Orphans are a propagation problem, not a hashrate problem. We measured how much faster blocks reach a connected miner than they do over public gossip, and modeled what that is worth. The answer depends heavily on where you are.
Based on our measured delivery advantage and the network's observed orphan rate. These are projections from measurement, not amounts any miner has been observed to save — see What we have not shown below.
We arrive 306 ms ahead of public gossip on full blocks here, which models to avoiding 69% of a miner's orphans.
| Your share | Blocks/day | Orphans/day now | Avoided/day | ZEC/year |
|---|---|---|---|---|
| 5% | 58 | 0.34 | 0.24 | 107 |
| 10% | 115 | 0.68 | 0.47 | 214 |
| 20% | 230 | 1.36 | 0.94 | 429 |
| 44% | 507 | 3.00 | 2.07 | 944 |
We arrive 42 ms ahead here, which models to avoiding 9% of a miner's orphans. Europe already has excellent peering, so there is far less for us to recover — we would rather you knew that now than after connecting.
| Your share | Blocks/day | Orphans/day now | Avoided/day | ZEC/year |
|---|---|---|---|---|
| 5% | 58 | 0.34 | 0.03 | 15 |
| 10% | 115 | 0.68 | 0.06 | 29 |
| 20% | 230 | 1.36 | 0.13 | 59 |
| 44% | 507 | 3.00 | 0.28 | 130 |
Stated in ZEC only. We do not publish a fiat figure: it would be a number we do not control, and it would be wrong by the time you read it.
When someone finds a block, it takes time to reach you. Until it does, your machines keep working on the previous one. If you find a block during that window, you have solved the wrong parent — the network keeps the other chain and your block earns nothing.
That window is the whole game. It is not about how much hashrate you have; a larger miner simply loses proportionally more to it. Shorten the window and you keep blocks you are currently forfeiting.
Our delivery cost barely changes with block size. Public gossip's does — and full blocks are exactly when the race is worth winning.
| Region | Sovright delivery | Public gossip | Difference |
|---|---|---|---|
| APAC | +0.110 ms/KB | +1.739 ms/KB | 16× |
| Europe | +0.145 ms/KB | +0.646 ms/KB | 4.5× |
Cost added per kilobyte of block. A near-flat line means a big block reaches you almost as fast as an empty one; gossip slows down as blocks fill.
| Median delivery | <5KB | 5–20KB | 20–100KB | 100KB+ |
|---|---|---|---|---|
| APAC — Sovright | 119ms | 119ms | 122ms | 134ms |
| APAC — public gossip | 183ms | 275ms | 313ms | 440ms |
| Europe — Sovright | 67ms | 69ms | 75ms | 89ms |
| Europe — public gossip | 82ms | 107ms | 130ms | 131ms |
For every block we take the earliest arrival anywhere in our fleet as the origin, then measure how long it took to reach a given region by our path versus that region's own public-gossip connection. Slopes are Theil-Sen over ~1,900 blocks per region, with the outer 1% of delivery times trimmed, so a handful of stragglers cannot drag the line.
The modeled saving assumes orphan exposure is bounded by how long a miner stays unaware of the new tip: at a 75-second target block time, hearing blocks Δ earlier removes roughly Δ/75s of exposure. Applied to 1,152 blocks/day and a 1.25 ZEC block subsidy.
The model is calibrated before it is used. At the measured 0.591% orphan rate it predicts about 6.8 orphaned blocks per day across the whole network. Our orphan monitor independently observes about 6 per day. We publish that check because a projection nobody tested against reality is just a sales figure.
We measured that blocks reach us sooner. We have not yet measured a miner's orphan rate falling after connecting to us.
The reasoning in between — faster delivery means an earlier switch to the new tip, which means fewer wasted blocks — is sound, and it is still reasoning. Nobody has run the before-and-after. Every ZEC figure above inherits that gap, and we would rather you weigh it than discover it.
Run a pilot. We measure your orphan rate before you connect and after, and you keep the result either way — including if it shows we did not help you much. That is the only way anyone turns the projection above into a fact, and we would rather have one real number than a persuasive model.
If you mine from APAC, you are the case where our measurements say there is the most to recover.