<p>It doesn’t make them “the same,” it makes them “not statistically different.” Tomorrow the 780 could get and 800 and the 800 could get a 780, all due to chance.</p>
<p>All that retesting would reduce the standard of error, and produce more reliable scores.</p>
<p>It’s also possible that many of the 780’s will end up as 800’s and vice versa.</p>
<p>By definition, those with highest scores after 100 tests have done better than those with lower scores. However, I would not bet any money on predicting the final performance on 100 tests based on one score?</p>
<p>What is the point that you say this illustrates?</p>