Talk:Torture Photos from "Caesar"/Photo Timeline

Folder Dates

The point of this sub-page is setting a timeline for the photos, and central to that is the dates given in "folder" names. These aren't folders of photos we can access, just ones apparently created once by "Caesar" (whoever exactly that is) and now reflected in "photo original name" with each SAFMCD entry. HRW apparently saw the files as insiders have, in Caesar's own file system, and call the dates folders ("so and so's photos were in a folder dated this month, which fits with what people told us ...").

I wish we had something better. The dates given ate just given - not even proven accurate. But they do make basic sense, and I suppose 2-2013 is probably a folder created on, or referring to, February, 2013. Just what was put in it is another story, but the starting presumption is photos from around that time. "Around" in what sense isn't totally clear - before or after?

• Is it a no-later than date for a person's death? (photos he collects that day, taken on days up to that point, starting probably from the end of his last collection)
• Is it an earliest date? (the date he starts collecting photos (on a recurring basis?), up to the start of the next folder)

I thought the first option made more sense, and the answer could even be mixed, But two cases suggest it's the date he started a folder, or maybe the start date of a covered span ina folder he creates later.

• Ayham Ghazoul, said to die on day 5 of detention, November 9, 2012. (HRW report) His photo is dated 1-11-2012.
• Ahmad Musalmani, allegedly arrested from a checkpoint on August 2, 2012 (HRW report) is listed as j-1129, folder date 1-8-2012.

So the basic presumption is the date means photo added after the given date, and I guess prior to the next folder.

Dates are vague/sporadic: 5-2012. 000. Etc. There's no folder for each day, and in most cases not even for each month. A good 90-95% of victims by bulk are collected in these primary folders, with only a few smaller ones in between or earlier:

• 1-8-2012
• 1-11-2012
• 1-2013
• 2-2013
• 3-2013
• 4-6-2013
• 24-6-2013
• 7-7-2013
• 26-7-2013
• 14-8-2014

Two points stand out for have an extra wide gap between two extra big folder, with an unclear cut-off point between them. These are between August and November, 2012, and March and June, 2013.

The final folder, August 14, will cover just the last few days until "Caesar's" defection, reportedly on August 20. --Caustic Logic (talk) 09:24, 1 March 2016 (UTC)

Graphic aid: seen hospital numbers plotted by matching SAFMCD entry's folder date.

Summer Oddities

In the graphic above, some confusing points in the second half.

• 24-6 and 7-7's covered spans overlap, suggesting all those were available earlier, but the last ones only gathered later
• A folder 6-2013 in 215 at least includes entry numbers higher than those in 26 July, but lower than 14 August. Is that really just June, when high-number guys were killed early, when there's already a huge June 4 and 24 folder and at least a 27 as well? Maybe they meant 6-8-2013? That's about where it should be. I went ahead and mapped it like that above.
• At least branch 215 and 216 have their 26-7 and 7-7 folder backwards as far as victim order compared to the norm there. Is that maybe them meaning 2-7? Or was that an actual effort to make some prisoners - a whole folder's worth - out of sequence? (that would be to support the argument these are prisoner numbers). Notice the H#s mapped above continue properly between these 2 folders. --Caustic Logic (talk) 13:51, 2 March 2016 (UTC)

Torture Branch Profiles

Listed in order by number of photos/victims, as (now) presented at SAFMCD site. Note: Human Rights Watch received - from the SAFMCD-affiliated Syrian National Movement in March, 2013 - a set of the photos with slightly different numbers in each folder. They give a "Number of Caesar Victims Identified with Branch," not a number of photos, for the top 5, added for comparison with the same top 5 below. However, it seems to be a count of photos, just a bit different. --Caustic Logic (talk) 09:15, 5 March 2016 (UTC)

Throughout, "hospital number insertions" refers to the issue explained below, and better here.

Branch 215 Profile

MI Branch 215: secret raids and intrusions ( سرية المداهمة و الاقتحام ) - SAFMCD folder 3,556 photos. HRW Number of Caesar Victims Identified with Branch: 3,532

This branch allegedly killed over 50% of all the victims shown in the "Caesar" photos. So it matters that this is one of the total kill branches. 3,556 photos minus ( (48?) hospital # insertion and (6?) other duplicates) = about 3,500 reasonably claimed, numbered bodies.

The prisoner numbers run up to 4091 (there's a 4242, but it's the same photo as for 2042) missing: app. 590, with over 200 in a huge missing span* - maybe just 3 entries in the whole 3400s and 3500s - likely also killed but not included somehow - let's say 400 not shown, possibly spared, out 3,500 = 88-89% kill rate, at least.

• huge gap at 3400 and 3500s - starts maybe at 3391, stops maybe at 3606. In between those I could only find:

215-3499-7-7-2013 (2)-630273 215-3527-7-7-2013 (3)-413783 215-3589-7-7-2013 (2)-429926 Anyone finds any in between, let me know. --Caustic Logic (talk) 10:00, 3 March 2016 (UTC)

So ... With at least 90% of 215's prisoners killed (allegedly), any who claim they were held, witnessed valuable insights, and were then released to talk about it should be treated with skepticism. Let's consider 3 examples I've seen:

• "Hanadi," another woman: alleged female prisoners of 215, "Hannadi" spoke to HRW says about Rihab Allawi. She says she was arrested first, will have a lower prisoner ID that Rihab (who was #2935) who arrived, and would have # issued, late Jan/early Feb 2013. They spent 24 days together before Hanadi's transfer, she says, and they lost touch. But later she bumped into another woman who was jailed with Rihab afterwards, and apparently described the night she was taken away and killed. 3 women jailed by 215, presumably with ID #s but hard to predict and check, one dead and 2 released to bump into each other, allegedly explaining two of the relatively few whose ID#s should not appear in the "Caesar" photos.

"M." - a "student from Damascus University who was arrested on the same day" as anti-government activist Ayham Ghazzoul, per HRW. M. "said that security officers took him to a room in the medical college of Damascus University. Ayham was already there. The officers tortured the two students ... They were later transferred to the 215 Branch (Military Intelligence)together, and detained in the same cell." Then M. witnessed Ayham's death, he says. If ID #s are given on transfer, as "Caesar" alleges, their numbers should be adjacent. So in a near-total-kill branch, there should be an empty slot.

Ayham, allegedly, had ID# 320. I think I have this stretch all filled-in, and it shows those both killed and shown include:

• ID#s before: 300-319, every slot but 303 is filled, all killed.
• after: 321, 323, 324, 325, 326 327, 328, 329, 330, 331, 333, 334, 335.

So it could be "M." was #322, or even 332, one or 11 other guys numbered in between them, and that's why he's not there. Or, whoever has the number 322 was just re-numbered into another folder or whatever, to make room for this known witness to a high-profile case. --Caustic Logic (talk) 10:00, 3 March 2016 (UTC)

Branch 227 Profile

227 Zone Branch (Damascus): SAFMCD folder 2,047 photos. HRW, victims: 2,043

2,047 photos, adjusted for duplicates, etc. is unclear, maybe 2010 credibly claimed and numbered victims shown? I have this folder about halfway processed, so there are still many gaps, but a good overall feel.

I don't think the credible and continuous victim numbers go any higher than 2825 - if 2010 are seen, it means a total of around 815 prisoners were either spared, sent elsewhere, etc. Implied kill rate: at least 71%.

227 alleged prisoner numbers - well over 1,000 of them so far - just roll upward steadily each month. If the SAFMCD version is true, this branch, like most others, kills nearly all of its prisoners. And it does so in the same order they were transferred/arrested and numbered, irrespective of case details, after what seems a uniform and pre-decided delay, with the delay maybe adjusted for changes in the overall death rate, etc.

Missing slots: the first 11 are, as usual, sporadic and confused-seeming low numbers, some near 200, then 316, 324, 339, and then more solid with 340, 343, 344, etc. up the line. Most slots from there seem filled. But out of first 338 #s, we see only about 13 entries (and this part seems complete now - if I don't have it, it seems SAFMCD doesn't either). So 325 of the missing entries are here, and the rest are spread around however. No massive gaps detected yet, unless we could really call a gap 6,279 entries wide, between #2825 and #9108 (see next). --Caustic Logic (talk) 09:01, 4 March 2016 (UTC)

Out-of-order:

At least two oddball entries in the 9100s could mean there were way more prisoners, with an implied kill rate then only about 22%. However, these might belong in the AF intel folder. 227 prisoners 9108 and 9109 are in the 2-2013 folder. AF intel was killing 9100s at that time, and has a gap between 9099 and 9119. I think these were supposed to be in there. It seems only these 2 appear in 227, so 16 prisoners might be spared/not shown/in yet another folder, maybe even with the numbers actually changed... (wouldn't that be genius?)

Further, the gap between 2825 (where continuous #s stop) and 9108 has at least 5 other entries, all in the 3,600s, and no others I've found. These even more clearly belong in branch 215. 3670 is the key - a faceless body shot run by ZaW here drove me nuts - Card says detainee #215-3670, H#3976/b. He's got a curved-sword Shia tattoo, 2 less-clear tattoos, and 2 (bullet?) holes in the right arm - he was captured, starved, killed, and now left pants-down by Caesar. And he's one prisoner # and one H# above 215-3669, 7-7-2013, also starved with a Shia word tattoo unusually visible. But I could not find this adjacent entry in the 215 folder at SAFMCD. I got all in the immediate range except for 3670, 3675, and 3677. Tried finding the one 3 times, no luck.

Finally, I checked other folders and already had a 227-3670 processed, same shoulder marks and light on collarbone, same number and date, just wrong branch. Note: we can see in one view no major body injury, and yet like so many, he died coughing up blood. There's also an oddball 227-3675, 3677, and a 3658 I've found so far, all dated 7-7 (215's 3658 also seems to be missing now that I check...). Oh, and a youthful 227-3654 dated 24-6 (also no 215 equivalent found). So that's at least 5 more misplaced here, for at least 7 total. (again, not that the first placement was even truthful...)

Others out of order (found so far) just by date, it seems:

• 227-1610-2-2013 (2)-561241 others killed 4-6 in the range 1040s to 1240s.
• 227-2048 and 2058, 1-11-2012 (H# insertions)
• 2569 and 2570 are slightly out of order

--Caustic Logic (talk) 13:28, 11 April 2016 (UTC)

Summer oddities folder order and prisoner # range represented

• 4-6 1484-2005
• 7-6 2030-2085
• 24-6 2086-2491
• 7-7 (partial) 2494-2519
• 26-7 2520-2563 + 2570
• 7-7 (resuming) 2577-2662-
• 000 2664-2779 + 2569
• (no "6-2013" entries yet)
• 14-8 2780-2825

Total range: 1484-2825 = 1,341 - whatever proportion of these appear (seems well above the 71% average). At any rate, half the span of eliminated prisoner numbers was covered in these folders, claiming a time span of just 2 months and 10 days. Alleged 227 prisoners from the first half of the number scale had taken nearly 27 months to accumulate. --Caustic Logic (talk) 09:01, 4 March 2016 (UTC)

AF Intel Profile

Air Force Intelligence Branch ( افرع المخابرات الجوية ): folder link - 350 photos. HRW # of victims: 352.

This is the first branch folder I completely processed - all 350 photos included correlated with the info they have, and so I can make authoritative statements about what's in there. numbers said to be issued by them are prefixed j, from the Arabic word for Air Force. Some of the few j-series numbers represented are among those 133 or more apparently concocted, by re-branding the hospital number (seems to mean "unidentified body," not coded prisoner) as a supposed ID number in one or another of several MI branches. ] j-1129, HRW's profiled teenage victim Ahmad Musalmani, is included in that pattern.

J-Jumps

First, the alleged prisoner numbers range wildly here, from 000 (several cases, maybe errors, etc.) to just over 9700. This implies they alone had arrested that many people prior to mid-August, 2013. We can only say they allegedly killed ... less than 350 (duplicates, count still in progress) out of this, maybe more but no others appear in the "Caesar" released set as shown on the SAFMCD site.

What's weird is the spread - I have a graphic in the works but mainly they pick groups numbered right around the 1000 or 100 mark. It's a bit muddled at the low end, but here's the pattern from 2000 onward:

• 2026-2313 = 60 out of 288 killed
  • 2324-3026 = none at all included
• 3026-3189, = 35/163 (inc 1 dup)
  • 3190-5005 total gap
• 5006-5063 = 11/58
  • another gap, almost 2,000 #s wide
• 7016-7027, 3/12
  • gap
• 8005-8093 = 40/89
  • gap
• 9013-9246 = 105 / 234 (nearly half of these killed and show)
  • gap
• 9500-9508 = 9/9 killed and shown
  • gap
• 9700-9704 = 5/5 included

graphic added: red bars are where ID #s appear, gray spans are where none do - except the 3,000s, which I somehow missed (revision coming) --Caustic Logic (talk) 10:02, 2 March 2016 (UTC)

The huge implied numbers of prisoners in between these number sets - not clear what the alleged story is there. Maybe they were killed and sent elsewhere, sent here but not included somewhere along the line, spared, killed later, etc. With alleged numbers so huge, surely the allegation would be some mix of fates. But it almost seems like those at the top of each page or whatever definitely got picked for decimation, from narrow bands totally killed to wider ones picked-off to varying degrees. Allegedly. --Caustic Logic (talk) 08:35, 1 March 2016 (UTC)

Prisoner # Assignation Order They say prisoner # is assigned in order of arrest or transfer - and I presumed it would be assigned sequentially - 215, 216, 227, at least (about 90% of all photos between these) seem to go sequentially - All 1240s-1260s, for example, must have all been arrested around the same time, and we can see they're almost universally recorded as killed around the same time.

But apparently not by AF Intel. They at least have a whole different alleged logic it's hard to grasp. I found so far 5 Prisoner #s attached. to AF Intel, all but #2 from HRW report (4 of their 8 profiles were from AF intel, way out of proportion), in what should be chronological order.

1. 1129 = 2012-08-02 2 August 2012 - Ahmad Mussalmani - issued on arest by AF Intel this day, killed quickly, in 1-8 folder (must be start date as with A. Ghazoul in 1-11) - j-1129-1-8-2012 (1)-026668
2. 1958 = 2012-07 - Mr. Abdul Wahab, arrested May 23, 2012, transferred to AF Intel around July 14, released Sept. 2 (SNHR report)
3. 8025 = 2012-03-28 or soon after - Oqba al-Mashaan - "The folder they appear in is marked March 2013, a year from the date of his arrest." - j-8025-3-2013 (1)-451848
4. 9189 = 2013-01 - Khalid Hadla - "Air Force Intelligence officers arrested Khalid in January 2013 during a raid on his home, while his wife and children watched" - not dated, Daraya, clashes had both sides in control at different times that month "The photo was in a folder dated June 4, 2013, about five months after Khalid’s arrest." - j-9189-4-6-2013 (1)-506879
5. ja-001 (issued later than the last?) = 2013-02 - Osama Hussein Salim, Palestinian from Yarmouk, accountant at Damascus University - arrested by PFLP-GC "around 10 a.m. on February 24, 2013" transfer unclear, killed within the month. "Osama’s body has the words “Air Force Administrative Branch” written in marker on his abdomen." (hence the "a") ja-001-2-2013 (2)-944416 --Caustic Logic (talk) 08:35, 1 March 2016 (UTC)

These 5 are included in the graphic above on the ID# scale, with arrest date (blue line) vs. death/shown/folder date (red), except for the guy who says he was just released (no red line). Consistently, his stated number is NOT in any of the kill zones. Did he know what he was talking about here? --Caustic Logic (talk) 09:54, 2 March 2016 (UTC)

Prisoner Kill Order It's also evident the numbers do not tick up chronologically with this branch, but follow their own weird pattern. Again, appearing in a folder means killed and shown, while not appearing can mean a few things. But in each folder, representatives of each of these little red groups appear. In each folder, no one from the gray spans in between does. So the reason for that is ongoing. It's almost as if they decided on this narrow killing in advance, and worked on it a bit each month. And had the full number range already filled from the start.

For one example, a heavy date, January 2013. That month this is the list of prisoner numbers that appear.

002

953 1113 1117 1118 1121 1123 1130

2120 2128 2131 2133 2132 2133 2138 2146 2173

3059 3070 3072 3075 3084

4012

5015

7016 7017

8012 8015.

Everyone else was spared/sent elsewhere or photo not collected/not shared/who knows. It's the same pattern month after month, a few others from each red span, none from the gray, until this pattern is filled in. --Caustic Logic (talk) 09:43, 2 March 2016 (UTC)

Well, the 9000s aren't represented here - maybe not arrested yet, or just not getting decimated yet. But oddly, as noted above, branch 227 had 2 unusual victims in that range - 9108, 9109 - appear in their 1-2013 folder. They're very out of place there (thousands of #s apart from the rest), but would fit perfect here. Was there a mix-up perhaps? --Caustic Logic (talk) 09:15, 5 March 2016 (UTC)

216 Profile

216 - patrols Branch SAFMCD folder 297 photos. HRW, victims: 293.

This is a total kill branch - little concern with interrogation, individual case details, etc., just with exterminating people. 297 photos minus (2+ duplicates, 2 hsopital # insertions) = 293 or so. I have this folder mostly processed, emphasis on the early and late ends. I don't think victim numbers run any higher than 384. Only 2 higher #s appear, 692 and 693, being the hospital number insertions, all alone in mid-2012, when apparently no one else was killed until November. In 1-11 are the lowest numbers seen, a lonely 14, then 45, 47, 49, 51, 52, and getting more solid. So 43 missing numbers at the start. Each month, generally the numbers go up. Minimum number range for each (not all slots in between are filled)

• 1-11-2012 includes up to #74
• 1-2013 includes #s 77-101,
• 2-2013 104-127
• 3-2013 includes #136 137 138
• 4-6-2013: 132 133 (late) and #150-221
• 7-6: 242, 245
• 24-6: 254 - 310
• 7-7 338 - 351
• 26-7 314 - 337
• 000: 353-363
• 14-8: 364-384

(note: 7-7 and 26-7 seem to be in the wrong order - 215 at least has the same issue. This will be addressed above under folder dates.)

At least 384 assigned prisoner numbers - 293 or so photographed victims = 91 numbered prisoners, at least, who were not killed and shown. 43 of these are in the range 1-44, the rest spread around in smaller, unclear gaps. But they killed at least 293 of perhaps no more than 384 people they arrested - a 76% kill rate, or maybe closer to 99%, depending, and it was mainly sequential - killed in about the same order arrested. --Caustic Logic (talk) 13:24, 2 March 2016 (UTC)

235 Profile

235 - Palestine Branch - ( 235 - فرع فلسطين ) - SAFMCD folder 128 photos. HRW: 127.

Logical prisoner numbers run from 008 (1-11-2012) up to 182, probably no higher. Three higher numbers (731, 823, 1242) are hospital # insertions. Gaps not clear yet. 026 is repeated on 2 different prisoners.

182 prisoners - app. 124 victims shown = app. 58 victims at most (in that range anyway) that were spared. Implied kill rate: at least 68%. --Caustic Logic (talk) 09:15, 5 March 2016 (UTC)

I've scanned all the images I have (most of them...) and got a ton of hospital/exam/body # (above, card bottom #). Every number on these cards is suspect, but this is the most likely to be reliable/valid and also seems to map out well over time. --Caustic Logic (talk) 15:26, 16 February 2016 (UTC)

By Hospital Number

moved from front page

One of the main points of this page and tables above was to check time correlation (not clear, covered below) and the claims that over 11,000 victims are suggested (they say bodies were numbered 1-5,000, then re-started with a /b at the end up to 5,000/b, and then started on a third letter. Other than the first 2,000 entries and the last part after 10,000 is reached, this is confirmed quite well, and two full sets of 5,000 are suggested. There is no (clear) coverage before entry 2065, and at the late end this odd spot - either 544 and a separate 4 for no clear reason, or this is body number 5444 with no /b. Maybe the switch was after this? Not clear enough to include in this quick table shows the lowest non- and highest (clear, and consistent) number in each series. --Caustic Logic (talk) 14:30, 18 February 2016 (UTC)

body numbers closest to 5,000/0 switches
Hosp#	victim#	date	sources
2065	227-394	1-11-2012 (6)	SAFMCD entry SNHR photo ZaW ZaW ZaW
4978	215-1385	3-2013 ?	SAFMCD
10/b	215-1414	3-2013 (2)	SAFMCD entry (shows card)
4994/b	227-2791	14-8-2013 (5)	SAFMCD ZaW

Timeline: I plotted out several dozen cases where I can match that victim to a SAFMCD entry, with original file name including folder date. The folder dates aren't gospel, and as we can see below they mainly aren't very specific - about one a month. At the moment, I'm taking the date as creation date, filled with new and maybe previous images, and also added to until the next one is started. Not clear.

It's hyped how "Caesar's" photos show torture and repression for the whole conflict from March, 2011 to August, 2013. But the numbers suggest only about 1,000 unidentified bodies - less than 10% of the total implied - were logged between the start and July, 2012. 1,000 in 15 months is an average of less than 70 a month.

Probably that many victims are listed, but no reliable #s yet. The earliest possible hospital #s are in the 500-1100 range – this is a thing that emerged analyzing Majed Abu Draa – it seems likely his hospital number 1075 was re-branded as branch 215 ID# way out of sequence - others in the span seem to fit the same pattern, often duplicating ID 3s in the process. … these seem to fit actually, correlating logically to June and July 2012 folders.

But otherwise, there's a huge gap; the first clear H #s are in low 2000s, already in the 1-11-2012 folder. Here I've mapped the numbers from here forward, from 2,000 to 5,000, repeating 1,000-5,000 with a /b suffix, and the suggested 1,000+ past that, although no examples at all have been seen yet.

It's not clear how far back the 1-11 folder covers, but likely some span - it's huge. And on the other end, #s reportedly go to over 11,000 and the last folder 14-8 is both huge and short on body shots with visible H#s - they probably do tick up that far unseen to us. So … this implies at least 8000 and perhaps over 10,000 unidentified or coded bodies passed through in 10 months, or the span covered by these folders. I suspect 1-11 covers the time frame of the Daraya in August and whatever else, from the size of it, so likely a span of one year. 8-10,000 over 10-12 month = Average rate: 700-1,000/month - at least ten times the rate over the first year. Final 5-6 months saw about 6,500 - at least 1,083/month.

What strange luck - just as "Caesar" got his insider documentation system down, someone decided to start really sending mass numbers of horribly abused bodies for him to document. And it might be the final folder was the worst yet (pending analysis, range of all folders a bit unclear...) and I bet it fell off once he stopped recording, and also the Ghouta massacre happened, and hostage stocks might finally have gotten low. --Caustic Logic (talk) 15:26, 16 February 2016 (UTC)

Also FWIW the branch victim #s tend to tick upward logically as if they covered a portion of the sequentially discovered unidentified bodies. At least 215 and 227 (about 75% of the total) do this, other than the few odd entries or groups like mentioned above. The one exception I've noticed is Air Force intel, J - their numbers bounce around a lot - 9000s killed and then 2,000s a month later, etc. But they have other weird patterns too I mean to explain somewhere. --Caustic Logic (talk) 15:39, 16 February 2016 (UTC)

And before I confuse many people, my suspicion now is the branch victim # is made up by the terrorist killers, written on the bodies before dumping or even before killing them. "Caesar" had the job of "verifying" that, maybe just by putting the real hospital # alongside the numbers on the body, on these official cards. So the supposed proof of bouncing numbers, which is rare anyway, is just a poorly written attempted proof. The terrorists mainly just numbered their victims sequentially based on the order they were murdered, or numbered within a batch about to gt it. This wasn't too bright as far as setting "Caesar" up to claim it's a military intel prisoner # ... but it would be easier to keep track of and avoid screwing up. But as we'll see, it got screwed up anyway. --Caustic Logic (talk) 15:21, 18 February 2016 (UTC)

Early Entries Re-Branded?

I apologize for going on in talk mode on a front page. It seemed right for this case, plus I've gotten rusty on site protocol lately. --Caustic Logic (talk) 16:39, 21 February 2016 (UTC)

Clearer but long explanation of this issue: Fail "Caesar" Part 7: Inventing More Coded Prisoners?. The following but for the big table becomes more like optional reading. --Caustic Logic (talk) 10:33, 27 February 2016 (UTC)

Finally just moved it here to the talk page. --Caustic Logic (talk) 09:24, 5 March 2016 (UTC)

As mentioned above, case of Majid Abu Draa has an odd issue with his 215 ID# 1075 (he also has a branch 227 ID#) being seen in this fuller view on a card - no 215 is specified, and no other number. Usually when we see that, especially on early photos, it's the hospital or unidentified body number. As we'll see, that number makes sense in context of mid-2012.

There might be another 1075, but not that I've found yet. What we do see is this ...

Supposed prisoner ID# is repeated three times here that I had found - 1071, 1072, 1123, each having on repeat an early, mid-2012 hospital photo like Majid's, the other like and dated like those around it. I've discovered more in the shown range since making this plate, further into the 110s and sporadically back into the 500s - maybe 70 entries or more, some with later duplicates found, some not yet. What are the odds they decided to kill several 500-1100 range prisoners in mid-2012 and the rest six months later (an unusual pattern), and happen to have several duplicate ID #s in this narrow range, and hardly anywhere else? Majid is the only case where we can see the number on a card, and it's shady - looks like "Caesar" decided to fill in some 215 victims that never existed, and re-branded the right hospital #s. It also helps prove his theory the branch victim 3 represents prisoner ID, not death or finding date - how can some of these people be found six months ahead of the rest? Right?

Anyway ... he's the only direct glimpse but it leads into a pattern that keeps making sense. I noticed the same issue to a lesser degree in the 227 folder - which is less complete on my end so far. But I filled it in some more and then found enough instances to show the same pattern continuing. Some duplicates, usually with one a 1-6, 1-7 or 1-8 hospital photo. (there are some other kinds of duplication, but just one other set I've included below). As we can see, it's a and continuous string of entries with this kind of number jammed in even when it conflicts with later entries, split between two branches with no repeats noticed yet within the string, just within the branches it's jammed into ... noticeable switch-over points. Those with an identified later duplicate number they trip over indicated with a D. --Caustic Logic (talk) 16:24, 18 February 2016 (UTC)

And also, the same seems to continue into other branches, at least 216, and Air Force Intelligence. Probably the others too... --Caustic Logic (talk) 14:54, 19 February 2016 (UTC)

Some appear in Sporadic Views, but numbers and dates don't lead to clear pattern fits (any that do witll be in the "other" column, marked S). More were in the natural place - Unknown or Anonymous branch - the MI division responsible wasn't clear in these cases - no forehead tape, just a coded victim # somehow known but no one it's pinned on. In each of the several cases I'll add below, it's an early hospital photo, dated that way 5-8 of 2012), with a mystery number fitting exactly into the pattern, and probably just what the hospital # would be at that time. In the other cases, they decided just who to pin that number on. --Caustic Logic (talk) 07:14, 20 February 2016 (UTC)

(select links at least - 6-digit # included lets you search the SAFMCD site for the picture)

• D = duplicate
• In "Other" column:
  • j=Air Force Intelligence
  • S="Sporadic Views"
  • U=Unknown/Anonymous Branch

Suspect branch ID #s in 215, 227, and other folders
215	227	other	date
	251 (Majid Abu Draa) (number relevance unclear - ignore)		3-7-2012
		U293-001?	5-2012
		j-352	5-2012
		j-353	216-353?
		216-353? (visual)	000 (summer 2013)
		j-354, 355, 364	5-2012
		j-365 - 865758	5-2012
	372 632999		5-2012
		S 393-regiment 274 - 065218	5-2012
		S 394 regiment 274 -001571	5-2012
		S 397-5-2012-274 regiment -459030	5-2012
		U409	5-2012
		251-428	5-2012
		U 434 435 436 438 439 440 441 442 443 444 445	5-2012
		U 456 457 458 459	1-6-2012
		463, 464 465 466	1-6-2012
		U 467	1-6-2012
		U 471 472 473	1-6-2012
		U 489
		U 490, 491 492, 493, 494, 495	1-6-2012
		U 496 - 011153	1-6-2012
		U 497	1-6-2012
		U 498 - 186407	1-6-2012
		U 499 - 122735	1-6-2012
		U534 - 735334	7-2013
		U 535? (AA Hariri) 765908 – S 535-ninth division-948944	7-2013 vs. 1-6-2012
	537 D		1-6-2012 vs. 1-11-2012
538			1-6-2012
539 D			1-6-2012
		j-540
541 D 980690 vs. 206942			1-6-2012
542 D 487909 vs. 505499			1-6-2012
543 D 008181 vs. 960082			1-6-2012
544 D
545 D
546 D 042116 vs. 530202
547			1-6-2012
548 D			1-6-2012
549 D 964084 vs. 001163			1-6-2012
		U 550? - 906179	7-2013
551 011433			1-6-2012
562 D 343870 vs. 309859			1-6-2012
563 D 708519 vs. 719701			1-6-2012
564 - 480173			1-6-2012
565 D			1-6-2012
566 D			1-6-2012
567 D 702949 vs. 731760			1-6-2012
568
569 D
570
571 D
		j-580 331677	1-6-2012
		U 600 – 289955	1-6-2012
		U 601 – 392704	1-6-2012
		U 604	1-6-2012
		U 607 - 091005	1-6-2012
		U 608	1-6-2012
		U 610	1-6-2012
		U 614	1-6-2012
		U 615 - 945650	1-6-2012
		U 616 - 311447
		U 617 - 950023
		U 618 - 949716
		U 619 - 718503	1-6-2012
		U 620 - 324061	1-6-2012
		U 621	1-6-2012
		U 635 - 553426	1-7-2013
		U 636 (dup - 646) 146643	1-7-2013 (killed with blades)
		U 638
		U 639-645
		U 646 (dup - 636) 710427	1-7-2012
		U 648 - 928096	1-7-2012
		U 649
		U 650
	651		1-7-2012
		U 652
		U 653
		U 654
		U 655
		U 656
		U 657 - 861455	1-7-2012
		U 658 - 139634	1-7-2012
		U 659 - 507403	1-7-2012
		U 660
	661 D		1-7-2012
662			1-7-2012
663			1-7-2012
	664
665 D			1-7-2012
666			1-7-2012
667 D			1-7-2012
668 D			1-7-2012 vs. 1-2013
	669 D		1-7-2012
		216-692 - 748420	1-7-2012
		216-693 - 526627	1-7-2012
	694		1-7-2012 vs. 000
	695 D		1-7-2012 vs. 000
725 D 046717 vs 830178
726 D 541496 vs 427450
727 D
728 D
729 D			1-7-2012 vs. 2-2013
730 D
		235-731 -- 495103	1-7-2012
732 D			1-7-2012 vs 2-2013
		S 293-733 - 217897	1-7-2012
		U 820 - 191816	1-7-2012
	227-821 - 206027		1-7-2012
822 D			1-7-2012 vs. 2-2013
		235-823 -- 911914	1-7-2012
	900		1-7-2012
901 D			1-7-2012 vs. 2-2013
	902		1-7-2012
1071 D 077754 vs. 241307
1072 D 250700 vs. 359487
1073 294133 D			1-7-2012
1074 D			1-7-2012
1075 (Majid Abu Draa) D			1-7-2012
	1121 - 71724
1122 - 874195			1-8-2012
1123 D 508274 vs. 667457
	1124 - 653553		1-8-2012
1126
		j-1129 - 026668	1-8-2012 Ahmad al-Musalmani,age 14 HRW report
		j-1140 -869397	1-8-2012
	1141 D 514561 vs. 885589		1-8-2012 vs. 1-2013
		j-1142--004384	1-8-2012 (1)
	1143 D - 391823 vs. 1-8-2012 (2)- 1-2013-duplicate (2)		1-8-2012 vs. 1-2013-duplicate
		j-1144 - 877001	1-8-2012 (1)
	1158-1-8-2012 (1)-390836
	1159-1-8-2012 (2)-209398
	1160-1-8-2012 (2)-126612
1161			1-8-2012
1162			1-8-2012
1163 D			1-8-2012
	1164 - 663667		1-8-2012
		235-1242 – 798873	17-8-2012
	1243 D – 288788 vs 181720		1-8-2012 (2) vs 1-2013
	1244 D – 342853 vs 1244-(duplicate)(1)-751447		1-8-2012 (2)- vs. 1-2013
	1252 - 445666		1-8-2012
	1253		1-8-2012
1254 - 347201			1-8-2012 (2)
		S 1263? 027601	12-2011
		U 1270 - 941539	1-2012 (prob means 1-8)
		U 1277 - 342103	1-2012 (prob means 1-8)
	1277-1287?		2-2013 (probably not matches)
		1285? (291-128-5-8-2012 (4)-441700)
	2048		1-11-2012
	2058		1-11-2012
		j-2084 -626746	1-11-2012
		j-2092? -884864	000
		j-2098? - 038005	000
	2099		1-11-2012
	2810?		14-8-2013 (but looks like mid-2012 hospital photo)
		U 2829? - 988693	000
		251-3634 - 609146	2-2013 same as mp1559 (233026)
		251 - 3831-3840	2-2013 (see "decayed victim" here
		S 3851	2-2013-Intelligence department
		S 3863	2-2013-Intelligence department
		U 4315 –	1-2013 (folder-bouncing is not uncommon)
big gap,	almost 5,000 more bodies,	with none I've seen	re-branded this way,
then,	as if	to be tricky ...
		U 4308 (/b)-578929	24-6-2013
		U 4643 -768559	7-7-2013

At right, a visualization of one well-filled-in stretch, likely re-branded hospital numbers 535-549 compared with duplicate entries in the 215 folder.

2048 and 2058 do not fit the main pattern, but instead re-illustrate it - in these two cases, there are no duplicates yet, but the November asphalt photos are shoehorned in with similar-numbered victims killed in July, 2013. The running victim numbers for most in the November folder was back in the 3-400 range. What's odd, as we see above with the lowest body number entry, he's 227-394 - same span ... body number 2065. What are the odds the few out-of-sequence prisoners killed here 2048, 2058, also had victim numbers matching with current hospital # range. Too slim, sorry. So this marks a second occurrence of the pattern illustrated by Majid. !! Cool sound of holes tearing? --Caustic Logic (talk) 16:24, 18 February 2016 (UTC)

Same for j-2084 - what odds! TWO different branches pulled the same thing at the same time, victim #s between 2000 and 2100, despite their entirely different systems. --Caustic Logic (talk) 15:05, 19 February 2016 (UTC)

Another later entry helps illustrate this is a re-branded body/hospital # - unknown branch contains 578929 labeled 4308-24-6-2013 (1). No branch but 215 and J have "prisoner ID" #s that high, and this didn't seem to fit. But if we take it as the other number, and presume it's missing a /b, it fits perfect between the closest entries:

• 4264/b -- 227-2615 -- 7-7-2013 - starved, eye damage/chemical signs (eyes, skin, blood/mucous)- other views show tattoos showing he was a Christian
• 4308/b? -- 4308 (branch unknown/never invented) -- 24-6-2013 - starved, eyes out/damaged, slight decay (dry), chem signs (skin irritation, poss. cyanosis, eyes)
• 4311/b -- 216-305 -- 24-6-2013 - decayed (discolored bloating, eyes popping out), possible chemical signs (purple color, poss. eye damage)

Note this even fits the switch from the 7-7 folder (first entry) to some later bodies documented earlier in the 24-6 folder (the others). For a third pretty clear point, we can see this is just the number of a body that was verified as just a body and cataloged at hospital 601. In all but the "U" cases above, someone on "Team Caesar" decided how to divvy these up to different MI branches, when it seems nothing did this in the first place. These victims tend to be early in the operation, and mostly lack the trademark forehead tape we saw appearing later. I suppose the whole idea hadn't gelled just yet. --Caustic Logic (talk) 14:53, 20 February 2016 (UTC)

Same for unknown 4643-7-7-2013 (1)-768559 - fits between 4450/b (248-63, 26-7-2013) and 4813/b (215-4002, 6-2013), with some more folder bouncing. --Caustic Logic (talk) 05:39, 21 February 2016 (UTC)

If we ignore the apparently fake branch attribution and take these non-repeating numbers as hospital/body numbers, and mapped it out by folder, it would look like this below. The purple span correlates to the start of the later span covered in the graphic further up the page. One might note how logically this fits with the established pattern of a steadily increasing death toll in the "Caesar" photos system. --Caustic Logic (talk) 16:39, 21 February 2016 (UTC)