Slides presented at the 2nd Snow Workshop (http://wiki.liquidpub.org/mediawiki/index.php/Second_Workshop_on_Scientific_Knowledge_Creation%2C_Dissemination%2C_and_Evaluation)
For instance if t =1, n = 4 p1(0) = probability that the first number is different from the first in the ordered ranking p1(1) 0 probability that the first number is equal to the first in the ordered ranking
Check numbers
Notice that if we restricted the world to the top N papers and accepted only 30% of them, or 20% of them, we would have essentially a random process. The bottom right suggests that it is not true that for weaker papers we can have a clear decision We were not looking for this results Acceptance rate is over-rated
Randomness in review Due to many reasons: subjective reviews limited size of the sample for a given reviewer absolute subjective marking vs. relative ranking