Are you, in fact, a pregnant lady who lives in the apartment next door to Superdeath's parents? - Commodore

Create an account  

 
Intersite Game - Turn Discussion Thread

Our enemies continue to build units:

[Image: RB%20Poly%20Civplayers%20T168.png]

Note these is good news and bad news here:

(Sort of) Good news - a full range of drafts by Poly this late actually suggests they may wait 1+ turns before coming in. Unless we're missing a city location, they only need to draft 4 cities the two turns before they move in (and 3 the turn before) to get every drafted unit into our lands.

Bad news - the RB portion of the spreadsheet will shrink a bit this evening when I log into the game and check builds. This is due to a few cities where we're putting overflow into infrastructure as well as cities where we won't be able to 2pop whip units out next turn.

We're getting a clearer picture of the Poly stack (meaning the lower bound and upper bound for their strength is converging) and I'm learning more about the collateral system. That should allow us to refine our 'vs Poly stack' sim. Before I try to do that, can people look over the info below and let me know if anything is wrong?

*Unpromoted cats apply 6 damage to rifles, barrage 1 cats apply 7 damage. They apply ___ damage to jans, none to cats.

*Rifles are at max collateral at 70hp, jans by ___, and enemy cats are collateral immune.

*Max collateraled rifles are 44% battles for our units

*Cats apply 6 'ticks' of collateral to an enemy stack [source: Novice, who corrected the link I used earlier]

*A collateral tick is absorbed by both an already redlined unit as well as a collateral-immune unit

*The top defender is excluded from the collateral calculation [source for the last three bullets: IMs with pindicator, Novice, Ruff, and Seven]

*The top defender will receive on average 17 damage from the standard combat (assumes top defender is a healthy rifle, will need to revise when/if we have more cats than rifles) [source: cyneheard combat calculator]

EDIT: So here's a super quick and dirty estimate: if we have 30 cats attacking rifles should be the top defenders each time. We're going to distribute 510 damage from the basic attacks. We're also going to apply 1,080 total collateral damage to the stack, however 420 of that is going to hit either immune cats or not-as-important jans. That works out to 39 damage per rifle, which gets them to the collateral limit. What I'm unsure about is how many rifles will become collateral immune during the fight, but I don't think it will be many, and given the 15% cushion we should be OK.

Let's use my 32 rifles, 5 jans, 20 mixed older units, and 30 cats estimate instead. That applies the same 510 damage via basic attacks, but the 1,080 collateral now has 570 applied to the 'wrong' units, leaving the rifles at around 72hp before factoring in rifle 'misses.' That gives us 41% battles.

If we want to take our every enemy unit during our attack we need a lot of units. In the lower bound scenario we're looking at 30 attacks on the wounded rifle stacks, killing 13 and losing 17 units. The remaining rifles should average just 27 health and should drop below the cats in the combat order, which is annoying. We then need 5 jan battles, probably winning 3, and redlining the other 2. Then 14 cats, losing only once. The 20 cleanup battles. All told that's 69 attackers to kill everything.

In my middle estimate we need more. 32 attacks @ 41% means 13 kills, 19 losses. Same 3 kills versus jans. Then, and this is approximate, 17 kills versus the old units. 30 cats losing twice. Then cleanup (19 rifles, 2 jans, 3 old units, 2 cats for 27 total). That's 108 attacking units needed.
Reply

6 ticks per catapult. Checked the xml and confirmed in worldbuilder. So if you attack a stack of seven rifles with a catapult, they all get hurt. (Well unless the top defender wins flawlessly.)

Also confirmed that Rifles cannot be collateraled beyond 9.66/14 i.e. 69 HP. Not sure why really, since catapults are supposed to do up to 50% damage to units through collateral. I guess relative unit strengths factor in in an undocumented way.


I have to run.
Reply

Thanks Novice, I adjusted my post accordingly. I also edited my post like 10 times, so people may want to make sure they've seen the newest version of the quick sim at the bottom.
Reply

(August 9th, 2013, 15:43)sunrise089 Wrote: If we want to take our every enemy unit during our attack we need a lot of units. In the lower bound scenario we're looking at 30 attacks on the wounded rifle stacks, killing 13 and losing 17 units. The remaining rifles should average just 27 health and should drop below the cats in the combat order, which is annoying. We then need 5 jan battles, probably winning 3, and redlining the other 2. Then 14 cats, losing only once. The 20 cleanup battles. All told that's 69 attackers to kill everything.

Maybe we can attack with knights after we've catted the rifles, and flank away their cats.
I have to run.
Reply

barrage 2 cats will do 9 hp of damage to rifles.

The formula for collateral damage is either

floor(10 * (3A + D) / (3D + A)) or floor(10 * (A + ((A + D + 1) / 2)) / (D + ((A + D + 1) / 2))
Where A is the attacker's base strength, and D is the defender's base strength.
Those are the same almost all the time.

I don't remember the collateral cap formula off the top of my head. I think, but am not certain, that it's
5 times the collateral damage (for cats)

Jans will take a little more damage

floor(10 * (15 + 9) / (27 + 5)) = 7
That becomes 10 for barrage 2.

Same for Knights
floor(10 * (15 + 10) / (30 + 5)) = 7

---

If we attack with cats after doing other things, remember that everything below the collateral limit will absorb collateral hits just like a cat.
Reply

So you're saying that barrage II cats can collateral a rifle down to 55HP? We should test that in worldbuilder. If it's the case we should save the barrage cats for last.
I have to run.
Reply

(August 9th, 2013, 16:49)Zargon Wrote: barrage 2 cats will do 9 hp of damage to rifles.

Thanks, I saw that in your (awesome) post in the Attack on RB thread, I just figure we'll have very few if any barrage II cats.

(August 9th, 2013, 16:44)novice Wrote: Maybe we can attack with knights after we've catted the rifles, and flank away their cats.

This would be great. I have to get on the road shortly, but it would be great to get an estimate of how many would be needed. Noble is going to have to balance knights for the poly stack, defending in the south (because their mobility can cover a larger front), and possibly attacking the civplayers border city. I personally favor only attacking the city if we lose just a couple units doing so, and only favor sending knights south in general if we know we're going to need a number of turns to get sufficient poly-stack-attackers in the northwest anyways.

If we were able to flank all the cats away that lowers the number of units needed for the lower bound attack to ~55non-cat units, which should be doable.
Reply

I've spent some time looking for written resources for collateral damage and it's eithee conflicting or incomplete. Definitely should trust our own WB tests for this
Suffer Game Sicko
Dodo Tier Player
Reply

(August 9th, 2013, 16:53)novice Wrote: So you're saying that barrage II cats can collateral a rifle down to 55HP? We should test that in worldbuilder. If it's the case we should save the barrage cats for last.
No. I meant that the collateral limit is 5 times a regular collateral hit before flooring it (I think). Barrage definitely doesn't boost the limit, it just gets you there faster.

We'll have a few barrage 2 cats from HF, but we might have decided to get other stuff instead.

Yeah, I've seen conflicting stuff on the formulae for collateral, and I anticipate that some of it is for vanilla Civ 4, and some for BTS. I even mentioned the 2 competing formulae I've seen for collateral damage, but I don't care much because they return the same value for pretty much everything. Still, a WB test might be worthwhile.

If cats collateral rifles to 69 hp, then I like the second formula for collateral damage better (floor(10 * (A + ((A + D + 1) / 2)) / (D + ((A + D + 1) / 2))), because using 50 for the first term there gives 31 for the collateral limit, whereas using 50 in the other formula gives 30 for the limit.
Reply

(August 9th, 2013, 04:56)darrelljs Wrote: When you can make me laugh after getting up at 4 to catch a flight, you've won lol. The only bit that might not be obvious is that we did indeed send the note to the Frenchies.

Darrelll

You and me both - I got up at 4:30 for my flight ... so extra sleep for me. That said, my return flight is delayed 3 hrs already ... blasted rain.
I have finally decided to put down some cash and register a website. It is www.ruffhi.com. Now I remain free to move the hosting options without having to change the name of the site.

(October 22nd, 2014, 10:52)Caledorn Wrote: And ruff is officially banned from playing in my games as a reward for ruining my big surprise by posting silly and correct theories in the PB18 tech thread.
Reply



Forum Jump: