All you're doing is proving more and more that you don't understand the principles at play here.
If I had selected 3 x 10^9 instead of 3 x 10^3 points, there still wouldn't be a hcokey stick. Because a hockey stick shape has a tiny chance of realizing ina random number generator.
How is this still being discussed? McIntyre and McKitrick sampled randomly from data that had a trend...of course you're going to get a hockey stick when the data has a hockey stick shape. How is that shocking? You put trended data into a noise model, you will still get a trend. If you take any time series that has a significant trend, and randomly sample from it, you're going to get roughly the same thing. It will then be just trend plus noise, but it's still trend, and when you take the long-term average of the time series, you get essentially the same thing.
If you want to call something random noise, then you can't start with significantly trended data. That is so uncontroversial, I can't believe that this is still being discussed.