Chia Calculator considered harmful

TL;DR: Set weekly network growth on Chia Calculator to 69% or 146%, and round the expected XCH it displays down to the nearest multiple of 2, e.g. expect that 0.9 XCH/month is most likely 0 XCH/month.

I’ve been working hard on modeling the likely profitability of my farm. At first Chia Calculator seemed really helpful - it does a lot of complicated calculations and even accounts for the growth of the network over time! However, after I had already spent a few hundred dollars without winning any blocks and wanted to check the numbers before spending more, I discovered an enormous problem with the Chia Calculator: the default network growth rate is off by more than 4x and maybe as 10x, which makes a world of difference. I ran a simple exponential regression over the network size data since transactions launched and found that the network has been growing nearly 69% weekly, much faster than the 15% used by default by Chia Calculator. Running the same regression only since network growth recovered after the outage on 11 November shows 146% weekly growth rate, although that’s based on only 10 data points across two days and may partly reflect growth that was hidden by the outage so take it with a grain of salt. With weekly growth of 69% or 146%, you have to double the size of your farm every 9 days or 5 days respectively just to keep the same chance of winning a block! A plot loses 99% of its value every 61 or 36 days.

The second major problem with Chia Calculator is that it does a poor job of showing how skewed the probability distribution is. While the calculation of XCH/day or XCH/month seems to be a statistically accurate expected value, it distorts the fact that, until pools become available, XCH can only be won 2 at a time, so any XCH/day or XCH/month value below 1 really means that you are most likely to win 0 but you do have a small chance of winning a block worth 2 XCH. Chia.Garden’s probability calculator does a better job of visualizing the likelihood of the possible outcomes but doesn’t account for network growth so I can’t recommend it either.

6 Likes

Problem with All calculators, you cant predict network grouth, because it is made by people and you cant predict people emotions. Soon i think expansion will slow down, as HDD prises go roketing high already.
Also some people will leave the network as expectationas are always high, and people want go rich fast.

2 Likes

That may be true but Chia Calculator tries to adjust for that by capping the exponential growth phase of the network and switching to linear growth after that. My point here is that their 15% default growth rate is irresponsible and misleading given that growth since the launch of transactions has been 4-10x that.

I have pretty good calculator (done itself :D) and actually I take into considerations a lot of factors. Actually it helped me to do some good investments, as long as I could verify what is more profitable for me. I also calculated chance to NOT win anything in 6, 12, 24, 36 months, which was pretty useful to estimate the risk. But basically there are 2 high risk factors: netsize and the price. You can farm only 1 block in next 5 years, but if Chia will be priced for 50000$, it will be anyway worth.

Interesting. Are you willing to share anything about how you are modeling the price and network size?

Sure, basically I am always calculating break-even scenario for very risky factors like netspace and the price: considering constant growth how high should be the price to be at 0 profit after 6, 12, 24 and 36 months. Same doing for netspace: with constant value of chia coin, how fast have to netspace grow to be at 0 profit after 6, 12, 24 and 36 months. For example I see: after 36 months I’ll be on 0 profit with price around 600$ and netspace growth 0,4EiB daily over three years (with some kind of decay over weeks).

What is interesting, is that I wanted to buy SSD to plot faster, but from calculations straightly comes out, that I will earn more when I will buy for this money HDDs, instead of SSD. I also calculate ruin scenario, so what is the probability of situation, where I won’t farm any Chia. I see risks, see potential upsides, and then I manipulate the factors to see how much I can earn (what is the potential considering other cryptocurrencies). I even cut it into something like: considering this price and this netspace growth I can earn monthly passive money of X $.

If You will provide me Your maximum TB You want to have, size of Your current farm and speed of plotting (TB/day) I can show You some calcs :slight_smile:

2 Likes

I’m curious. I’ll bite.

Let’s try an example.

Maximum TB = 100TB
Current TB = 10TB
Speed = 2TB/day

Me too.

Maximum = 500/TB
Current = 25TB
Speed = 5TB/day

Let’s see - Your break evens:

Break even price (in $) 628,95
Break even growth (EiB per day) 0,72

So if You will be selling all chia You will earn for 628$ You will be on 0 after 3 years. On the other hand - considering stable price, netspace should grow 0,72 EiB daily to be on 0 after 3 years.

Below 3 scenarios:

  1. Expected case - in my opinion - most probable
Considering current price and 0,37 EiB daily growth Theoretical total earnings (gross) Chance to NOT farm anything Costs Net earnings (before fees and taxes) Total chia farmed
6 months 4 904,67 12,07% 3 833,78 1 070,89 4,66
12 months 6 739,49 4,24% 4 638,07 2 101,42 6,34
24 months 9 016,22 1,46% 5 442,36 3 573,87 8,44
36 months 10 804,56 0,64% 6 246,65 4 557,91 10,08
  1. Scenario with slower netspace growth and price which oscillate around today
Considering current price and 0,215 EiB daily growth Theoretical total earnings (gross) Chance to NOT farm anything Costs Net earnings (before fees and taxes) Total chia farmed
6 months 8 231,32 2,94% 3 833,78 4 397,54 7,84
12 months 11 737,13 0,40% 4 638,07 7 099,06 11,04
24 months 16 163,32 0,05% 5 442,36 10 720,96 15,14
36 months 19 663,37 0,01% 6 246,65 13 416,72 18,34
  1. Growth from expected scenario with price increase to over 5K $
Considering 5300$ price and 0,37 EiB daily growth Theoretical total earnings (gross) Chance to NOT farm anything Costs Net earnings (before fees and taxes) Total chia farmed
6 months 24 523,36 12,07% 3 833,78 20 689,58 4,66
12 months 33 697,43 4,24% 4 638,07 29 059,36 6,34
24 months 45 081,12 1,46% 5 442,36 39 638,76 8,44
36 months 54 022,78 0,64% 6 246,65 47 776,13 10,08

Some facts about speed plotting in expected case:

  1. If You will decrease plotting speed to 1TB/day, You will earn only 1 XCH less in the 3 year horizon
  2. If You will double your plotting speed, You will earn only 0,79 XCH more in the 3 year horizon

I have more calcs, but still working on them :smiley: (for costs I have just assumed similar to mine per TB and plotting speed)

1 Like

You’re probably right – nobody predicted this much network growth, this fast! I think filecoin was at 5 exabyte?

I don’t have much time now, but here is the expected case (maybe later I’ll do the rest):

Considering current price and 0,37 EiB daily growth Theoretical total earnings (gross) Chance to NOT farm anything Costs Net earnings (before fees and taxes) Total chia farmed
6 months 17 657,39 0,056261435% 15 335,12 2 322,27 16,89
12 months 26 743,65 0,000301760% 17 735,12 9 008,53 25,19
24 months 38 060,19 0,000001483% 20 135,12 17 925,07 35,66
36 months 46 961,69 0,000000023% 22 535,12 24 426,57 43,79

Sorry but if u didn’t saw that the netgrowth was still the one from the early days, you maybe should rethink investing at all, it was very obvious, the graph looks closer to 100% /week. Always double check everything when a lot of money is involved. Not only with crypto with everything, your bills your tax declaration, your paycheck. All the errors I catch during the last doublecheck over my lifetime adds up to a few thousand…

Seriously… I felt pretty dumb when I realized that I had skipped verifying such an important number and that it turned out to be so far from reality. Hence my decision to post this warning to prevent others from making the same mistake :slight_smile:

2 Likes

Even with the right growth rate, without knowing when the inflection point of the curve is or is going to be and without knowing the price in the future, its impossible to predict… maybe the price is going to raise by 15% a week and your calculation is going to turn out correct :wink:
I think that most people suffering from FUD over chia netspace these days are still going to see their ROI within a year or 2, and its still going to be profitable to just let the harvester run for years…