During his intensive initial round of media interviews as commander in Afghanistan in August 2010, Gen. David Petraeus released figures to the news media that claimed spectacular success for raids by Special Operations Forces: in a 90-day period from May through July, SOF units had captured 1,355 rank-and-file Taliban, killed another 1,031, and killed or captured 365 middle or high-ranking Taliban.
The claims of huge numbers of Taliban captured and killed continued through the rest of 2010. In December, Petraeus's command said a total of 4,100 Taliban rank and file had been captured in the previous six months and 2,000 had been killed.
Those figures were critical to creating a new media narrative hailing the success of SOF operations as reversing what had been a losing U.S. strategy in Afghanistan.
But it turns out that more than 80 percent of those called captured Taliban fighters were released within days of having been picked up, because they were found to have been innocent civilians, according to official U.S. military data.
Even more were later released from the main U.S. detention facility at Bagram airbase called the Detention Facility in Parwan after having their files reviewed by a panel of military officers.
The timing of Petraeus's claim of Taliban fighters captured or killed, moreover, indicates that he knew that four out of five of those he was claiming as "captured Taliban rank and file" were not Taliban fighters at all.
Checking on the claims of the number of Taliban commanders and rank and file killed is impossible, but the claims of Taliban captured could be checked against official data on admission of detainees added to Parwan.
An Afghan detained by U.S. or NATO forces can only be held in a Forward Operating Base for a maximum of 14 days before a decision must be made about whether to release the individual or send him to Parwan for longer-term detention.
IPS has now obtained an unclassified graph by Task Force 435, the military command responsible for detainee affairs, on Parwan's monthly intake and release totals for 2010, which shows that only 270 detainees were admitted to that facility during the 90-day period from May through July 2010.
That figure also includes alleged Taliban commanders who were sent to Parwan and whom Petraeus counted separately from the rank and file figure. Thus more than four out of every five Afghans said to have been Taliban fighters captured during that period had been released within two weeks as innocent civilians.
When Petraeus decided in mid-August to release the figure of 1,355 Taliban rank and file allegedly captured during the 90-day period, he already knew that 80 percent or more of that total had already been released.
Major Sunset R. Belinsky, the ISAF press officer for SOF operations, conceded to IPS last September that the 1,355 figure applied only to "initial detentions".
Task Force 435 commander Adm. Robert Harward confirmed in a press briefing for Journalists Nov. 30, 2010 that 80 percent of the Afghans detained by the U.S. military during the entire year to that point had been released within two weeks.
"This year, in this battlespace, approximately 5,500 individuals have been detained," Harward said, adding the crucial fact that "about 1,100 have come to the detention facility in Parwan."
Harward did not explain the discrepancy between the two figures, however, and no journalist attending the Pentagon briefing asked for such an explanation.
Petraeus continued to exploit media ignorance of the discrepancy between the number of Taliban rank and file said to have been "captured" and the number actually sent to the FDIP.
In early December, ISAF gave Bill Roggio, a blogger for "The Long War Journal" website, the figure of more than 4,100 "enemy fighters" captured from Jun. 1 through Nov. 30, along with 2,000 rank and file Taliban killed.
But during those six months, only 690 individuals were sent to Parwan, according to the Task Force 435 data – 17 percent of the 4,100 Taliban rank and file claimed captured as "Taliban".
The total of 690 detainees also includes an unknown number of commanders counted separately by Petraeus and a large number of detainees who were later released from Parwan. Considering those two factors, the actual proportion of those claimed as captured Taliban who were found not to be part of the Taliban organisation rises to 90 percent or even higher.
Three hundred forty-five detainees, or 20 percent of the 1,686 total number of those who were detained in Parwan from June through November, were released upon review of their cases, according to the same Feb. 5, 2011 Task Force document obtained by IPS. The vast majority of those released from the facility had been sent to Parwan in June or later.
Detainees are released from Parwan only when the evidence against them is so obviously weak or nonexistent that U.S officers cannot justify continuing to hold them, despite the fact that the detainees lack normal procedural rights in the "non-adversarial" hearing by the Task Force's Detainee Review.
The deliberate confusion sowed by Petraeus by referring to anyone picked up for interrogation as a captured rank and file Taliban was a key element of a carefully considered strategy for creating a more favourable image of the war.
As Associated Press reporter Kimberly Dozier wrote in a Sep. 3, 2010 news analysis after an interview with Petraeus, he was very conscious that "demonstrating progress is difficult in a war fought in hundreds of small, scattered engagements, where frontlines do not move and where cities do not fall."
SOF raids, however, could be turned into a dramatic story line. "The mystique of elite, highly trained commandos swooping down on an unsuspecting Taliban leader in the dead of night plays well back home," wrote Dozier, "especially at a time when much of the news from Afghanistan focuses on rising American deaths and frustration with the Afghan government."
Petraeus made sure the impact of the new SOF narrative would be maximised by presenting the total of Afghans swept up in SOF raids as actual Taliban fighters.
The deceptive nature of those statistics, as now revealed by U.S. military data, raises anew the question of whether the statistics released by Petraeus on killing of alleged Taliban were similarly skewed.
Gareth Porter is an investigative historian and journalist specialising in U.S. national security policy. The paperback edition of his latest book, "Perils of Dominance: Imbalance of Power and the Road to War in Vietnam", was published in 2006.