apparently, the Americans have a deathwish.

Anyway, that very turn I generated another leader, and I held him for Bach's. Just before rushing Bach, though, I traded Music Theory to the overseas civs for gunpowder & some cash (suicide galley run #3 was a success).

Rome has 3 cities (all offshore). Persia has 2. India has 1 or 2. Oh, yeah! China! Almost forgot about them. Over the course of several turns, I upgraded my vet horsemen to knights (approx 10-15 of them) and then unleashed them on Mao, along with my 3xGS army and other remnants. Happily, 2 turns prior to my attack, Mao's source of iron culture flipped to me. Not that it would have mattered much, since the iron supply was routed through the remnants of Persia, and I got the Persians to ally with me.

It is 900AD, and I'm still running a Monarchy, but as my southern empire gets built up a little more and my campaign vs. China winds down, I will switch to republic. I'm researching Astonomy (the Americans are the only civ that has it, otherwise I'm even in tech), after which the plan is to beeline for Smith's. I have a city (the one I built up on the river by the iron using the settler I got from the hut) that has been building a Palace for centuries and centuries. That city is now actually nearing completion (20 turns). Switching to a university (160 shields) would waste 495 shields, meaning he has built up 655 I figured I might not get all the leaders necessary to rush every medieval wonder, but have so far been wrong (though had I gotten the Sistine, perhaps that city would have built Bach). Considering the geography, however, I'm considering just allowing the Palace to complete. What do you think? Would that be better Palace/FP positioning?

-Arrian