Ahlers Loft

PWDU now has its own calibrated model in the AI ETA predictor.

Not a shared regional model anymore. Their own. Thank you for sharing your data with us.

Every PWDU result shared for 2024, 2025 and 2026, each winning bird matched to its loft, and 130 races replayed on the real archived app model — wind and weather sampled along the actual line of each race. That fit is now the PWDU model in the app.

Pick PWDU, enter your loft GPS and liberation point, and it pulls the wind along your flight path.

pigeonraceanalyser.co.za

To measure is to know...

Two races. One truck. Two different numbers.

TPA sent Gariep 1 and Gariep 2 off the same truck this morning, half an hour apart — 07:00am and 07:30am. Same line, same wind map. So why two calls?

Because half an hour matters. The model reads the wind at the times the birds are actually along the route, not at the moment we press the button. Move the release by thirty minutes and the birds meet a different morning.

Gariep 1 — 1 199 m/min
Gariep 2 — 1 187 m/min
Band ±38. Roughly 592 km on the truck's own GPS.

Both were pinned this morning before Gariep 2 went up, and both go on our public record — hit or miss, beside every call we've made. Eleven races scored so far: typical miss 23 m/min, best 0.4, worst 69. We publish the worst one too.

Think you can read it closer? Drop your two numbers in the comments before the birds are home. We'll hold them up against the official result.

🕊 pigeonraceanalyser.co.za — free, no sign-up
To measure is to know

RACE 4 AFRIKAPRO — our call, on record before liberation.

AFRIKAPRO published their liberation town ahead of the race. That’s rare, and it’s good for us.

So here is Race 4, called the evening before liberation:

Koppies → Bela-Bela · 280 km · liberation 06:45am

🏁 Predicted winning velocity: 1409 m/min
🕙 First birds: around 10:05am — window 09:55am to 10:20am

Our read of tomorrow morning matches theirs: an easterly early, easing and swinging north-easterly — a right-shoulder wind with a slight headwind in it. Nothing severe. The birds should be home well before the heat of the day.

We’ll post the actual result here tomorrow. Whether we called it right, or missed.

pigeonraceanalyser.co.za

Graaff-Reinet → Pretoria. 782–833 km, liberated at 07:20am this morning. The longest race of our season so far.

Our model reads a 1 259 m/min velocity — which puts the leader home at 17:54. Sunset is at 17:51.

Three minutes. That gap is smaller than our own typical miss on a race this long, so we are not calling it either way.

Will the leaders make it home before dark?

Our second race today, Victoria West, we did call at 1 305 m/min, home at about 17:19 in Pretoria, comfortably in the daylight.

pigeonraceanalyser.co.za

We can't tell you when your first pigeon will hit your landing board. Nobody can...

What we can do, is read the wind and the heat along the route, and tell you roughly when the first birds are due.

To measure is to know...

pigeonraceanalyser.co.za

We called the Richmond winner to within eleven seconds! 767km... 11 seconds. Let that sink in.
We published a winning velocity of 1 359 m/min at 07:20 this morning — fifteen minutes before the birds were let go at 07:35.

The winner flew a 1 359.4m/min Just one beat the prediction.

That's 0.4 m/min out... Three hundredths of one percent. Put another way: over a flight of nine and a half hours, our number was about eleven seconds off the bird. Of 363 pigeons in that race, exactly one beat our call — the winner himself.

Now the other half, because we publish both.

Noupoort, the youngbird race the same morning: we called 1 398, the winner flew 1 467. Sixty-nine out, 4.7%, and 57 birds beat our number. Noupoort was liberated at 07:15 — five minutes before we pinned our call.

Same morning, same model, same forecast.

Ten calls on the record. Average miss 35 m/min. Every one under the winner. All of it on one page:

https://pigeonraceanalyser.co.za/model-vs-reality/

The two TPA races are away this morning — Noupoort with the youngbirds at 07:15am (~680 km) and Richmond open at 07:35am (~760 km).

Our calls, both pinned at 07:20am:

Noupoort — 1 398 m/min
Richmond — 1 359 m/min
Both ± 38 m/min

Eight races scored so far, average miss about 36 m/min, and every miss is on the page next to the hits.

Think we're too fast, or too slow? Put your number in the comments — we'll hold it against the real result.

https://pigeonraceanalyser.co.za/this-week/

Called before the baskets even opened.

Saturday's two TPA races, both races predicted at 07:00 — before a single bird was in the air:

Middelburg (JO): we called a 1192 m/min velocity. The winner flew a 1217m/min. Only 6 pigeons of the 547 beat our number.

Britstown (Open): we called a 1183 m/min winner. The winner flew a 1233m/min. Only 31 pigeons of the 431 beat our number.

See every call we've ever made, scored against the official result: pigeonraceanalyser.co.za/model-vs-reality

Free to use. "To measure is to know".

OFFICIAL CALL — TPA, Saturday 8 August 2026

Middelburg (Youngbirds): 1 192 m/min
Britstown (Open): 1 183 m/min

Both froze at 07:00 this morning, before either race was liberated. Middelburg went up at 07:30, Britstown at 07:45.

Two races at practically the same distance — about ~ 710 km on the TPA average — but on different lines. Middelburg comes down from the Eastern Cape, Britstown from well out west.

Both calls, the live wind maps and the predicted flight lines:
https://pigeonraceanalyser.co.za/this-week/

Free. No sign-up. Your data stays on your own device.