Crop Price Analysis

[AgroDude]

EDIT: There is a selling summary and median price (seen during testing) here (updated for 1.2.0.0): viewtopic.php?f=963&t=185048&start=45#p1452250

Prices are for EASY. Divide by 2 for Normal, and divide by 4 for Hard.

Well... that was interesting.

It took nearly 3 hours just to compile 3 game-years of data for each of the storable crop types, but I now have a huge spreadsheet with max/min prices for each month and some calculated data for 3 years worth of prices.

Here is the data for wheat.

There is some random variation in the prices, and it CHANGES BETWEEN SESSIONS OF THE SAME SAVE GAME! Very important.

What I don't think is in doubt however, is when to sell. While there will always be some fluctuation in the price, it appears the max/min over the year are relatively fixed. For wheat, it appears if you sell in Jan or Mar you can't go wrong and will get close to the best prices at this time.

The average of the average of the high and low prices is actually closest to the best price. On that basis, January is the best month to sell wheat.

January data is 1406, 1269, and 1324 (small variation), whereas March data is 1104, 1645, and 1337 (large variation).

[FarmCatJenkins]

Nice work. Just can't wait for great demands now!

[AgroDude]

Nice work. Just can't wait for great demands now!

Thank you! I cleaned up the chart and added two more pieces of data.

I need to mess around with the data to get Excel to do what I want, so it will be another day before I have complete charts for all the crops.

I'll create a summary of the best time to sell each crop type, as I think that will suffice for the majority of people.

I have a theory, but it'll need to wait until the charts are completed to see if it works out.

[StoneTheCrows]

[george.earlslight]

https://docs.google.com/spreadsheets/d/ ... 2032357996

[AgroDude]

https://docs.google.com/spreadsheets/d/ ... 2032357996

My data is very much at odds with what is presented there.

Yes, base files form the basis, but I've taken the actual data from the game for 3 years. There are patterns that are clear.

Interestingly, when the spreadsheet you linked states is the best time, vs. my actual game state data, my spreadsheet shows it is actually a bad time to sell.

Remember, my data is based off what the game actually gives as a price, to include any computational trickery beyond raw fixed data values.

There is a random element to the pricing as well, which I have attempted to compensate for through multiple runs and multi-year data collection, followed by statistical method of averaging averages to smooth out the fluctuations.

[AgroDude]

My High and Low averages are a good fit for the 130% data point in the linked spreadsheet. It is very close indeed.

Also interesting is the Low Average in my data set (Easy mode) is equivalent to the High Average (Normal mode) at 130%.

I'm satisfied my data is accurate.

I have seen Soybean as high as 3718 in testing.

[NateS]

Pulling data from the xlm files shows that the best time to sell wheat should be Mid Winter with a price of 21% over the baseline. The lowest would be late summer at 81% of the base price. Wheat, Barley, Oats, and Straw should all fluctuate the same according to the xlm files.

Soybeans have a much higher fluctuation with the high being 42% over baseline in early summer and 64% of baseline in late winter! Selling Soybeans at the wrong time can cost you more than double!

[Seedy]

I would be happy if the in game chart would only show actual historical data, instead of whatever fantasy it shows now.

[AgroDude]

Pulling data from the xlm files shows that the best time to sell wheat should be Mid Winter with a price of 21% over the baseline. The lowest would be late summer at 81% of the base price. Wheat, Barley, Oats, and Straw should all fluctuate the same according to the xlm files.

Soybeans have a much higher fluctuation with the high being 42% over baseline in early summer and 64% of baseline in late winter! Selling Soybeans at the wrong time can cost you more than double!

I will check here.

From my testing, absolute best price for wheat occurred in March, with the highest lowest prices occurring at this time. It concludes that March is the best month to sell (my graph for wheat is in the first post).

Reading XML files is all well and good, but do they show the random variance or account for any other computations after the values have been read? I suggest not.

I'm apparently unable to attach PDF documents. I'll post a bunch of images shortly.

[mackintosh]

Looking forward to your analysis. I wonder if the game gets confused by season length. I have seen some crops hit high prices at a time the charts suggest they absolutely should not. Maybe the chart and xmls take a 1 in-game day per month into account and calculations get thrown off when we use different time periods. Perhaps someone forgot to test time scaling. Or maybe it's just RNG throwing oddballs.

[george.earlslight]

https://docs.google.com/spreadsheets/d/ ... 2032357996

My data is very much at odds with what is presented there.

Yes, base files form the basis, but I've taken the actual data from the game for 3 years. There are patterns that are clear.

Interestingly, when the spreadsheet you linked states is the best time, vs. my actual game state data, my spreadsheet shows it is actually a bad time to sell.

Remember, my data is based off what the game actually gives as a price, to include any computational trickery beyond raw fixed data values.

There is a random element to the pricing as well, which I have attempted to compensate for through multiple runs and multi-year data collection, followed by statistical method of averaging averages to smooth out the fluctuations.

Sorry for posting the link and flying off, reading the forum at work can be problematic at times.

Surely, the random factor exists but after averaging over a few years the results should be conclusive and that's what should be taken as a selling time indicator.

I'm not sure I understood what you mean about the spreadsheet vs your data mismatch.
The spreadsheet lists January as the best month for Wheat, while you also note January as the best month.
One could wait over the years to get peak values like that 1645 on March, but that's just gambling.
It would be like a farmer waiting for March to have a huge hurricane that would destroy everybody else's silos and take advantage of that.

[norfolk farmer]

Looking forward to your analysis. I wonder if the game gets confused by season length.

I wonder the same and if this is the underlying reason that Giants were so reluctant to increase the season lengths initially which was also said was because of the new productions.

[AgroDude]

I've not forgotten - just ran out of time today. I'll post them all tomorrow, and respond to everyone!

[humbe]

https://docs.google.com/spreadsheets/d/ ... 2032357996

This spreadsheet looks great at least. Is it based only on game XML files? I started my first game on hard, so I've got a sizable loan, poor equipment and little land, so actually figuring out when to sell is important, as even the best price is way below average price on normal.

One could hope to see the multiplication factors for each month, so you could see how much the difference between the months actually are. That would allow you to make an informed decision for whether -10% price is acceptable if you get the money 4 months earlier. But this'll do nicely

As there's a lot of info in spreadsheet, I'm guessing it's based on game XML files, and that the random sampling of 3 years done by original poster don't match up exactly because the random component on top needs sampling for a lot more than 3 years to average out.