Ferry Avenue wrote:I'm fully willing to agree that the sample size doesn't permit one to conclude there was a significant difference in Embiid's performance, statistically speaking, between the regular season and postseason. However, that's completely consistent with the point being made -- there isn't a significant difference.
No it's not, the point being made, again. Is that the test is always gonna say there is no difference because there is too much variance.
The reality is that the test you're using is flawed and you should not draw any conclusions from it, in any direction.
Ferry Avenue wrote:Again however, change a mere two playoff games -- 35 points instead of 50 in game 2, and 25 points instead of 39 in game 6 -- and we have a significant difference in PPG. So with the very same sample size, we could've obtained statistical significance with a relatively minor change in Embiid's scoring. The sample size was hardly prohibitive here.
Holy hell, you have done this 3 times and i already told you this is precisely the point. I don't understand how to repeat it.
If two games can change the result, then what you're showing is a big variance swing, you cannot prove this two gams are a random occurence if they're 1/3 of your entire example.
You're making a point that contradicts your premise.
Ferry Avenue wrote:Surely your contention isn't that that there was a statistically significant difference in his scoring in the postseason? If it is, how can you possibly support that?
No, I contend that the test you're doing is useless and you cannot draw any conclusions from it, in favor or against. I have repeated this 4 times in different ways, i don't know if you don't understand statistics or you're being obtuse on purpose at this point.