Camel archers and Numidian cavalry

[Tobiyogi]

In a recent AW game with Hannibal, I had some really positive fights with the Numidian Cavalry that comes with Flanking I for free (+20% withdrawal), so with barracks+stable (or GG), you can get a Combat I + Flanking II guy with 50% withdrawal chance which is really nice, especially when going into desperate fights (like the unlucky initial attackers who try to weaken garrisoned archers).

I was wondering if even more is possible and I just found an additional +30% withdrawal through a special promotion unfortunately being only available to GG-units. So 50% withdrawal, and that's it.

And now I have seen that Camel Archers are "born" with 15% withdrawal, not acting like a promotion. Does that mean you can upgrade them to 65% withdrawal chance?!? (that would change my ugly opinion about Arabia quite a bit), This is more interesting for AW games though, as a great weapon to flank damage enemy stacks a lot.

In general, as I am not really experienced with mounted attacks, I wonder why not make more use of the withdrawal chances, especially for the first attackers who most likely lose anyway?
(from what I have seen in videos, the usual strategy seems to promote your Horse Archers Combat 1+2 or shock)

 

[Tobiyogi]

Otoh, that does not seem to be so unique as I thought, even Cuirs get 15% and Cavalry even 30% withdrawal chance. Keshiks 20%.... So you could promote Cavalry up to 80% in that case? I never realized that before.

 

[AcaMetis]

You could use a GG to get a Camel Archer (or variety of other Mounted units) up to 65% (or higher) Withdraw Chance, yes. The problem, and reason why many people go for Combat promotions instead, is that a weaker unit has a lower chance of actually dealing damage before either retreating or dying horribly. Combat II mounted units are more fragile and likely to die, but the attack is more likely to succeed. Units tend to be easier to replace than the turns spend healing them before continuing the assault, unfortunately.

If you want to play with Withdraw Chance shenanigans in an AW situation, by the way, I'd recommend playing around with the Celtic UU/UB. Guerilla III gives +50% Withdraw Chance, and the Dun gives all eligible units Guerilla I for free, so it's any one of Vassalage/Theocracy/settled GG to get Guerilla III units right out the gate. It's not as effective as the usual City Raider/Garrison promotion lines, usually, but double movement on hills can be fun to play with.

 

[Tobiyogi]

65% + "tactics" (I think that's the name) = 95% withdrawal chance, almost immortal.... That's funny. Maybe only useful if travelling with a Medic III at the same time. Of course, when you withdraw, you were almost dead, that's the logic behind. And most HA attacks benefit from their "momentum" or however this is called...

The Celts, yeah. Avoided them for really bad starting techs. Guerialla III = +50 withdrawal, I did not know that, thanks. Must mean something like disappearing in the mountains

Does that mean that any unit "born" in a Dun-City has access to the Guerilla-Line? (mounted, melee....). Normally, it's only archers, no? Wait, I am confused, you said "elligible" units. But the Gallic Warrior is also a Swordsman, normally not "elligible"....?

 

[AcaMetis]

65% + "tactics" (I think that's the name) = 95% withdrawal chance, almost immortal.... That's funny. Maybe only useful if travelling with a Medic III at the same time.

Not quite. Camel Archers get 15% Withdraw Chance base, Flanking I adds 10%, Flanking II adds 20%, and Tactics adds 30% = 75% Withdraw chance. Not sure where I got 65% from earlier, must have been looking at another unit. The highest you're going to get is a Cavalry (base 30%) with 90% Withdraw Chance. I'm actually not sure if that's higher than the game's hard cap on Withdraw Chance, I know it has one but I'm not sure how high it is. And yes, Withdraw Chance builds really need a strong Medic unit to go with them, otherwise they're out of the fight for far too long.

The Celts, yeah. Avoided them for really bad starting techs. Guerialla III = +50 withdrawal, I did not know that, thanks. Must mean something like disappearing in the mountains

Does that mean that any unit "born" in a Dun-City has access to the Guerilla-Line? (mounted, melee....). Normally, it's only archers, no?

The Dun only gives the free promotion to units that can get the free promotion, which is Recon, Archery and Gunpowder units. Although the only Gunpowder unit you can get without Rifling, which obsoletes the Dun, is a Musket. A few unique Melee units also have access to Guerilla as well, but those obviously aren't relevant in this case. Even the Celtic UU, ironically enough, since it gets Guerilla I for free just on it's own.

 

[dankok8]

Otoh, that does not seem to be so unique as I thought, even Cuirs get 15% and Cavalry even 30% withdrawal chance. Keshiks 20%.... So you could promote Cavalry up to 80% in that case? I never realized that before.

Yea Knights are the only mounted units IIRC that don't get any withdrawal chance.

Camel Archers are nothing special as a UU because it's hard to beeline Knights in general and they don't ignore Walls and Castles like Cuirassiers do. However, Camel Archers do better busting cities defended by Longbows than ordinary Knights and even better than Cataphracts who while they have 12 strength have no FS immunity. Also a point that Camel Archers are resourceless so you can build them in every game.

I think if you build a swarm of mounted, the first few attackers should have flanking because they are attacking the healthiest defenders and have the highest chances of dying and then the others should have combat. That's the conventional wisdom anyways.

 

[Tobiyogi]

It was me who was misled about the Numidian cavalry in the first place. They start with an (inherent) 20% wd chance plus have Flanking I and when you add Flanking II they are at 50%. Not realizing the inherent wd chance, I concluded only that Flanking I means 20% and Flanking II 30% (because most of my units had 50% anyway) and transmitted that to the Camel Archers....

 

[Tobiyogi]

So Camels have +45% in the best case, not too great, but I would not consider to rush with Knights anyway. On a good IMM map, you could have some fun with Cataphracts alone, if you send Spies into the cities, probably worthwile, but maybe only with gold/gems, maybe Oracle in the first place. I don't know, don't have too much exp with those civs anyway. On a sidenote, Saladin and Justinian seem to be almost identical, same starting techs, a unique knight, both are SPI, ....

 

[drewisfat]

I want to say the withdrawal cap is 70%, but that's a guess. I remember being totally dismayed in a game where I farmed withdrawal chance.

Sadly, Flanking just isn't a good promotion line for several reasons. For starters Flanking 1 with only a +10% withdrawal is an absolute garbage promotion, while strength 1 is fine. Flanking 2 looks more attractive with +20% and granting immunity to first strikes. This would indeed make it a decent promotion.... except most mounted units already are immune to first strike. If we look at just which is better, a unit with strength1-2 or flanking 1-2, the strength is going to outperform - and of course there are substitutes to strength 2 that could outperform even more in certain situations. And it does nothing on defense, and only yields 1 exp if you do withdraw. If we look at specific situations where flanking should make sense, it still kinda falls flat on itself.

1. They're going into a battle with low odds of success. Since they're going to die regardless, isn't it a good idea to give them a chance to survive?
The problem here is you're only considering that unit. If they're going into a battle with low odds of success, there's a chance that the defending unit survives with enough HP to survive the second unit's attack. Those odds increase if you're sending in a weaker unit, erasing any overall unit gains you would expect. In the immediate short-term, it just increases the number of units you'd need to take a city that turn.

2. What about that + 30% withdrawal promotion for a GG? Seems really great, because I want to keep my GGs alive, and that's like 2 flanking promotions combined!
The problem here is you don't want to risk your GGs on risky fights. Withdrawal chance gets the maximum value on risky fights. I'm not risking GGs unless I get 90%+ odds, ideally 99%. While 30% is a larger number, it's always going to be in the context of a very small number when it comes to GGs. No one is throwing away their GGs at low odd fights and trusting that 1/3 of the time they didn't lose their GG.

3. Numidian cavalry are a good support/defensive unit, but typically when we think HAs we think "only HAs" to take advantage of the speed bonus. Comparing a bunch of HAs attacking a city vs a bunch of Numidians attacking a city doesn't compare favorably, unless the defender is using mostly melee instead of archers for some reason (silly Monty). I've already said I think grabbing 2 strength promos > grabbing 2 flanking promos. If we disregard the melee bonus, Numidians starting at 5 instead of 6 strength is like starting 2 strength promotions in the hole, while only getting one flanking! This is not good. They're mostly a support unit. I wouldn't consider using them like an HA rush.

4. Camel archers are an interesting UU just because they require 0 resources while Knights need Iron and Horses. This at least lines up with Saladin being PRO, so his resourceless archers can cover you until resourceless camels. But the obvious issue is Knights are just in a terrible position tech-wise and don't bypass walls. You can mitigate this somewhat by running an espionage economy, so you can steal some of the awkward techs and flip cities. Saladin is PRO, SPI, starts with myst and wheel, and UB is madrassa. That's actually a good setup for an EE. I doubt it'll do as well as playing normally, but hey it's unique, and can give you a solution on an unfair map devoid of strategic resources.

 

[dankok8]

So Camels have +45% in the best case, not too great, but I would not consider to rush with Knights anyway. On a good IMM map, you could have some fun with Cataphracts alone, if you send Spies into the cities, probably worthwile, but maybe only with gold/gems, maybe Oracle in the first place. I don't know, don't have too much exp with those civs anyway. On a sidenote, Saladin and Justinian seem to be almost identical, same starting techs, a unique knight, both are SPI, ....

I did a test with Cataphract a few months ago. They were really hyped up when BTS came out but they are a very subpar UU. They get barely better (like 1% better) odds against Longbows in cities than stock Knights but because of no FS immunity they have a lot more fights where they fail to even damage the defender. Basically their expected HP left is lower than stock Knights. Sometimes odds don't even tell the whole story. I used the Advanced Combat Mod when I did these tests. Now obviously against Pikes they'll do better than normal Knights but if they have Pikes odds are they have Castles as well and that war will be a painful slog anyways plus there is still a lot of Longbows you gotta kill.

So Camel Archers are actually better than Cataphracts.

I'll run some tests with Numidians but I suspect they do a bit worse than stock HA's overall. You mentioned a CI/FI/FII Numidian which has 5.5 strength and 50% withdrawal chance. A CI/CII stock Horse Archer has 7.2 strength and 20% withdrawal chance. The latter will do a lot better in all non-melee fights and even against melee like Spears where NC's have +50% I suspect they won't do much better than stock HA's. The more defensive bonuses for the Spears, the weaker the +50% bonus is because it's subtracted from the defender not added to the attacker.

 

[jmas]

@dankok8 , Do you have a link to the Cataphract test? I'd like to see, to know whether it's worth trying a Cata war in the future or not.

 

[dankok8]

@dankok8 , Do you have a link to the Cataphract test? I'd like to see, to know whether it's worth trying a Cata war in the future or not.

No I don't. I did a clean re-install of Civ IV. Got the Steam version for like 8 bucks maybe a month ago. But I remember clearly that the Cataphract in common had either marginal or somewhat clear edge in terms of odds (anywhere from 1% to 15%) but the expected HP left favored the stock Knight and the Cataphract also had much higher odds of leaving the Longbowman unscathed. I could be mistaken though and want to redo this.

Just did a test tonight for Numidian vs. HA though... Will try and post the results the tomorrow but here is the save if anyone wants to have a look. You need Advanced Combat Mod though which breaks down probabilities of battle outcomes in more detail including expected HP left.

 

[ArchGhost]

I think if you build a swarm of mounted, the first few attackers should have flanking because they are attacking the healthiest defenders and have the highest chances of dying and then the others should have combat. That's the conventional wisdom anyways.

Yeah, the flankers should go first. Doesn't make much sense otherwise since their combat odds are lower to begin with and one banks on their retreats to do anything with them.

Tossing in some flanking units is really only helpful when you would otherwise lose a lot of units to even score an injury -- the kind of of odds where you could throw 3 cuirs at an entrenched pike before it drops out of the defender rotation, or Cavs vs. PRO rifles with no siege or revolts, that kind of stuff. I think of it quite similar to the idea of a "can-opener" GG unit but applied across more units and easier to obtain. If you're able to punch through after only a few losses of C2 units you'd be better off just eating it and slowly building your numbers as you push (especially as survivors start to gain C3), as Flanking units are inferior in every other situation than severe disadvantage vs. entrenched defenders :/

 

[sylvanllewelyn]

"Starts with Flanking I" means, with barracks and 1 more xp somewhere, Numidian Cavalry get "mobility".

I like UUs that grant extra movement. Gallic warriors have 2 moves across hills. Jaguar warriors have 2 moves in forests. The Impi have 2 moves. Musketeers have 2 moves and draftable.

 

[lymond]

Love me some Numidians. I'd rather go with the Flanking and stronger promos. Adding mobility isn't really helping much since you already have two moves.

 

[Tobiyogi]

First of all, some of my considerations were adapted to AW games which I play a lot. In that case, numidians are great units (camel archers probably too), but yes, more like a supportive unit (stack-weakening, pillaging, scouting, ).
In a normal game, I don't know, rushing against archers is not really nice..... Only advantage here is that a NC + C1 + Shock outweighs even spears.

Secondly, there seem to be two different approaches to HA rushes. Either promoting the Flanking line (which I would also tend to do) for the first fights, or to promote down the combat line. I can see the Point 1 from @drewisfat that combat makes it much more likely to do substantial damage to the defenders. And if a flanking guy doesn't do damage at all, well, you will probably lose the 2nd attacker as well.

 

[dankok8]

I completely forgot about this. I'm finally posting my NC vs. regular HA comparison in terms of their effectiveness in city attack.

I chose an enemy city with 50% tile defense which is pretty typical for a classical age city (Walls + roughly 200 culture). I did it for three different defenders, a regular Archer, a Pro Archer, and a Spearman. For the attackers, I compared a F1/F2/CI Numidian, a C1/C2 Horse Archer, and a F1/F2 Horse Archer.

vs. a regular (CG1) Archer

F1/F2/C1 Numidian: 27.23% W/36.38% R/36.38 %L
C1/C2 Horse Archer: 67.59% W/6.48% R/25.93 %L
F1/F2 Horse Archer: 32.18% W/33.91 %R/33.91 %L

HA's come out on top here as expected. And combat promotions make the most sense for HA's when fighting regular Archers giving the highest win rate and highest survival rate. No contest.

vs a Protective (CG1/CG2/D1) Archer

F1/F2/C1 Numidian: 21.88% W/39.06% R/39.06 %L
C1/C2 Horse Archer: 36.47% W/12.71% R/50.82 %L
F1/F2 Horse Archer: 26.16% W/36.92 %R/36.92 %L

This one is closer. Combat HA's have a much higher win rate than Numidians but also lower survival. However HA's with flanking promotions win this one with higher survival and win rates than the Numidians. It seems that when fighting Pro Archers, using flanking promos for HA's makes a lot of sense.

vs. a (C1) Spearman​

F1/F2/C1 Numidian: 9.45% W/45.27% R/45.27 %L
C1/C2 Horse Archer: 11.21% W/17.76% R/71.03 %L
F1/F2 Horse Archer: 3.35% W/48.32 %R/48.32 %L​

This is where Numidians shine but it isn't clear cut. Combat HA's do worse against Spears but flanking HA's do very well with only 3% worse survival odds. But Numidians are the best unit when facing Spears. Just not by a lot. It surprised me how good HA's are with flanking promos. Might go for those more often instead of combat especially when expecting tough fights. Of course retreat is only valuable if you have enough units to finish the job.

Another thing I found but it's too tedious to post results of every single battle is that when tile defenses get lower (less culture) Numidians tend to do better but when tile defenses get higher (hills, more culture) then HA's do better. Against Spears that is so Numidians do a significantly better job at killing Spears if you can lure them on open ground but HA's do pretty equally good when attacking heavily fortified positions. The reason is that when the defender gets so many defensive bonuses the +50% attack becomes less and less valuable because it subtracts from the defensive bonus. For instance a C1 Spear on flatland has a modified strength of 8.4 (C1 + 100% vs mounted) when fighting HA's but only 6.4 (C1 + 100% vs. mounted - 50%) when fighting Numidians. But with very high say 100% tile defense the Spear's modified strength drops from 12.4 to 10.4 which doesn't make as big of an impact on the odds which use the ratio of attacker to defender strength. In that case the HA with higher modified attacking strength than the Numidian can actually get better odds.

 

[Fippy]

Imo Numidians are better on deity where AIs connect metals so fast, and worse on other diffs.
Sometimes a deity AI will have no copper or iron, but it's rare.

But more important than HA or Numidian for any rush will always be: how fast can i get them and attack.

 

[jmas]

Thank you, I found this very helpful @dankok8

 

[Gumbolt]

Both will be great for attacking. With high withdraw odds it sometimes does not matter a city has hill defences as you know the first few attackers might withdraw.How often do Ai not have metal defenders? Once they get metal units archers are produced a lot less. You will always pick your target in any way. Chances of protective Ai are unlikely.

As Mylene suggest the sooner you declare the easier the war should be. With cha NC can get 3-4 promotions easily. Only 1-2 attacks if you have stables and barracks. Barb units really help too.