Most “solo vs pool” discussions get stuck on expected value. That’s fine if you’re a fund. If you’re a person with a single miner, the real question is variance: what are the odds you get nothing, and how often does solo end worse than pool over your horizon?
This post gives a simple decision framework using three numbers: P(0 blocks), P(loss), and regret probability (P(solo underperforms pool)).
If your expected value is positive but P(0 blocks) is high, solo mining behaves like a lottery: most outcomes are “no blocks,” plus rare large wins. Pools turn that into something smoother.
A practical way to think about it
If you cannot tolerate long dry spells, you care more about distribution than EV. Pools are basically “variance insurance” (for a fee).
If this is high, solo is a lottery profile. Your median outcome will look like electricity burn (and maybe hardware depreciation if you count it).
This includes the case where you hit zero blocks (often the dominant loss outcome), plus cases where you get a block but still lose money due to cost inputs.
This is the decision signal: the probability that solo ends the horizon with less net USD than pool using the same inputs.
In the Compare page, you’ll see:
These are three “anchor” scenarios I’ll reference elsewhere (e.g., Reddit replies and future posts). Each one links to the Compare form with inputs prefilled and set to autorun. After it computes, click Copy share link. I don’t hardcode tokens here because the network snapshot updates over time. The share link you generate freezes the snapshot inside the token, so people can compare the same assumptions.
{
"coin": "bch",
"hashrate": "9.4TH",
"horizon_days": 365,
"coin_price_usd": 450,
"electricity_cost_per_kwh": 0.09,
"asic_power_watts": 200,
"pool_fee_pct": 0.01
}{
"coin": "bch",
"hashrate": "9.4TH",
"horizon_days": 365,
"coin_price_usd": 100,
"electricity_cost_per_kwh": 0.09,
"asic_power_watts": 200,
"pool_fee_pct": 0.01
}{
"coin": "bch",
"hashrate": "9.4TH",
"horizon_days": 365,
"coin_price_usd": 450,
"electricity_cost_per_kwh": 0.20,
"asic_power_watts": 200,
"pool_fee_pct": 0.01
}This tool models solo blocks as a Poisson process over the chosen horizon. That gives you a distribution over block counts (0, 1, 2, …). Pool is treated as expected value (deterministic) for now.
The network snapshot is frozen into the share token so results are stable and comparable when shared. Details are documented on the Methods page.
Solo mining makes sense when you explicitly want a lottery-shaped distribution and can tolerate long dry spells. If you need stable outcomes, you’re buying the wrong thing.