InformNapalm international intelligence community has produced five reports with weekly comparisons of data on incidents of shelling by the Russian occupation forces provided in reports of the JFO (Joint Forces Operation) and the OSCE SMM (Special Monitoring Mission). These reports are available here: Week 1 , Week 2 , Week 3 , Week 4 , Week 5. 
From the very beginning, skeptics kept saying something along the line of “Do you know that the JFO and the OSCE have different counting methodologies?”
Yes, we are aware of the counting methodology applied by the JFO and how it is different from the OSCE’s methodology. Of course, we make calculations in two stages and do not resort to the primitive one-to-one comparisons of the reports. Before comparing data from the two reports, the OSCE shelling incident data are restated on the basis of the JFO principles—30 or more discharges are counted (the way the JFO does it) as a single shelling incident if it has taken more than one hour till the next shell burst.
How are we doing this? Our volunteers print out OSCE tables, use date markers, and look at each settlement in the government-controlled area attacked by shelling from the enemy side. If there is one hour of difference between the shelling incidents, they are entered into the table at the relevant dates. We know that the same date and place can appear in three different OSCE reports, but we only count them once in our table. In their reports, OSCE indicate illumination flares as infractions, but we do not count them as shelling incidents, in spite of listing them in the table.
The following is a practical example of the calculations.
We print out OSCE reports for a week and use colored markers to highlight all violations of the ceasefire regime where enemy shells the government-controlled areas. This is, for instance, a Ceasefire Regime Violation Table as at 7:00 p.m. on August 9, 2019:
A camera in Berezove recorded an explosion of undetermined origin on 1:08 a.m. on August 9. We count it as one violation—1 shelling incident. It may have been a landmine detonation. It is, however, much likelier that it has been a shell hit because the militants tend to fire shells during the nighttime. Thus, this incident is entered into the comparison table.
A camera in Chermalyk recorded shells and explosions from 11:05 p.m. on August 8 to 3.28 a.m. We highlight August 8 violations with a yellow marker—14 shells from 11:05 till 11.13 p.m. as a single shelling incident. We highlight August 9 violations with a blue marker—78 shells from 2:07 till 3:08 a.m. as a single shelling incident.
In Maiorsk, a camera recorded three shells between 2:50 and 3:28 a.m. on August 8—we report it as a single shelling incident. On August 9, four shells were recorded on 0:58 a.m. After more than one hour, further ten shells were recorded from 3:31 to 3:33 a.m., and we report these attacks as two shelling incidents.
This is what it looks like in the aggregate table  we attach to each comparative report.
Thus, whenever we find that the JFO understates the number of shelling attacks on the government-controlled territory by 30 percent in comparison with the OSCE data, the differences in counting methodologies have already been reconciled. It is for this reason that we believe that the understatement of the number of shelling attacks from the enemy side can be caused by an initiative of officials guided by political considerations. There may have been no direct demand for understating the number of shelling incidents; however, there could have been statements along the lines of “Let us demonstrate that the truce is possible.” It would suffice for some officials to make necessary conclusions and take steps on their own. It may have been for this reason that the difference has gone down to 30 percent in the weeks after the reports about the 80 percent downward bias in JFO reports of shelling incidents in comparison with the OSCE reports. However, a difference of 30 and more percent determined on the basis of the unified counting methodology looks like an anomaly taking into account that the OSCE does not cover the entire line of separation and obtains fewer shelling incident reports than the JFO headquarters.
This situation drew attention of InformNapalm’s volunteers who keep preparing weekly calculations.
Distribution and reprint with reference to the source are welcome! (Creative Commons — Attribution 4.0 International — CC BY 4.0 ) Follow InformNapalm on
Facebook  / Twitter  / Telegram Contact us and get updates about new investigations. Repost and share with friends!
You can also support InformNapalm with your donation  so that you help the volunteers to cover the expenses on the site maintenance, and to promote our volunteer movement.