Micromanagement is alive and well in Civ 4!

I dowloaded the spreadsheet, and it looks very useful, but unfortunatelly my english is very weak, since i learnd it by myself. Im here wrestling my enligh dictionary and i cant understand how the spreadsheet "works". I changed the number in the yellow cell to 90 since i always play in marathon, but i cant understand how a city with 250% production bonus can have a efficiency of 29% by rushing something when 1 hammer is left. Dosent the hammers from the sacrifice get multiplied as well? So thats 90h x 250% = 315 hammers out of 1 pop lost, thats alot! Rushing 1 pop when finishing something in order to speed the construction of a wonder or something seems a good idea, since so many hammers will overflow; but not according to this spreadsheet. Can someone give me 'pratical examples'?

Those words in the spreadsheet have each 4 or so different meanings in my dictionary, i tried writing them with a different combinations of those meanings and that didnt helped! damn language barriers..

I'm at work now, but this evening I'll check up on the output from the scenario you describe. I haven't done much testing on Marathon, so there's a decent chance I'm wrong.

It sounds like you're reading the spreadsheet accurately for the most part, though. I'm not sure which words with 4 meanings you're talking about... let me know and I'll try to clarify.

That 29% efficiency is based upon getting only 90 hammers when you expected 315. That part I'm sure of. What I'm not sure about, is how the overflow hammers would actually be calculated. My current understanding is that, if you had 89 overflow with a 250% bonus, the 89 would have the bonus un-applied and be reduced to 25. Then, they'd re-apply the bonus for the next item in the queue (still 250% in this case) and get back up to ~89.

EDIT: Confirmed that my understanding was reflected in-game. On Marathon with a 250% bonus, rushing something with 1 hammer remaining only gets you 90 total hammers for your 1 population spent.

Wow, you got way to much time on your hands. And I thought I was a civ junkie.

Brancaleone said:

but i cant understand how a city with 250% production bonus can have a efficiency of 29% by rushing something when 1 hammer is left. Dosent the hammers from the sacrifice get multiplied as well?

Click to expand...

Short answer : no.

In plain (to you) French :

L'ordinateur calcule combien do population ça te prend pour terminer la production. Pour ce calcul, il tient en compte ton bonus. Ensuite, il te donne une quantité de marteaux qui est le multiple de 30 (90 à la vitesse marathon) le plus bas nécessaire pour terminer ta production. Ce montant n'est pas multiplié par ton bonus. Donc, s'il reste 90 marteaux ou moins à faire à la production, ça te coûtera 1 point de population et ça te donnera exactement 90 marteaux, peu importe le bonus que tu as normalement. Dans le cas d'un bonus de 250%, tu obtiens donc 90 marteaux au lieu de 90 fois 350% (250% plus le 100% de base). 90 marteaux au lieu de 315, ça donne 28.6% de ce que tu devrais avoir si le jeu était programmé comme il faut.

Malekithe, you might want to make tests at different speeds and refine your table accordingly. Because of the way the game multiplies (i.e. without keeping fractions between operations), the numbers are sometimes slightly off. In most cases (e.g. case A below), it doesn't make a big difference. In some cases however (e.g. case B below), it makes a pretty big difference.

Here are some examples at normal speed.

A. 25% bonus, get 60 hammers from 1 pop (i.e. 2 for the price of 1)
You can do this with anywhere between 31 and 38 hammers left to production. Normally, it should have been between 31 and 37.

B. 50% bonus, get 120 hammers from 2 pop (i.e. 4 for the price of 2)
Normally, you shouldn't be able to do this at all. With 2 pop, you should always get exactly 90 hammers, getting no bonus and no malus. However, due to rounding errors, with 91 hammers left (possibly also with 92, i haven't checked), you can get 120 hammers while only paying for 60 base hammers. Precisely 91 hammers is a new sweet spot i'll now be looking for at 50% bonus, since i discovered it yesterday.

Actually, (to my surprise) both of those cases are covered in the spreadsheet, at least on normal speed. I try to truncate the numbers as appropriately as possible. The main thing I'm unsure of is the order of opperations. The formula I've settled on gives correct results for every scenario I've tested on normal. I'm positive there are some issues at other speeds, though. The code in the SDK that relates to pop-rushing has been a pain to crawl through (probably why the bugs are there to begin with). I've given up at it twice so far, but each time I've come away having learned something new. I'll give the code another look this weekend and see if I can't glean anything new.

The spreadsheet may be especially inaccurate for Epic, as the 44 hammers per pop is a bit misleading. In reality, it's (30 / 0.67) hammers per pop which gets rounded down to 44. I imagine, though, if you sacrifice enough population, you can get around the rounding "error" and get back a few hammers here and there. There are strange inaccuracies on Quick as well, leading me to believe that I'm applying the game speed modifier at the wrong time. Marathon seems to be fairly accurate, though, at least as far as I've tested.

Also of note, in addition to your scenario B, there's another very efficient spot at 61 hammers left with a 100% production bonus. You get 90 hammers, having only sacrificed one point of population. Probably not as useful, as few cities, if any, in my games ever get much above a 50% bonus.

EDIT: The formula I was using was surprisingly close. After figuring out what the code was doing, though, I've put in the correct formula. Any errors in the spreadsheet now would be very surprising.

Nice catch. Actually, i find it much more useful than the one i pointed out, because any city with 50% bonus can already get 2 for 1 by whipping 1 pop instead of 2, while a city with 100% bonus can't get any extra at any point normally.

In my current game (where i found out about the 91 spot), almost all my cities have 50% bonus (organized religion + forge), but my capital is at 100% (because of bureaucracy). Now i've already made all the buildings that can be made in my capital, and it's at about +5 unhappiness from pop rushing alone; but if i had known about this 61 spot earlier, it would have come really handy.

I should really install winzip on my computer and take a look at your file!

Hmm... I've always thought slavery was overpowered, but I never really understood how much.

Unfortunately, I don't think I can bring myself to use this strat on marathon game speed, with 30 turn unhappiness penalties. I know the game speeds are theoretically balanced, but 30 turns is a really long time.

However, it may give me an excuse to switch back to normal speed. The main reason I play marathon is that it makes early war so much easier, because your units only take 100 years to march to their destination, instead of 1000. But this might be offset somewhat if I pop rush units, something I really never do.

What do you think?

RemoWilliams said:

However, it may give me an excuse to switch back to normal speed. The main reason I play marathon is that it makes early war so much easier, because your units only take 100 years to march to their destination, instead of 1000. But this might be offset somewhat if I pop rush units, something I really never do.

What do you think?

Click to expand...

I think it's really better to try to ignore these glitches, until (hopefully) they get fixed. You're not going to enjoy the game more if you turn it into an artificial exercise in exploiting bugs. At least, I doubt it.

I think if you want the easiest game (you said you play marathon because it's easier to wage war), then you should remain on marathon AND whip. 30 turns on marathon is no worse than 10 turns on normal speed. Like it says in the article, whip 'til your hands bleed!

RemoWilliams said:

Unfortunately, I don't think I can bring myself to use this strat on marathon game speed, with 30 turn unhappiness penalties. I know the game speeds are theoretically balanced, but 30 turns is a really long time.

Click to expand...

I always play marathon speed because otherwise the eras just fly by and units become obsolete too quickly for my taste. Even on marathon speed, though, I'll whip like crazy. Thirty turns sounds like a lot of time, but things in general are slowed down. The way I look at it, instead of asking "How long will whipping cause unhappiness?" I ask "What are my options?" Invariably, whipping turns out to be one of the best options. If so, then it doesn't matter that the unhappiness lasts 30 turns. The best option is the best option.

"For a nice spreadsheet showing the points that provide more hammers than you paid for at each game speed, see this file graciously provided by Malekithe."

I tried to get the file.. and got this:

"Invalid Attachment specified. If you followed a valid link, please notify the administrator."

I just fixed the link. I have no idea why the URL changed. I hope it won't keep doing this.

Zombie69 said:

I just fixed the link. I have no idea why the URL changed. I hope it won't keep doing this.

Click to expand...

Thanks, man!

Zombie69 said:

I have no idea why the URL changed. I hope it won't keep doing this.

Click to expand...

Probably because I updated it over the weekend. I don't anticipate any more updates, so this shouldn't be an issue anymore.

Since whipping is best exploited when you have a production bonus, I'm wondering how much you prioritize getting to Metalcasting and forges. In general, is this something you do before heading for writing the tech paths that opens up?

Very good question!

I prioritze granaries, as those are much more important than the 25% bonus. Granaries always double the power of whipping, while the 25% bonus only doubles it if you rush one pop at a time, and requires micromanagement which may otherwise be detrimental to you (e.g. forcing you to wait a few more turns before whipping, thereby getting the building later, although for lower cost).

I do try to get either forges (usually a priority if industrious, because of cheap forges) or organized religion (more a priority if spiritual) sooner than rather, but i don't plan my entire tech tree on getting one of those at the detriment of everything else.

Since i play at very high difficulty levels (immortal and deity), happiness is a huge issue. This is compounded by the use of slavery. Therefore techs that allow me to increase my happiness limit are often more of a priority to me than those that increase my whipping production. Sometimes though, these are one and the same. If i have one or more of gold, silver or gem, forges can provide happiness as well as more efficient whipping. In that case, i definitely prioritize metal casting more.

As with everything else, it depends. But certainly i prioritize metal casting more than someone who wouldn't use the whip this way, even if i don't necessarily make it a priority from the very start.

At game start, my first two priorities are always :
1. Bronze working, so that i can switch to slavery and start whipping, which is efficient even without bonus and without granary. The facts it also allows me to chop, find copper and build axemen are all very nice, but even with no forest and no intention of using copper, i still would go for bronze working first if only for slavery.
2. Pottery, for granaries. Access to cottages is also very nice, since i always play financial civs.

After that, it depends on many factors, and i'll probably get metal casting sooner than later, but it's not an immediate priority.

Thanks for the prioritization insight. Very sensible.

[Edit: Here begins a series of posts wherein I disagree with Zombie69 about some details. You might enjoy the reading, but the end result from here to post 276 is that 1) Zombie69 is right. Follow his advice on whipping if you want to optimize your city's productivty and 2) in a particular case where you might (I did) think working mines for productivity in a commerce city would be a good idea, whipping still came out as about 16% better in both productivity and commerce.]

I'm still working out the whipping micromanagement ideas for myself. I'm trying to frame it in terms of "opportunity cost". Bear with me as I think as I write, if you will.

Say you have a city that's happiness limited to size 5 and you've got at least that much health. Using the whip every 10 turns essentially makes this a size 4 city. It sounds like the best return you can get from your whip with a 25% production bonus is 60 hammers for 1 pop. Assuming your food situation is such that you can grow back that one pop in the 10 turns of unhappiness, you're essentially trading 1 pop point for 60 hammers/10 turns. That's a mine and a half, which is awesome and is why you exploit it.

If you whip more than one citizen, your opportunity cost goes up because it'll take you time to bring that citizen back online. So, if you whip from size 5 to size 3, you're out 2 tiles instead of one until you get to size 4, your new happiness limit.

You micromanage so that on the 10th turn, you grow from size 4 to size 5 and so never have to pay civic costs on an unhappy citizen.

So from an opportunity cost point of view if you whip at size (NormalMaxPop=5), the tradeoff is:

Whip Option: However many hammers you get for your whip

versus

Opportunity Cost: Last Marginal tile's use for 10 turns (due to unhappiness) + Other marginal tiles use for fewer than 10 turns because you have to grow your pop back to NormalMaxPop-1.

If you are keeping an unhappy citizen or two around and whipping them, then your opportunity cost is your Last Marginal Tile x 10 turns plus the increased civic maintenance of the unhappy citizens (assuming you have surplus food to feed then and still grow back to the point where you whipped).

I seem to recall reading that although it depends on the cost of the civics you're running, civic costs are around 0.5 GPT per population point. That would imply that unless your marginal tiles are unimproved (and why would you do that if you can at all help it?!), you are probably better off always running your city at NormalMaxPop-1 or greater and only whipping citizens who are unhappy for reasons other than previous whips.

Do I have it?

Compromise said:

Assuming your food situation is such that you can grow back that one pop in the 10 turns of unhappiness, you're essentially trading 1 pop point for 60 hammers/10 turns. That's a mine and a half, which is awesome and is why you exploit it.

Click to expand...

6 hammers per turn is a lot more than a mine and a half. Heck, it's more than two mines. In fact, it's better than three mines! Imagine 2 grassland mines. They give you 6 hammers, and cost you 2 food (because staying two pop below instead of working them gives you a net food surplus that's two points higher). 3 grassland mines give you +9 hammers and -3 food. If you consider 3 food to be worth the same as 3 hammers, then that's the same as +6 hammers. If like me, you consider one food to be worth more than 1 hammer (notably because you can whip 11 food into 60 hammers), then you must conclude that 6 hammers per turn is actually even better than 3 grassland mines. As for plains mines, they're even weaker, so let's not even go there.

The only reason not to whip every 10 turns is if the city has very low food and can't grow sufficiently to get to its max size. For any city that has at least +3 food (which only requires at least one tile with 3+ food, since the city tile itself gives you +2 food), you're always, and i do mean ALWAYS, better off whipping at least every 10 turns, and definitely whip as soon as there's no unhappiness anymore from the whip. For any city that has at least one 3+ food tile, you should always have at least -1 unhappiness due to the whip, permanently.

Hi Zombie69,

Thanks for the replies; I appreciate the attention. But I hope you'll bear with me a little longer, because I think I'm still not seeing the whip as the super deal you have described it as, at least not without a good amount of micromanaging.

I've set up a little experiment on paper. (Away from the game at the moment.) nF means n food. nH means n hammers. Size 6(30) means size 6 with 30 food in the food bar. +nF or -nF means the extra/deficit food, so a farmed grassland is 3F because it produces 3 food, but +1F because it takes 2 food to feed the citizen who's working that tile. Here's the setup:

Look at one city and concentrate only on food an production (we're trying to maximize production. Your only building is a granary. Playing at normal speed. Your happiness limit is 7. Your health limit is at least that, say 10.

You have just grown your city from size 6 to size 7, so all your people are happy and working. More people would be unhappy. It took (20 + 6x2 =) 32 food to grow so you have 16 food in the bank from the granary: size 7(16).

The tiles surrounding your city are 1 grassland pigs with a pasture (6F), and as many mined grassland hills (1F3H) and irrigated pre-biology farms (3F) as you want. Your city is located on a normal square and is 2F1H.

To maximize production and not grow, you would set your 7 citizens to work the pigs and 6 grassland hill mines. Per turn, and including the city center, that's:

Center(+2F1H), Pigs (+4F), 6 GHMines (6 x -1F3H): Total 0F 19H

If you work that setup for 10 turns, you get no growth and 190 hammers.

Below, I've separated the Turn numbers (what you see when you look at your cities during your turn), from the calculation (where the game runs the update and the AI's turns.

In the whip case you have:
Turn 0: (whip) 7(16) --> 6(16) + Whip Hammers
. . City(+2F1H), pig(+4F), 5 grass hill mines (5x-1F+3H): +1F +16H
Turn 1: 6(17)
. . City(+2F1H), pig(+4F), 5 grass hill mines (5x-1F+3H): +1F +16H
Turn 2: 6(18)
. . City(+2F1H), pig(+4F), 5 grass hill mines (5x-1F+3H): +1F +16H
Turn 3: 6(19)
. . City(+2F1H), pig(+4F), 5 grass hill mines (5x-1F+3H): +1F +16H
Turn 4: 6(20)
. . City(+2F1H), pig(+4F), 5 grass hill mines (5x-1F+3H): +1F +16H
Turn 5: 6(21)
. . City(+2F1H), pig(+4F), 5 grass hill mines (5x-1F+3H): +1F +16H
Turn 6: 6(22)
. . City(+2F1H), pig(+4F), 5 grass hill mines (5x-1F+3H): +1F +16H
Turn 7: 6(23)
. . City(+2F1H), pig(+4F), 5 grass hill mines (5x-1F+3H): +1F +16H
Turn 8: 6(24)
. . City(+2F1H), pig(+4F), 5 grass hill mines (5x-1F+3H): +1F +16H
Turn 9: 6(25)
. . Work City(+2F1H), pig(+4F), 3 farms(3x+1F), 2 GH mines(2x-1F3H): +7F +7H
Turn 10 6(32) --> 7(16) pop growth
. . This is back to where we started, so we can whip again (or not!)

After ten turns...
Non-whip total: 0F +190H
Whip total: 0F +(16x9 + 1x7 + Whip = 151+Whip)H

So, you get 190 hammers over 10 turns from not whipping and 151+whip hammers from whipping and working as many production tiles as you can. You need to get 39 un-bonused hammers per whip to get the same total number of hammers as the no-whip case.

This implies that it hurts not to be working improved tiles. And unhappy citizens, such as those incurred by cracking the whip, can't work improved tiles.

There are many things neglected in this analysis:

1) Exactly what tiles you have--and whether or not you've improved them--makes a difference.
2) When you get those hammers can make a (big?) difference. (Aaahh...sneak attack!)
3) What your population size is and what your happy/health limits are make a difference.
4) Whether or not there is a production bonus makes a difference, and when you have a production bonus, you're also looking at the possibility of roundoff losses.

But unless I made some error--definitely possible!--I think you have to be exploiting an error in the pop-to-hammers calculation to get more total hammers over the total turns. I think the absolute best you can do is 48H/pop. Here, that would be a total of 211 compared to 190 hammers over 10 turns, or about 2H per turn (and you get many of them up front). But you have to get more than 39 hammers per whip to get a better-than-not-whip total number of hammers. [Edit: I had originall used 60H/pop, but Zombie69 points out that it is 48H+25% bonus. So to be fair you compare to 48 base hammers. This particular example doesn't use a production bonus, but the issue is discussed below.]

I realize that you need to have a production bonus of some kind to exploit the pop->hammers bug, but here I think I'm showing that without a production bonus, whipping will actually net you fewer hammers than you'd otherwise get. (Though again, you do get them sooner!)

For the example above, if you added a forge for a 25% production bonus, you'd lose 0.75H per turn in the no-whip case because of roundoff, but this is just an example. In real-life, you would have different tiles and might be able to micromanage to get most of that back, so maybe 6 or 7 extra hammers over 10 turns for the no-whip case. That's about the bonus (7) that you'd be getting from the population point too if it were calculated correctly.

It seems to me that unless there's an error in my analysis (and somebody please check it to find out!) that it is possible to use the whip to get more hammers than you otherwise would, but:

1) you need lots of micromanagement,
2) you need to exploit the pop-to-hammers calculation error
3) the max benefit is around 2H/turn which is less appealing the bigger your city is.

If that analysis is correct, then the take-home lesson is, Whip when:

1) you need/want it finished right now
2) the last tile your city is working is unimproved
3) you really like optimizing and are willing to extremely micromanage and take advantage of the bug for a max 2H/turn/city increase in hammers.

If I made an error, let me know by post or PM, and I'll correct the analysis so that readers don't have to flip between posts to find out the truth.