The Russian-Ukrainian War Thread

[Pukovnik7]

Maybe they are of better quality, but they lack the editorial oversight of actual media publications. It's one thing if you can show your work demonstrably, in a manner that is easy to understand. But a project like Oryx's is practically screaming for multiple independent reviews, and for some reason media and bloggers cite their figures without even the slightest hesitation.

Modern media, especially online newspapers, have minimal editorial oversight IIRC. Can't say for certain as I never wrote for printed newspapers, but in the online newspapers, "editorial oversight" tends to be zilch. Pressure to publish first is simply too much.

 

[vikingatespam]

Somehow this is always the response, and yet no one does it (or if they do it is never posted publicly). I lack the time, and the expertise to do it properly. You would think someone who has both would try it, but instead all you get when you bring up the lack of oversight are calls to do it yourself.

There is nothing wrong with comparing a few random images in a category (for example, T-80 tanks). So far as I can tell the other armor loss trackers (lostarmour, etc) were tracking with Oryx the last time I checked.

Oryx also lists its methodologies, which is rather important.

 

[vikingatespam]

Maybe they are of better quality, but they lack the editorial oversight of actual media publications. It's one thing if you can show your work demonstrably, in a manner that is easy to understand. But a project like Oryx's is practically screaming for multiple independent reviews, and for some reason media and bloggers cite their figures without even the slightest hesitation.



So I'm supposed to trust Oryx because they audit their own work (to an unknown extent)? I realize they have been around awhile. I'm not saying they are up to anything nefarious. I'm saying it's weird/frustrating that the best and maybe only source for equipment loss statistics, one that is cited by practically every media publication in the West, is an independent blog with no oversight, and apparently nobody with credentials cares to check their work.

Imagine if we applied these standards to anything that actually mattered.

While I cannot speak for the others, these sum of these issues is why Oryx is trusted:

- it has no relationship to either the RU or UKR governments (both of those official sources are garbage)
- it existed years before the RU invasion
- it lists its methodologies
- it has pictures of each and every item it listed
- it tracks well with other loss trackers (although it has been a while since I cross checked)

One wonders how they would generate pictures of each item loss if they were not real - attempts to slip Photoshopped items into the list have failed in the past.

Oryx is treated as the best OSINT source of information in these matters, due to the above reasons. May I suggest taking a few items from a list, comparing them, and then running them through a digital forensics site, such as FotoForensics.

 

[Feanor]

While I cannot speak for the others, these sum of these issues is why Oryx is trusted:

- it has no relationship to either the RU or UKR governments (both of those official sources are garbage)
- it existed years before the RU invasion
- it lists its methodologies
- it has pictures of each and every item it listed
- it tracks well with other loss trackers (although it has been a while since I cross checked)

One wonders how they would generate pictures of each item loss if they were not real - attempts to slip Photoshopped items into the list have failed in the past.

Oryx is treated as the best OSINT source of information in these matters, due to the above reasons. May I suggest taking a few items from a list, comparing them, and then running them through a digital forensics site, such as FotoForensics.

The mistake I recall finding was a BTS- 4 or 5 with a ZU-23-2 on top. It was attributed to Russian forces but I knew for a fact that a pre-war Ukrainian brigade had exactly this type of vehicle in service including the modified superstructure on the hull (photos of them had come up from exercises a few months before the war). This is how a mistake can creep in.

On a side note with sources liks Oryx, lostarmour, Ledok, the main issue is when you have something like this;

Рубрика от Ледка

Т-64БВ - уничтожен Северо-восточнее с. Работино в Запорожской области. Геопривязка LA: (47.475917, 35.901304) 64-ка ВСУ была подбита бойцами 58-й ОА ВС РФ. Оригинал неизвестен

t.me

It's a destroyed T-64, like BV. Is it Ukrainian? Well... probably. We don't have any info on former LDNR T-64s in Zaporozhye. But make no mistake LDNR forces were spotted in Kherson and Zaporozhye regions previously. They just didn't have T-64s. The source indicates it's a destroyed Ukrainian T-64 near Rabotino. Here likelihood is high based on location but not based on vehicle ID.

Now consider this;

Рубрика от Ледка

БМП-1 О/з ▲6 - уничтожена Возле с. Работино в Запорожской области. Кем и как была подбита БМП ВСУ неизвестно… Датировано: 26.07.2023. Оригинал ; ID LostArmour

t.me

It's a BMP-1 destroyed, also allegedly Ukrainian near Rabotino. But could it be a Russian BMP-1 near Rabotino? Theoretically yes. Two things cut against that. One is the fact that we haven't spotted any Russian units riding BMP-1s in that area. The other is the tactical marking on the side. It matches tactical markings from that Ukrainian mech coy riding BMP-1s that got shredded in one of the attacks on Rabotino. This is likely a vehicle from either that coy or another one in their btln.

Now consider this;

Рубрика от Ледка

БМП-2 - уничтожена Где-то в зоне СВО. Была подбита летом 2023 года, обстоятельства неизвестны. Источник

t.me

The source lists it as a Ukrainian BMP-2 destroyed some time in the summer of 2023, location unknown. How do we know it's Ukrainian? We don't. There are entries Oryx has that look like this. Presumably, and based on his prior comments, the source that provided it to him ID-d it as Russian or Ukrainian. However they could be lying, they could be wrong, they could be guessing and passing it off as a fact.

Or consider this;

Рубрика от Ледка

Т-80БВ - уничтожен Севернее с. Новомайорское в Донецкой области. Геопривязка LA: (47.757139, 37.022167) Танк подбит бойцами 40-й гв. ОБрМП. Датировано: 17.09.2023. Источник

t.me

It's a T-80BV. Ukraine had near 0 in service pre-war (a few coys in their Marine units). However they captured a few from Russia. Nonetheless most T-80BVs on the battlefield are Russian. Ok, but we have the cross on the side right? And what looks like a story, the vehicle got immobilized, was abandoned, and then got finished off by something that blew the turret off. Unless.. is the T-80BV with the cross and the destroyed tank the same one? They look sort of close, and the piece of bush by the track is similar. The rest of the surrounding vegetation got destroyed by the explosion. So... probably.

Now do this for the nearly 20 000 entries in Oryx. What Sinister is asking for is not really doable. I know he's made mistakes, beucase I've caught it. It would be absolutely incredible if he didn't make mistakes. The only time I looked at his stuff systematically iirc it was 1 mistake on ~200 entries, an astounding 99.5% accuracy rate. I'm sure there were other mistakes that just fell into the category above where you can't really tell for sure. However overall he clearly did a very good job assesing the data. He was objective about, and on some level he had to trust his sources to do his work at all. Lastly before this war's losses reached the insane levels he still had comments enabled on his site, and would respond. When I pointed out the mistake I mentioned above, he fixed it. So there was clearly a willingness on his part to correct any errors.

 

[SinisterMinister]

Now do this for the nearly 20 000 entries in Oryx. What Sinister is asking for is not really doable. I know he's made mistakes, beucase I've caught it. It would be absolutely incredible if he didn't make mistakes. The only time I looked at his stuff systematically iirc it was 1 mistake on ~200 entries, an astounding 99.5% accuracy rate. I'm sure there were other mistakes that just fell into the category above where you can't really tell for sure. However overall he clearly did a very good job assesing the data. He was objective about, and on some level he had to trust his sources to do his work at all. Lastly before this war's losses reached the insane levels he still had comments enabled on his site, and would respond. When I pointed out the mistake I mentioned above, he fixed it. So there was clearly a willingness on his part to correct any errors.

This is a reasonable point, and I appreciate the detail of your post. Seems like from everything you said of your experience there is no reason to distrust Oryx's intentions or capability. But a couple things I'd point out:

1. If Oryx (no idea how many people are on their team) is capable of doing this project, it is a shame that more reputable organizations (whether media or government) are not doing parallel projects. From the NATO Discord leaks it seems like they were using Oryx' statistics in their briefings. If the DOD wanted to audit Oryx' work and create their own report they would easily have the resources to do it; apparently they just don't care.

2. I think anyone who blindly cites Oryx' statistics without caveat is doing a disservice to readers. Obviously this is referring more to information tailored to large audiences; posting on a forum is a different story. But everyone should be aware that no project that relies entirely on interpreting often difficult-to-parse photographic evidence can be either 100% accurate or 100% complete in it's scope.

My concern is mainly that the widespread citation of Oryx without context has created an environment where many people following the war (especially the general public) are under the impression that those statistics are rock solid. They are probably the best thing we have, but I would be completely unsurprised if post-war scholarship found them to be significantly off (the most likely culprit being the amount of missing photographic evidence).

 

[vikingatespam]

My concern is mainly that the widespread citation of Oryx without context has created an environment where many people following the war (especially the general public) are under the impression that those statistics are rock solid.

I suspect any informed reader realizes they are not rock solid. After all, visual evidence of losses is biased low if considering total losses (i.e. visual < total). However, outside of having the official RU/UKR numbers, Oryx is likely the most accurate assessment of armor loses available to anyone.

 

[Jaykaro]

This is a reasonable point, and I appreciate the detail of your post. Seems like from everything you said of your experience there is no reason to distrust Oryx's intentions or capability. But a couple things I'd point out:

1. If Oryx (no idea how many people are on their team) is capable of doing this project, it is a shame that more reputable organizations (whether media or government) are not doing parallel projects. From the NATO Discord leaks it seems like they were using Oryx' statistics in their briefings. If the DOD wanted to audit Oryx' work and create their own report they would easily have the resources to do it; apparently they just don't care.

2. I think anyone who blindly cites Oryx' statistics without caveat is doing a disservice to readers. Obviously this is referring more to information tailored to large audiences; posting on a forum is a different story. But everyone should be aware that no project that relies entirely on interpreting often difficult-to-parse photographic evidence can be either 100% accurate or 100% complete in it's scope.

My concern is mainly that the widespread citation of Oryx without context has created an environment where many people following the war (especially the general public) are under the impression that those statistics are rock solid. They are probably the best thing we have, but I would be completely unsurprised if post-war scholarship found them to be significantly off (the most likely culprit being the amount of missing photographic evidence).

As mentioned earlier, you can verify this yourself; otherwise, you would subject any other source to doubt, necessitating further verification.

Two months ago, OSINT Archi Irving took the time to check Oryx for duplicates (yes, all 15,554 records from both sides), and he posted the results.

For Russia, out of 11,487 records, 136 duplicates were found (1.2%). For Ukraine, out of 4,067 records, there were 114 duplicates (2.8%).

In total, out of 15,554 records, there were 2% duplicates if rounded.

When Archi finds duplicates, he sends them to Oryx, and they remove them from the lists. That's why you might sometimes see on the website that there were, for example, 11,490 losses, but it became 11,487.

Two conclusions:

1. Duplicates among Ukrainian losses occur twice as often as among Russian losses. This is related to the "one 'Leopard' or 'Bradley' from 30 different angles" effect from Russian "war correspondents."
2. The overall margin of error at 2% is exceedingly small - the documented Oryx losses are highly reliable.

 

[Bob53]

A claim here that Russians attacking Avdiivka in brigade strength at least, and getting absolutely mollywopped. 90 APCs/IFVs and 30 tanks confirmed lost already. Is there any other confirmation if this?

https://www.reddit.com/r/CombatFootage/comments/175mu2b

 

[Larry_L]

A claim here that Russians attacking Avdiivka in brigade strength at least, and getting absolutely mollywopped. 90 APCs/IFVs and 30 tanks confirmed lost already. Is there any other confirmation if this?

https://www.reddit.com/r/CombatFootage/comments/175mu2b

This is not confirmation. Just another source saying the same thing in more detail. Flavored with some salt he can be informative.

A partial Quote:
"Avdiivka….the last few days it became clear why the Russians inserted their 25th CAA into the sector Svatove-Kremina, and withdrew the 1st GT and the 41st CAA from there, back in September: they’ve re-filled battered units, and then re-deployed them in the Avdiivka sector. Four days ago, these have launched a pincer attack, supported by heavy air strikes (mostly ‘spray & pray’ style, by Su-25s, but few by Ka-52s and Mi-28s), and extensive volumes of artillery fire.

Sufficient to say that as of the morning of 10 October, this counteroffensive ended in a veritable catastrophe. Essentially, at least two, possibly three huge Russian columns were detected (by UAVs) while still some 5,000-10,000 metres behind the frontline. And hit by artillery."

Ukraine War, 11 October 2023

Hello everybody!

xxtomcooperxx.substack.com

 

[vikingatespam]

A number of sources are claiming that the Pavel Derzhavin (RU patrol ship) was destroyed by a blast. One source:

It remains to be confirmed, but if the link is correct, RU naval losses continue to mount.

 

[T.C.P]

A claim here that Russians attacking Avdiivka in brigade strength at least, and getting absolutely mollywopped. 90 APCs/IFVs and 30 tanks confirmed lost already. Is there any other confirmation if this?

https://www.reddit.com/r/CombatFootage/comments/175mu2b

Total confirmed visual losses so far are30 armoured targets destroyed

https://twitter.com/x/status/1712285063453261874

If Ukr sources are correct and Russia has lost over 130 pieces, then this is reaching Vuldehar levels of destruction.

Rybar and Suriyaki maps claim that the high points have been captured

https://twitter.com/x/status/1712239369321394179

As per Deepstate- Russia has some partial successes on this front so far

"A tough battle continues for Avdiivka. The Defense Forces are making efforts to prevent the enemy from gaining strategic success. Each of our mistakes can be costly later, so high-quality forces are used to repel attacks. It's difficult for 110 OMBr, but they manage. 53 OMBr maneuvers well and keeps the enemy at a safe distance. Well done and TrO, and "presidents", and many other units that we cannot mention. 59 OMPBr has good success, the Katsaps lose a lot of equipment, and the fields are littered with the corpses of the "second army of the world""

 

[Feanor]

As mentioned earlier, you can verify this yourself; otherwise, you would subject any other source to doubt, necessitating further verification.

Two months ago, OSINT Archi Irving took the time to check Oryx for duplicates (yes, all 15,554 records from both sides), and he posted the results.

For Russia, out of 11,487 records, 136 duplicates were found (1.2%). For Ukraine, out of 4,067 records, there were 114 duplicates (2.8%).

In total, out of 15,554 records, there were 2% duplicates if rounded.

When Archi finds duplicates, he sends them to Oryx, and they remove them from the lists. That's why you might sometimes see on the website that there were, for example, 11,490 losses, but it became 11,487.

Two conclusions:

1. Duplicates among Ukrainian losses occur twice as often as among Russian losses. This is related to the "one 'Leopard' or 'Bradley' from 30 different angles" effect from Russian "war correspondents."
2. The overall margin of error at 2% is exceedingly small - the documented Oryx losses are highly reliable.

The biggest grey area aren't the duplicates, it's the Soviet-era vehicles that could belong to either side and are too destroyed to PID. A destroyed BMP-1/2, BTR-70/80, T-64/72/80 could belong to either side. There also aren't any BMP-3s, T-72B3s, etc. listed for Ukraine. But we know captured BTR-82s, BMP-3s, T-72B3s, etc. were destroyed in Ukrainian service. Which raises some questions.

A claim here that Russians attacking Avdiivka in brigade strength at least, and getting absolutely mollywopped. 90 APCs/IFVs and 30 tanks confirmed lost already. Is there any other confirmation if this?

https://www.reddit.com/r/CombatFootage/comments/175mu2b

Russian sources are silent on the casualties but are claiming certain advances. I'd like to see some of that visual confirmation. Russian sources have been very scarce with footage. More-so Rozhin commented on how it seems Russia is releasing very little footage from the current offensive.

 

[Feanor]

Update Avdeevka.

Russian sources are reporting a slower advance in Avdeevka. The main assault on Berdychy is apparently halted. Russia redirected some efforts towards Petrvoskoe, and has solidified their control of the refuse mound in the north of the town. Russian forces have also allegedly captured a couple of strong points from the south, the first time the southern attacks towards Avdeevka have gained any ground. The second pincer out of Severnoe seems to have completely halted. It's possible it wasn't a particularly major effort, the ground there is disadvantageous. The halted pincer near Berdychy in a position that doesn't seem to be something Russia can reasonably hold in the long run is at least indirect confirmation that things aren't going as planned. On the other hand Russia could expand the salient northward and southward, and make the positions more tenable. It wouldn't help them capture Avdeevka this time around, but would put them in a better position for the next time they try. The fight doesn't seem to be over yet so we will have to wait and see.

Рыбарь

❗️ Битва за Авдеевку: обстановка к исходу 11 октября 2023 года У Авдеевского укрепрайона российские подразделения продолжают штурм опорных пунктов. После первоначального успеха ВС РФ противник перегруппировался и сейчас пытается сдержать наступление. На северном фланге бойцы 11 полка ВС РФ в...

t.me

https://rybar.ru/piwigo/upload/2023/10/12/20231012000405-358012f9.jpg

Massive amounts of heavy smoke can be seen rising over Avdeevka from Donetsk. Footage of the combat itself remains ominously absent.

Colonelcassad

Противник обстрелял район железнодорожного вокзала в Донецке. ВС РФ с утра продолжили обстрел позиций ВСУ в Авдеевке. Сегодня в течение дня ожидаются интенсивные бои к северу и югу от города. Об оперативном окружении Авдеевки пока говорить преждевременно.

t.me

Colonelcassad

К вечеру продолжает гореть Авдеевский коксохим. Вокруг Авдеевки продолжаются ожесточенные бои. С обеих сторон ощутимые потери. Можно отметить, что это первая за длительная время операция, где командование постаралось создать информационный колпак и дозировано выдает информацию о ходе...

t.me

Colonelcassad

На фоне Донецка продолжает возвышаться огромное черное облако горящего со вчерашнего дня Авдеевского коксохима.

t.me

 

[Pukovnik7]

1. If Oryx (no idea how many people are on their team) is capable of doing this project, it is a shame that more reputable organizations (whether media or government) are not doing parallel projects. From the NATO Discord leaks it seems like they were using Oryx' statistics in their briefings. If the DOD wanted to audit Oryx' work and create their own report they would easily have the resources to do it; apparently they just don't care.

As I have explained, those "more reputable" organizations often have neither interest nor capacity for doing such things. Majority of the modern media outlets don't even write the news themselves, they just receive news from "news dispensers" (news agencies and the like), rewrite the few lines and then publish them with no fact checking at all.

If you want reliability, interested amateurs such as Oryx are your best bet.

 

[Feanor]

A number of sources are claiming that the Pavel Derzhavin (RU patrol ship) was destroyed by a blast. One source:

It remains to be confirmed, but if the link is correct, RU naval losses continue to mount.

Allegedly the Pavel Derzhavin, alive and well. Of course with no hull number of ship name visible, it's hard to be sure. Maybe someone up to speed on minute difference between the 22160 series can confirm.

Черноморский Флот

В условиях СВО наш телеграмм-канал "Черноморский Флот" никогда не публикует свежие фото кораблей ЧФ. Но в свете (очередной, увы) истерики "икспердов", сделаем исключение. Патрульный корабль "Павел Державин" Черноморского флота России у причала. Сегодня. Посеять сомнения, смуту, раздор...

t.me

 

[Redshift]

This is a reasonable point, and I appreciate the detail of your post. Seems like from everything you said of your experience there is no reason to distrust Oryx's intentions or capability. But a couple things I'd point out:

1. If Oryx (no idea how many people are on their team) is capable of doing this project, it is a shame that more reputable organizations (whether media or government) are not doing parallel projects. From the NATO Discord leaks it seems like they were using Oryx' statistics in their briefings. If the DOD wanted to audit Oryx' work and create their own report they would easily have the resources to do it; apparently they just don't care.

2. I think anyone who blindly cites Oryx' statistics without caveat is doing a disservice to readers. Obviously this is referring more to information tailored to large audiences; posting on a forum is a different story. But everyone should be aware that no project that relies entirely on interpreting often difficult-to-parse photographic evidence can be either 100% accurate or 100% complete in it's scope.

My concern is mainly that the widespread citation of Oryx without context has created an environment where many people following the war (especially the general public) are under the impression that those statistics are rock solid. They are probably the best thing we have, but I would be completely unsurprised if post-war scholarship found them to be significantly off (the most likely culprit being the amount of missing photographic evidence).

You are basically trying to undermine oryx withoit presenting any evidence.

You can read what Feanor is saying, that he personally has checked a number of sources and found very few errors, this is what you originally asked for.

You appear exasperated that no-one is doing something that you believe should be done, so my suggestion would be that, rather than use this lack of cross checking to suggest unreliability is that you a) acknowledge that Feanor has done some of what you wished to see and b) do it yourself.

 

[SinisterMinister]

You are basically trying to undermine oryx withoit presenting any evidence.

You can read what Feanor is saying, that he personally has checked a number of sources and found very few errors, this is what you originally asked for.

You appear exasperated that no-one is doing something that you believe should be done, so my suggestion would be that, rather than use this lack of cross checking to suggest unreliability is that you a) acknowledge that Feanor has done some of what you wished to see and b) do it yourself.

How have I tried to undermine Oryx? I am pointing out the limitations of their project, and the recklessness with which people use their numbers. I did not ask for someone on this forum to audit their work, though I can appreciate the work Feanor had done in this regard. What would be nice is an effort by multiple creditable sources to audit the entire body of work and publish their findings; this is done in every area of scientific research, not sure why it is such a provocative ask even if I know it will not be done.

You seem very agitated by this, not sure why. I'm surprised / annoyed that everyone of substance who actually has an interest in using the numbers does not have a better method than just quoting a blogger without caveat, regardless of how thorough we think that blogger is. I think that is a reasonable position.

I would apply the same standards to many other methods of statistical analysis. Nobody would have been ok with the media using a blog's investigative work to establish COVID-19 fatalities, for example.

 

[vikingatespam]

Update Avdeevka.

Russian sources are reporting a slower advance in Avdeevka. The main assault on Berdychy is apparently halted. Russia redirected some efforts towards Petrvoskoe, and has solidified their control of the refuse mound in the north of the town. Russian forces have also allegedly captured a couple of strong points from the south, the first time the southern attacks towards Avdeevka have gained any ground. The second pincer out of Severnoe seems to have completely halted. It's possible it wasn't a particularly major effort, the ground there is disadvantageous. The halted pincer near Berdychy in a position that doesn't seem to be something Russia can reasonably hold in the long run is at least indirect confirmation that things aren't going as planned. On the other hand Russia could expand the salient northward and southward, and make the positions more tenable. It wouldn't help them capture Avdeevka this time around, but would put them in a better position for the next time they try. The fight doesn't seem to be over yet so we will have to wait and see.

Various sources are all over the map in regards to this front. ISW claims the RU attack was novel and powerful and other sources have the RU taking a beating. I have seen a "confirmed" RU loss list which is pretty grim, but I will see what Oryx shows.

 

[vikingatespam]

How have I tried to undermine Oryx? I am pointing out the limitations of their project, and the recklessness with which people use their numbers. I did not ask for someone on this forum to audit their work, though I can appreciate the work Feanor had done in this regard. What would be nice is an effort by multiple creditable sources to audit the entire body of work and publish their findings; this is done in every area of scientific research, not sure why it is such a provocative ask even if I know it will not be done.

You seem very agitated by this, not sure why. I'm surprised / annoyed that everyone of substance who actually has an interest in using the numbers does not have a better method than just quoting a blogger without caveat, regardless of how thorough we think that blogger is. I think that is a reasonable position.

I would apply the same standards to many other methods of statistical analysis. Nobody would have been ok with the media using a blog's investigative work to establish COVID-19 fatalities, for example.

Well, do you propose to pay someone for this, if you dont want to do it yourself ?

Why not spot check a statistical random sample of entries and check them yourself ?

 

[SinisterMinister]

Well, do you propose to pay someone for this, if you dont want to do it yourself ?

Why not spot check a statistical random sample of entries and check them yourself ?

This is getting tiresome. I am proposing that massive organizations with millions of dollars to spend and the ability to employ experts of every stripe are lazy for using Oryx's numbers without even attempting to verify them themselves, much less create their own database. It's surprising that Oryx is the only person doing this at scale given how much interest there is in the war from both governmental and non-governmental organizations.

As I said previously, I lack the expertise needed to properly identify the many extremely murky photos used. I would not expect random people on the internet to take this on, and I never suggested I would.