Abusing IQ tests in death penalty cases?

This seems pretty strange to me that IQ tests done after the person is charged or convicted of a crime should be relied on. There is too much of a possibility of gaming the system. From the New York Times.

A 44-year-man whose lawyers claim he is mentally retarded is scheduled to be executed Thursday evening in Huntsville, Tex., unless the United States Supreme Court agrees to hear his case. . . .

Mr. Woods’s lawyers argue that his intelligence scores are low enough that he should be spared because of the Supreme Court ban in Atkins v. Virginia. But several courts have rejected that claim. . . .

His intelligence was tested twice in grade school, and he received scores of 80 and 78, but defense lawyers argued that those scores should be adjusted downward to account for the age of the tests. As an adult, he was tested just before his trial and scored 70. A second test done in prison in 2002 showed him with an I.Q. of 68. . . .

An IQ of 84 being declared as mentally retarded seems bizarre.

Some state courts in Mississippi, Alabama and Texas have held that inmates with scores as low as 66 are not impaired, while an inmate in California with a score of 84 was declared mentally retarded. . . .

I found this: "The standard deviation used in many tests, including the Weschsler IQ test, is 15." If that is accurate here, that implies that 66.6 percent of the population falls between an IQ of 85 and 115. That means that people just slightly outside the normal range are classified as mentally retarded. The notion that about 16 percent of the population is mentally retarded seems pretty high.