do me a favor and...
...compute these odds for me.
In one of the older accu-stats videos, one of the commentators (don't remember who) were talking about how they were present at this one tournament where Earl ran 6 racks...ran another 5 racks...and then another 6 racks on three different opponents...all in the SAME tournament!
I don't think the law of averages should apply to Earl.
...compute these odds for me.
In one of the older accu-stats videos, one of the commentators (don't remember who) were talking about how they were present at this one tournament where Earl ran 6 racks...ran another 5 racks...and then another 6 racks on three different opponents...all in the SAME tournament!
I don't think the law of averages should apply to Earl.
The fractional error is roughly one over the square root of the number of samples. That means that if you have 30 samples, your estimate is likely off by 18% and it would not be surprising if your estimate is off by twice that. (This depends a lot on exactly what you are trying to "measure.")
In polls you often hear "the results are about +- 3%". That's one in 33 which means they had about 1000 people in the poll (roughly 33x33).
Another example is flipping coins. If you flip a coin 100 times you expect 50 heads on average but the expected variation is 1/10th of that or 5. And a deviation from 50 of 10 would not be surprising.