This is a very problematic estimate due to

your biased samples. Your rms error could be
1000% from so limited samples.You should confirm your
information before posting here. The poor people here
are already very nervous. Please don't post any
unconfirmed information here. Thanks.