multiplication

The passage around this formula

​ Rating deviation $\mathrm{RD}$ , bounded to Showdown’s 25–130 range 95% interval: X $\pm$ 1.96 $\cdot$ $\mathrm{SEM}$ 95% interval: R $\pm$ 2 $\cdot$ $\mathrm{RD}$ [ 3 ] Reliability rises with more independent items RD falls with more games played in a period [ 2 ] Reliability falls with time since calibration RD rises with time since the player’s last game [ 3 ] This is the article’s original contribution: not the observation that a rating has “uncertainty,” which anyone who has seen a percentage sign after a GXE could guess, but the derivation that Showdown’s published RD update schedule is doing formally, on Glickman’s own stated terms, exactly the job Novick’s 1966 axioms assign to a standard error…

Read this part in the article →