File talk:SuccessRatevsEditsatRfA.png

y-axis label/units

I may be terribly confused and please correct me if I am wrong, but doesn't this graph say that, for example, a candidate with about 2400 edits has about a 0.9% success rate? Shouldn't the y-axis scale be 0, 10, 20, 30, ..., 100 to match the label Success Rate %? Katalaveno^T_C 00:46, 16 March 2007 (UTC)

I noticed the same thing, and was for a second very confused (well under 1% of RFAs are successful, regardless of edit count?). Could this be fixed? Rigadoun (talk) 19:33, 1 November 2007 (UTC)

Type of graph

Shouldn't this (or future versions of this graph) be a scatterplot? I don't think a line graph is really appropriate for this type of data. ~ Booya ^Bazooka 20:44, 18 July 2007 (UTC)

Third-party reply to both previous comments

Type question first: This type of graph is appropriate because the underlying assumption that two editors with very close edit counts will, all other things being equal, have very close probabilities of RfA success, is likely to be true in real life on Wikipedia. If a scatterplot was used, it would show masses of points along the top and bottom axes which would likely only be distinguishable around the 500-2000 edit range (where there would be almost none at the top). With all this said, what exactly the line is which is being plot is another story, and which relates to the y-axis comment.

Katalaveno's observation is correct; this is merely a typo probably. Their suggestion would work, as would removing the "%" and calling it instead "probability". I do have a question for Durin (the creator of this plot, unless I mistake). What type of smoothing or binning did you do to get the values that are actually plotted? I would suggest that it is "undersmoothed"; it in reality is likely less jumpy. With that said, I would like to ask a big favor: Do you have the original data for this (that is, the user, their EC at RfA opening, and RfA result)? I would love to have a crack at this... Baccyak4H (Yak!) 19:29, 6 August 2007 (UTC)