[00:06:02] <wm-bb>	 <gtisza> For one thing, IIRC IP vs logged-in user is a strong signal for ORES (sometimes even too strong - bias against IP edits was something the ORES team was worried about). IP masking will change the meaning of having an account. It wouldn't be surprising if ORES prediction quality would be negatively affected. (re @fuzheado: It would be interesting to find out whether that is true. Y <clipped message>
[00:06:04] <wm-bb>	 <gtisza> ou may be right but in my RC patrolling and AIV work, I've found the ...)