Receipts
The accuracy program
Monthly report card
Every receipt of the month added up: how far each day's published call was from what the NOAA station paired with it measured, for wind, seas and period, by fish and by station. A receipt grades the forecast, not the fishing.
Days graded
10 so far
of 31 in October, month not over
Receipts
73
calls with a buoy reading
Wind error
2.7 kt
mean absolute, n 70, bias 0 kt
Seas error
0.3 ft
mean absolute, n 22, bias -0.1 ft
History rebuilt 8 h ago; 11 days kept.
October 2026
11 days so far · Oct 5 to Oct 1510 days graded so far of 31. The card is complete when the month ends; with this few days every average below can still move a lot.
Wind
2.7 kt
mean absolute, n 70, bias 0 kt
Seas
0.3 ft
mean absolute, n 22, bias -0.1 ft
Period
1.6 s
mean absolute, n 22, bias +2 s
| Fish | Receipts | Wind error | Seas error |
|---|---|---|---|
| Bonefish | 5 | 1.9 kt (n 5) | -- |
| Grouper | 3 | 2.9 kt (n 3) | 0.3 ft (n 1) |
| Kingfish | 2 | 4.2 kt (n 2) | 0.2 ft (n 1) |
| Mahi | 1 | 1.4 kt (n 1) | -- |
| Mutton Snapper | 8 | 3.5 kt (n 7) | 0.2 ft (n 7) |
| Permit | 7 | 1.9 kt (n 7) | -- |
| Redfish | 10 | 2.1 kt (n 10) | 0.3 ft (n 2) |
| Sailfish | 3 | 3.0 kt (n 3) | 0.6 ft (n 2) |
| Snook | 8 | 3.0 kt (n 8) | -- |
| Swordfish | 7 | 3.2 kt (n 7) | -- |
| Tarpon | 8 | 1.2 kt (n 7) | 0.4 ft (n 3) |
| Wahoo | 1 | 0.3 kt (n 1) | -- |
| Yellowtail | 10 | 4.0 kt (n 9) | 0.2 ft (n 6) |
| Station | Receipts | Wind error |
|---|---|---|
| Key West (KYWF1) | 8 | 2.9 kt (n 8) |
| Lake Worth Pier (LKWF1) | 9 | 4.4 kt (n 9) |
| Long Key (LONF1) | 25 | 2.2 kt (n 25) |
| Virginia Key (VAKF1) | 30 | 2.5 kt (n 28) |
Worst wind miss: Yellowtail at Charted structure, 9.9 nm NE of Angelfish Creek on Oct 13, called 11 kt against 3 kt measured at Virginia Key (VAKF1) (error +8 kt).
How the card is made
- Only what was graded. Each row is a receipt from the Receipts page: the call of record for one fish on one day, against the station within reach of its place. Days with no buoy reading add no row.
- Mean absolute error and bias. The error is ours minus the station's; the bias is its average, so a positive bias means we called it higher than it was.
- No projection. A month that is not over shows the days it has so far, and every average on it can still move.
Observations: NOAA National Data Buoy Center, read through the Anglrithm relay. Forecast: the published call of record (weather data by Open-Meteo.com, CC BY 4.0). Summed on the device from public/data/receipts/history.json.