Would you really call adjusting a statistical window "playing around with numbers"?

It doesn't look like they tried multiple parameters to get a result they liked, they just looked at the time frame that made sense.