Rebalancing Civ4: RtR Mod

[Cyneheard]

Not sure what I did, but reinstalling fixed it. And I was careful not to lose any files that I actually needed.

[Cyneheard]

Looks like I replied to the wrong T-Hawk post above (See PBEM4 IT for where I should have gone).

Since I'm trying to make sure that this is a thorough test (including when rounding comes into play, especially for Quick Speed), I do want to double-check that the only penalties for whips come from either:
Whipping from scratch
Whipping a wonder

I can't think of any other time where you get less than the full amount of hammers.

[T-hawk]

That's all I can think of. The wonder penalty is not always exactly +100% hurry cost (equivalently -50% whip yield). There are a few national wonders with +50% hurry cost (-33.3% whip yield), and the United Nations and Apostolic Palace are +300% hurry cost (-75% whip yield.) This is iHurryCostModifier in CIV4BuildingInfos.xml.

Also you could test the Kremlin, though that sees some rounding oddities even in a normal game where 1 whip yields 44 hammers but 2 pop is 90.

[Cyneheard]

Ah, right, the Kremlin.

The other National Wonder costs don't matter:
+50% is the same as the New_Hurry cost, i.e., same as whipping from scratch. I assume we don't care if there are slight errors if you want to whip the Heroic Epic from scratch. Likewise, with the AP and UN, not an issue.

Good thing tomorrow's a holiday.

[T-hawk]

The +50% for a national wonder is the same effect as the scratch penalty, but is calculated at a different step, so should be tested separately. (Yeah, I wouldn't worry about combining the scratch and wonder penalties, but we should make sure they work separately and both in conjunction with Kremlin. Whipping the AP and UN is sometimes part of a quick win strategy, though not from scratch.)

Also, you probably know this, but remember that Quick speed applies a modifier of 67% to most things, meaning exactly 67% and not 2/3 as in 66.6666...%.

[Cyneheard]

Whipping the AP and UN is sometimes part of a quick win strategy, though not from scratch.

This isn't a HOF contest, where I could see such a thing happening (and wouldn't the cleverer UN player either a) have a GE or 2 saved up, or b) use whip cascades?).

That extra 1/3% for Quick does add up.

Worldbuilder Pyramids + Shwed + Globe (can't cripple my Whiptesting city) + Spiritual is going to be my friend with these tests. Lots of situational modifiers to play around with without constantly having to go into WB and/or reload.

Tested the new Financial both in and out of Golden Ages, and it works right. Land-based tiles need to be worth 3 commerce to get the bonus, and water tiles only 2 commerce. A tile that is only worth 2 base commerce on land (such as a FP Cottage) improves to 3 commerce when under a Golden Age, not 4. This Golden Age behavior is as expected: a Golden Age cannot be the cause of getting Financial's bonus commerce in a regular game. Shame I wasn't playing that game for real...as Mansa, I had a natural 2 gold (both on PH), several flood plains, (dry) corn and Marble start. Oh, and copper in the 3rd ring.

[T-hawk]

This Golden Age behavior is as expected: a Golden Age cannot be the cause of getting Financial's bonus commerce in a regular game.

I can actually fix that. In CvPlot.cpp, the addition for Golden Age comes right after the threshold check for Financial's bonus. Simple matter to move the Golden Age check before the Fin check. But I think the original behavior is by design, or at least by inertia in that's just how it was coded and never changed.

[Cyneheard]

I don't want to be messing with Golden Age behavior. There's no reason to, so let's leave it be.

[Cyneheard]

See OP. Version 0.4, with the DLL fixes, is up.

Financial works as promised. Slavery looks to be functioning as expected. Normal speed: 30/50/70/90. Quick Speed: 20/33/46/59. (No modifiers. And, yes, 59 is less than 67% of 90. Blame Firaxis for using integers in situations where division happens)

Please try a game or two, and see how the different changes feel. We've made enough changes at this point where we need both theorycrafting and some actual experience.

I've not finished testing for rounding purposes, but on Normal Speed, your math should be correct unless you're whipping something from scratch or a wonder and have production modifiers: there may be 1h lost due to rounding in those cases.

Other changes made:
War Elephants:
7 str, +75% vs. Mounted
Ballista Elephants:
8 str, +75% vs Mounted (Buff b/c of the Vulture Problem if they didn't get this), still target Mounted units first outside of a city.
Spear/Pike changes relative to WE have been removed.

Praetorian nerf removed. Something may happen to the Praet, but it won't be a city attack penalty.

[T-hawk]

Quick Speed: 20/33/46/59. (No modifiers. And, yes, 59 is less than 67% of 90. Blame Firaxis for using integers in situations where division happens)

Ha, actually this is my fault and I see why it happened. My code applies the 2/3 nerf modifier for each population point, before multiplying by the number of population killed. So you get 20 + (20 * 2/3) + (20 * 2/3) + (20 * 2/3) = 59. I can fix it to give 20 + (20 + 20 + 20) * 2/3 = 60 as it should. Check your email in a few minutes.