the sample size doesn't support the conclusion in a significant way, it is mathematically true, but lacks significance.
Australia goes from infinitely awful with no medals to demi-gods with one medal. That's binary.
In your "huh?" example you dropped Australia (necessarily) because had they been kept in, what I said would be true ... 20% of the population would have to win 100% the medals to win.
Most small countries get nuttin (underperform). Most large countries outperform their relative populations.
And yes, China should get a 59:1 ratio with Australia, but whether that is occurring cannot be determined with significance on the basis of 32 medals. With the 32 extant medals, China needs 100%, and that is your math.
There needs to be some 600 golds awarded or Australia needs to get a second one before Australia can be at all measured.
Yes, Australia shouldn't have one any, but several of the various Australia's should have won some. The situation is where none of 10 should individually win any, but the ten together should win 2 ... which of the 10 that happen to win the 2 not being signfiicant.
Now, were one to use all medals instead of just gold, that would start to help things a bit.
| 