I Spent 47 Hours Building a Mining Profitability Model. Here’s What I Found.

I Spent 47 Hours Building a Mining Profitability Model. Here’s What I Found.

The spreadsheet didn’t lie. But it nearly cost me my marriage.

It started with a simple dinner question.
“So what’s the actual ROI on those mining machines you keep talking about?”

I said the four most dangerous words in personal finance: “Well… it depends.”

Forty-seven hours later, I had 23 Excel tabs, 4,847 formula cells, and a model detailed enough to make an energy economist proud. I also had a wife who gently suggested I might want to “go outside for a bit.”

Worth it. Because now I have clarity.

Hero: The Model That Took Over My Life

Before this, I used online calculators like asicprofit.com for quick estimates. They’re great for snapshots. But I wanted to understand the engine under the hood.

So I built one.

I modeled 34 variables across hardware, electricity, difficulty, price, uptime, fees, and financing. Then I stress-tested everything with scenarios and Monte Carlo simulations.

Here’s what actually matters.

The Variables That Truly Drive Profit

After ranking every factor by impact over a 36-month horizon, the results were brutally clear.

Tier 1: The Big Four (about 85% of the outcome)
Electricity cost ($/kWh)
Hardware efficiency (J/TH)
Network difficulty trend
Bitcoin price

Tier 2: Meaningful Adjustments (about 12%)
Uptime percentage
Cooling overhead
Pool fees
Hardware purchase price

Tier 3: Mostly Noise (about 3%)
Transaction fee variance
Minor firmware tweaks
Exotic optimizations people argue about online

Most of what miners obsess over barely moves the needle. The first four variables decide almost everything.

How the Model Was Built

I structured the model like a financial system, not a hobby spreadsheet.

Data sources included:
Hardware specs from manufacturers, cross-checked with asicprofit.com
Electricity rates from utilities and hosting providers
Difficulty projections from historical regression (2020–2025)
Price scenarios based on long-term trend models and volatility bands

The workbook had:
A hardware database (47 ASIC models)
A location matrix (23 electricity regions)
A difficulty engine with three growth paths
Three Bitcoin price paths
Per-machine sheets generating nine scenario combinations each
NPV, IRR, and break-even calculators
A Monte Carlo simulator (1,000 runs)

For the theory behind mining economics and difficulty mechanics, I leaned on educational material from btcfq.com.

What the Numbers Changed My Mind About

1. Efficiency Has Crossed a Critical Threshold

The gap between 29 J/TH and 17 J/TH machines is not just “a bit better.” At scale and over time, it’s survival versus shutdown.

A few J/TH differences can mean thousands of dollars per unit over three years. Efficiency now directly determines who survives bear markets.

2. Electricity Cost Is a Cliff, Not a Slope

I expected a smooth decline in profit as electricity got more expensive. Instead, I found a sharp break point.

Around $0.12–$0.14 per kWh, profitability falls off a cliff. Above that, you’re often mining for the power company, not yourself.

3. Uptime Quietly Destroys Returns

Most calculators assume 100% uptime. Real life does not.

The difference between 98% and 88% uptime on one machine over a year can be over $1,000 in lost revenue.

That’s why I model professional hosting with documented uptime. Providers like OneMiners publish uptime targets around 98%, which I use as my “serious operation” baseline. A home garage setup rarely matches that once you count reboots, overheating, and random outages.

Scenario Analysis: Bull, Base, and Bear

I ran nine combined scenarios using three price paths and three difficulty growth paths over 36 months.

In the base case (moderate price growth, moderate difficulty growth), a modern efficient machine at competitive power rates produced a strong positive NPV and a high double-digit IRR.

In bull scenarios, returns were extreme.
In bear scenarios with fast difficulty growth, losses were very possible.

The key insight: mining is not “always profitable,” but under reasonable assumptions, the probability of a positive outcome was much higher than I expected.

Geography: Where You Mine Matters More Than You Think

I compared 23 locations, including home mining and professional hosting.

At typical U.S. residential rates, home mining often struggles unless you have unusually cheap power. In high-cost regions, it’s simply unworkable.

At scale, professional hosting in low-cost states consistently outperformed home setups once I accounted for:

• My own time

• Infrastructure

• Cooling

• Risk and downtime

For larger deployments, B2B hosting providers such as Circlehash became more attractive past roughly 15 machines, where infrastructure and bulk pricing offset management fees.

Financing vs Paying Cash

One surprise: financing sometimes improved IRR even when the total cost was higher.

By spreading payments over time, capital stayed free for other uses. When I discounted cash flows properly, certain “Pay Later” structures from providers like OneMiners produced better internal rates of return than paying everything upfront.

Old accountant instincts said, “Avoid installments.” The model said, “it depends on your cost of capital.” The model won.

Sensitivity Analysis: What Breaks Profitability

When I stress-tested variables one by one, electricity cost and hardware efficiency dominated.

Price mattered a lot, but efficiency determined how long a machine stayed above break-even as difficulty climbed. In downturns, inefficient machines died first. Efficient ones kept breathing.

Conclusion: Mining Is a Game of Margins, Not Hype

After 47 hours of modeling, thousands of formulas, and more scenario testing than I ever want to see again, one thing became clear: Bitcoin mining is not magic, and it’s not madness. It’s math.

Profitability doesn’t come from chasing small tweaks or obsessing over minor settings. It comes from getting a few major decisions right — especially electricity cost, hardware efficiency, and operational reliability. These aren’t “nice-to-have” optimizations. They are the foundation that determines whether a mining operation thrives, struggles, or shuts down.

The model showed that mining behaves less like a lottery ticket and more like a high-volatility infrastructure investment. There is real risk. Bear markets, rising difficulty, and poor operational setups can absolutely lead to losses. But with efficient hardware, competitive power rates, and strong uptime, the odds shift dramatically. Under realistic assumptions, the probability of long-term profitability was much higher than the common narrative suggests.

The biggest mindset shift was understanding the difference between profit and survival. Cheap electricity maximizes upside in good times. High efficiency protects you in bad times. The miners who last through multiple cycles are not the ones who got lucky once — they are the ones who built operations that can stay above break-even when conditions get tough.

In the end, the answer to “Is mining worth it?” is still “it depends,” — but now that answer has structure behind it. It depends on your power cost. It depends on your hardware. It depends on how professionally you run the operation. When those pieces are aligned, mining stops being a speculative gamble and starts looking like a calculated, long-term strategy.

So yes, after all the charts, simulations, and sleepless spreadsheet nights, I would mine.

Just not blindly.