Ground footage forecast
Where footage will likely be geolocated tomorrow and the following days, based on the last 4 weeks. With reliability measured on past days, filmed units and posting accounts.
Situation note
- Last 7 days: 3 geolocated images, about 3 once late geolocations arrive; down 21% on the previous 7 days.
- Tomorrow (Fri, Nov 6, 2026): about 1 images expected, between 0 and 3 on an ordinary day.
- Most likely sectors within 7 days: Alfredo V. Bonfil (Mexico) (21 %), Cedro (Mexico) (21 %), Juan Aldama (Mexico) (21 %).
- Most filmed units this week: MF (2), Cabreras - MF (1).
- Footage comes first from “X account (OSINT, individual)” accounts (67 % of main sources), ahead of “News outlet” (33 %).
- Reliability measured over 24 past days: 1 of the 10 sectors announced were filmed within 7 days, against 0 for the “same as yesterday” rule. 86 % of images fall in a sector with no image in the previous 4 weeks: nothing can announce those.
This forecasts footage, not fighting: what GeoConfirmed volunteers will geolocate, based on what they already have. A sector without network or witnesses stays silent, and no past figure announces a new offensive.
images expected tomorrow
1
from 0 to 3 on an ordinary day
sectors above 50%
0
within 3 days
measured reliability
0.3 / 10
announced sectors filmed; “same as yesterday”: 0
unpredictable share
86 %
images in a sector with none in the previous 4 weeks
01Map of likely sectors
unlikelyvery likely
One circle per 8 km sector, redder when an image is more likely within 3 days. Click a circle for details.
02Most likely sectors
| Sector | Tomorrow | 3 d | 7 d | Trend |
|---|---|---|---|---|
| Alfredo V. Bonfil (Mexico) Sinaloa · filmed 1 d out of 28, last Oct 28 · Chapo Isidro | 1 % | 5 % | 21 % | falling |
| Cedro (Mexico) Dgo. · filmed 1 d out of 28, last Nov 2 · MF | 1 % | 5 % | 21 % | new |
| Juan Aldama (Mexico) Chihuahua · filmed 1 d out of 28, last Nov 2 · Cabreras - MF | 1 % | 5 % | 21 % | new |
| La Tuna (Mexico) Sin. · filmed 1 d out of 28, last Oct 23 | 1 % | 5 % | 21 % | falling |
| Las Pilitas (Mexico) Chihuahua · filmed 1 d out of 28, last Oct 21 | 1 % | 5 % | 21 % | falling |
| Rosamorada (Mexico) Nayarit · filmed 1 d out of 28, last Oct 27 | 1 % | 5 % | 21 % | falling |
| Siqueros (Mexico) Sin. · filmed 1 d out of 28, last Oct 24 · MF | 1 % | 5 % | 21 % | falling |
| Zavala (Mexico) Sin. · filmed 1 d out of 28, last Nov 1 · MF | 1 % | 5 % | 21 % | new |
| Queréndaro (Mexico) Mich. · filmed 1 d out of 28, last Oct 18 · CJNG | 1 % | 5 % | 21 % | falling |
| Villanueva (Mexico) Zacatecas · filmed 1 d out of 28, last Oct 18 · MF | 1 % | 5 % | 21 % | falling |
| San Juan (Mexico) Sin. · filmed 1 d out of 28, last Oct 14 · H12/Cabreras - MF | 0 % | 5 % | 21 % | falling |
| Puerta de San Marcos (Mexico) Mexico · filmed 1 d out of 28, last Oct 11 · Mayito Flaco | 0 % | 5 % | 21 % | falling |
| Zirándaro de los Chávez (Mexico) Guerrero · filmed 1 d out of 28, last Oct 11 · Gente del Serio - CJNG | 0 % | 5 % | 21 % | falling |
| La Mesa del Cerro Azul (Mexico) Gro. · filmed 1 d out of 28, last Oct 10 · La Familia Michoacana | 0 % | 4 % | 21 % | falling |
| La Sábila (Mexico) Sinaloa · filmed 1 d out of 28, last Oct 9 · MF | 0 % | 4 % | 21 % | falling |
03Regions: images expected tomorrow
| Region | Expected | Ordinary day | Last 7 days | 3 weeks before, per week |
|---|---|---|---|---|
| Sin. | 0.2 | 0 → 2 | 1 | 1 |
| Chihuahua | 0.1 | 0 → 2 | 1 | 0 |
| Dgo. | 0.1 | 0 → 0 | 1 | 0 |
| Sinaloa | 0.1 | 0 → 0 | 0 | 1 |
| Nayarit | 0.0 | 0 → 0 | 0 | 0 |
| Mich. | 0.0 | 0 → 0 | 0 | 0 |
| Zacatecas | 0.0 | 0 → 0 | 0 | 0 |
| Guerrero | 0.0 | 0 → 0 | 0 | 0 |
| Mexico | 0.0 | 0 → 0 | 0 | 0 |
| Gro. | 0.0 | 0 → 0 | 0 | 0 |
“Ordinary day”: 10th to 90th percentile of the last 4 weeks' days. On replayed days, the actual figure of the main regions fell within this range 98 times out of 100.
04Measured reliability
The forecast was replayed for each day from Sep 5 to Oct 22, with only what was known the evening before. Probability recalibration is learnt on the first half (Sep 5 → Sep 28); figures below come from the second (Sep 29 → Oct 22, 24 days), unseen by the method.
Out of 10 announced sectors, how many filmed
Announced probability and observed frequency · 3 days
| Announced | Observed | Sector-days |
|---|---|---|
| 1 % | 3 % | 234 |
An honest forecast says 30% where about 30% is observed.
05Units filmed this week
| MFOther · Dgo. | 2 |
| Cabreras - MFOther · Chihuahua | 1 |
3 images out of 3 name a unit this week.
06Who posts this week
| News outlet | 33 % |
| X account (OSINT, individual) | 67 % |
| X @Toluco_R3 · Soy FrizX account (OSINT, individual) · Mayo Zambada CDS | 2 |
| X @All_Source_News · All Source NewsNews outlet · Mayo Zambada CDS | 1 |
Share of images by type of the account that posted them first. Type inferred from the account name: an indication, not a check. 30-day detail →