The Numbers Behind Food Consumption and Production

[REMworlder]

Lately I've been thinking about the relationship between food and work (I'm really excited about refrigeration and rotting raw foods coming next patch). I thought I'd try to neaten up and share some of the numbers I've gathered (though I'm sure many of you have a good idea of the numbers involved already). One question I had was how much food I needed to keep one colonist fed, below.

Boring Numbers: I used 80 saturation as a rough benchmark. Saturation is lost at .004 a tick, with 20,000 ticks in a day, for -80 lost each day cycle. Hunger strikes when only 30 saturation is left (70 saturation is lost). More saturation (let's say 10) may be lost as the colonist walks to the stockpile, grabs food, and wanders away to eat it because your colony doesn't have tables, you monster. I mention these assumptions to point out this chart doesn't cover all the nuances involved, such as colonists saturation never exceeding 100, or the productivity impact of colonists eating foods that don't satisfy their hunger 100%. Inputs represent the raw food involved in preparation.

THE INTERESTING STUFF:
-Basically, a single colonist averages just under a meal a day. One takeaway here is food preparation adds significant value to raw foods, essentially cutting the amount needed to feed a colonist in half. You can see that not only is eating raw food often unsatisfying to colonists, but it requires a lot of material just to satisfy the equivalent of a meal.

-Another application of this chart I've used is your starting 3 colonists can survive off of 24 survival meals for 10 days; I usually find a similar number scattered around the map.

-I was surprised to see a lavish meal uses just as many inputs as eating raw food (of course this ignores mood effects, meat vs plants used, etc). This, plus the cap on a colonist's saturation at 100, means lavish meals are pretty wasteful for purely hunger. Though Fine meals use raw foods most effectively, this doesn't take into account how long hunting takes and how unreliable it can be.

Here's the other side of the coin, growing the edibles. This should give an idea of how terrain affects output. Days to grow is used to figure out how many plants need to be planted for a colonist to be fed. Something to keep in mind is average output can be unreliable. Looking at worst yield can be used to figure out a minimum planting level. This also doesn't take blight into account.

The rough number of active plants needed to keep a colonist fed for a day results in the table above. I multiplied the Inputs Used for X Meals for X Days (table 1) by the corresponding Yield/Day values (table 2).

The takeaway for this graph is how many plants you need to average for each colonist, depending on the meal type used. If you're feeding your colonists raw berries on marsh, for example, you'll need to average 5.8 bushes per colonist, best case scenario.

The utility here comes in estimating how many plants you'll need to get by in the desert, or in reckoning how many hydroponic tables you'll need at a minimum (worst case) if you want to devote more of your production capacity to cash crops. As a rule of thumb, you can probably also get by with one hydroponics table (4 plants) per colonist, or 4.25 tables if you want to insure against (only) bad harvests.


Anyways, thanks for reading! I know this is a wall of text and numbers, so I hope it was at least somewhat readable. If you see mistakes anywhere please let me know! Here's the workbook I used, on Google drive, if you want to see what I used.

I realize these numbers don't cover all the variables that come into play in Rimworld, but I think these are definitely a start to making decisions with data. Let me know what topics you think deserve additional attention.

^{Here are some areas I'm working on:
-effect of blight and solar flares on minimum food productivity
-economic meal stockpiling, taking spoilage into consideration
-selling meals to traders versus raw foods and cash crops
-mixed food options
-trading off rich, normal, and marshy soil by distance}

Updated last two tables by removing min and average yield values since we're assuming plants don't get harvested til 100% grown. Thanks to Cimanyd for catching that!

[DaveStrider]

this is super helpful, it also makes the argument for berries over potatoes even stronger.
one question though, how does the nutrient paste resynthesis tech affect these numbers? if i'm reading this right the 10% decrease in input cost would make a paste dispenser more food efficient than simple meals. whether that small increase is worth the morale penalty is another question though

[REMworlder]

Good question! I totally forgot about that since I rarely use paste.

So basically a 10% reduction in inputs changes usage from 10 to 9.
http://i.imgur.com/NqNYHTM.png

It turns nutrient paste into (just barely) the most saturation-efficient form of food there is, needing around 4% fewer potato plants (2.7 vs 2.6) to maintain daily saturation.

[Cimanyd]

From what I've seen in-game, when a plant is harvested at 100% growth (like they always are automatically) the yield is always the max yield. So I think only the "best" numbers really matter there.

[REMworlder]

I'd never really noticed before, so I tried to reproduce it by setting up some hydroponics farms with hauling disabled (I also gave the colonists survival meals so they wouldn't eat any potatoes). Here's a picture of the results: http://i.imgur.com/qfgGjU6.png

While many of the amounts were cleanly divisible by 4.5, some amounts weren't, and the totals weren't either. Is it just a weird rounding thing, or am I missing something else?

[Cimanyd]

Harvesting one plant can't give 4.5 potato items, there aren't fractions of them. It appeared to always be either 4 or 5. I don't know how it's chosen, but if it's just random, you could get any number from combinations of fours and fives as long as it isn't too small, so the total numbers don't really matter, and I don't see any stacks there that wouldn't fit that.

Berries are max 3. They'd be simpler to try.

[Tynan]

Amazing analysis. Heck, this is really useful to me as a game designer.

It's also inspired me to rebalance nutrient paste for Alpha 8. It has sort of become useless, since the paste dispenser is expensive and nutrient paste is such bad food and cooking is so easy. Now paste is the most efficient food by a healthy margin, which should hopefully balance out the mood effect.

[skullywag]

Just when you thought it was safe to go back in the kitchen....

[REMworlder]

Berries are max 3. They'd be simpler to try.

You're right! All multiples of three. http://i.imgur.com/nwdLCEP.jpg This definitely makes things simpler, thanks for catching my mistake.

It's also inspired me to rebalance nutrient paste for Alpha 8. It has sort of become useless, since the paste dispenser is expensive and nutrient paste is such bad food and cooking is so easy. Now paste is the most efficient food by a healthy margin, which should hopefully balance out the mood effect.

Cool! Never really touched paste before because of cooking being easy, looking forward to using it more.

[Wex]

Actually, when you have a colony of 60+ colonists, having a paste dispenser is a blessing.
That or you can have 4 stoves full on preparing meals (both simple and fine).
If only the malus for the taste, after a research would go down a couple of points... 

[nuschler22]

Great research and thank you.

[silentdeth]

Do you have new numbers for alpha 8?

[amul]

So if a sun lamp covers (11x11 - 12(rounded edges) - 1 (lamp) = ) 108 tiles, then a grow zone designed around a sun lamp will feed about 20 colonists!

I'd love to see productivity per sun lamp added to your stockpiling tab, as I plan to absolutely keep an eye on this document! Thanks for doing the math!