Farming Efficiency and Performance

Support Previous Build Reports
1

Here are some interesting numbers for you. As a disclaimer, I’m not that smart. so this discussion is meant to foster more intelligent people expanding on my ideas and findings. I want to learn too!

I created 1x1 farms in the space of an 11x11 farm. I compared the efficiency of this farm next to an 11x11 farm to get some numbers. I’ve found that a 1x1 farm at the size of 11x11 is 83.3% more efficient at space usage, 71.4% more effective for food harvested, and overall better for soil quality. Here are details and pictures:

QLpGtam.png1410×571

Food Harvest
1x1 = 12 Baskets (took 1 from 11x11 crop, hungry halfway through harvest, fully harvested early)
11x11 = 7 Baskets (took 4 from 1x1 crop)

Usable Quality (Poor, Fair, Good, Rich)
1x1 = 28, 39, 29, 25
11x11 = 30, 35, 1, 0

jdRiwsa.png1215×505

I should note these numbers will also change depending on the stats of your farmer(s). I don’t know of a resource that thoroughly explains how each stat affects actions in-game, and I recall hearing a dev mention this was left hush-hush so we would learn on our own. The Hearthling I used for farming had the following stats with no additional gear:

lMt8WmX.png719×862

Due to obvious flaws in these numbers (e.g. my farmer getting hungry halfway through harvesting a basket, pulling food from the other crops to fill a basket, etc.) I am going to run through some control tests. These will be performed like this:

Control test will consist of one farm type (11x11 or 1x1), harvested only after the Hearthing eats, with a tally of baskets gathered each season for 10 seasons (a season consist of a fully planted and harvested round), with screenshots of before and after each season for soil quality (numbers may or may not be provided as they’re painful to gather).

I’m aware this is Alpha 10, and new game features may break this or otherwise make it ineffective. I welcome your discussion on this matter anyhow. Thanks for reading.

Updated — Please see this reply for performance issues related to this method.

4 Likes
2

we already know this is going to be changed when they require water to be present when farming so spaces in between the plants will be necessary in the future.

1 Like
3

How about performance? Does having lots of 1x1 farms make the game slower than having 1 big farm? I’m curious :neutral_face:

2 Likes
4

@AdriannaVentara This is also dependant on how much space is required for water to affect the surrounding farm. Will one row of water work on two or more rows of farm, or will there literally have to be a row of food next to a row of water and so on? As an aside, you can still do a single row of 1x1 farms with a row of water then another single row of 1x1 farms and be far more efficient and productive.

@Relyss I will test this now. (Updated — The game becomes almost unplayable when the farmer is plowing the second 1x1 farm. See below image. Given this was also while all other workers were cutting down trees. I’m going to let the game perform badly and see if it gets better after the initial plowing of the field.)

2 Likes
5

You make a very bold bold statement, but you’d need to do more trial runs for it to be of any statistical value! :smile:
On average, doing this will make your farms 100% more efficient, given a large enough amount of trials. The random variations in soil quality, etc., will average out between trials.
I had brought this up a while back, as it does bring in a level of ‘cheapness’ to farming, but any benefits given by the doubling of food output are negated by:
a) the future plans for aqueducts and irrigation
b) the need to change 11 fields when you wish to switch crops
c) the need to place 11 fields when you can drag two of them out, especially since space is not an issue at this point!

Anyways, it would be nice to see a follow-up of this when more water features are implemented!!

Edit: You masochist! 121 farm plots?! I thought you were doing 11 ‘1x11’ plots!!!

6

The statistics I provided were from a single initial run. As I said, I’m not a smart individual, I merely ran the numbers through a calculator when comparing a single harvest with a 1x1 and an 11x11 farm.

I know full well they hold little value right now, which is why I wanted to run through a number of control tests. Given the recent performance hits I also found using this method, the control tests will have to wait, or may not be happening at all, so anything more with this will wait until these additional water-related features are implemented.

I’m still waiting to see how the performance settles after a complete plowing and planting of 4 farms, but seeing the growth rate is so quick, further planting will be required and thus further performance hits. In the end, this method sounds nice, the numbers hold some weight, but performance is impossible otherwise.

2 Likes
7

It would be interesting to see if smaller plots have an inherently higher ‘value’ in terms of their soil quality, but the grindwork to get through all the trials would be ridiculous! The advantage could be something coded in (or miscoded, even, if you dare anger the farm master). I think looking at the game code could be of use here, but I don’t have the know-how to make anything of it

1 Like
8

I have 4 1x1 farms. Performance is degraded when plowing, planting and harvesting. After any of those steps the game runs fine, but when it comes time for any of those say goodbye to a playable game. Here is a screenshot that may help (I noticed this was what happened during all three of the above steps):

After all the crops were mostly planted (my farmers seemed to have forgotten five plots needing to be planted), I demoted all farmers back to workers, and made everybody cut down trees. Performance worked as usual with no issues. This was while the plants were growing, so that doesn’t seem to have any effect on performance as I see. Here is the finished work:

1639×408 36.1 KB

Anyhow, I’m done testing for now since performance makes this a lost cause. Thanks again for reading! I welcome continued discussion if there is any more to say.

4 Likes
9

i think it would be interesting, and add more depth to the game, if each crop/food had its own preference.

1 Like
10

Since performance bottlenecked when plowing, planting and harvesting the second of my 1x1 farms, I wanted to see how well the game performed with many regularly created farms. I’m sad to announce, my 11x11 farm broke the game.

gjmqh1G.png1038×720

It started well before the above picture, in fact, it started right about the time of this screenshot:

oviQctp.png992×567

I can’t even count how many farm plots this is. Somewhere between 10-12? Something needing to be tested further I think is if the already designated zones, but not yet plowed plots, are part of the cause. Will the game bottleneck with 10 farm plots by itself?

I guess before I submit a bug I need to test this. I noticed that lua, lua cb, and idle are all fighting for what seems to be 150% resources. (Don’t trust the overlay in the second screenshot. That’s where the fun was just starting. When you get to the first screenshot, all the plots being plowed and stuff, is when you start to see 150% or more being used by primarily these three processes.)

3 Likes
11

Can we get this moved to the Performance category since there are obvious bottlenecks involved? I started this topic with the intent of exploring maximized farming but uncovered other issues.

12

alrighty! i did it for you, but that doesnt seem like the right category… @Relyss @SteveAdamo, thoughts?

1 Like
13

If it involves lag and throughput of the game, then it’s fine under the performance category.

2 Likes
14

So, 100 farm plots is bad, and it looks like only 22 had been partially plowed in your second screenshot. Of those 22 the combined number of actual farmland appears to be 12.2-ish 11x11 plots worth.

15

Right. Now the question, and one I will discover the answer to tonight, is do the unplowed farms affect performance too? Can I create 12 farms with no additional plots created and expect the game to run the same?

I understand some of this performance loss is going to be related to my system, running a Core i5-5300U at 8GB RAM and an Intel HD 5500, but those specs while not gaming worthy aren’t bad. I have to point out that the game shows 150% or more resource usage between idle, lua and lua cb. I think the answer is somewhere in there, but what do I know?

2 Likes
16

The physicist in me loves this thread. Nice work Jeb

2 Likes
17

So four tests for all possibilities: 12 full farms may be the cause, 22 incomplete farms, or some combination of those two with several more plots yet to be plowed.

18

The game runs fine no matter how many incomplete farm plots have been created. The performance lags only spike when a farmer is actually plowing, planting or harvesting, and only when there are about 10+ farms. This is what I’ve gathered so far anyhow.

19

I’m just gonna jump in here with some information @Albert provided in the last twitch stream. There is a bug (which he’s fixed for the next patch I think), where every time the Farmer brings out his hoe to till a patch of soil, he creates a new hoe and then drops it without actually destroying it (it’s just floating around somewhere).

So if you try and till 1000 tiles of farmland, the game is going to gum up with 1000 copies of a farmer’s hoe. Which is obviously really bad for performance.

There may still be a performance issue with large numbers of farm zones (in fact it’s probably really likely), but its hard to tell if the farmer’s hoe memory leak is in effect at the exact same time.

5 Likes
20

[quote=“Tuhalu, post:19, topic:13560, full:false”]
I’m just gonna jump in here with some information @Albert provided in the last twitch stream. There is a bug (which he’s fixed for the next patch I think), where every time the Farmer brings out his hoe to till a patch of soil, he creates a new hoe and then drops it without actually destroying it (it’s just floating around somewhere).

So if you try and till 1000 tiles of farmland, the game is going to gum up with 1000 copies of a farmer’s hoe. Which is obviously really bad for performance.[/quote]

I was just about to bring this up. However, I do think there are two different performance issues, since this issue would exist no matter what size the farms were (1 ‘11x11’, 11 '1x11’s, and 121 '1x1’s would all have the same number of “ghost hoes”.)

1 Like