vertical rails, aim stability

those are wild guesses

ye i think stability is the main thing. I use higher pitch instead cause my hand moves easier left/right then up/down on the mousepad.

Just ask Princeton.

Czm's m_pitch has been very low for a really long time.

I don't know of any others that used such a low m_pitch, except for maybe Rix but that's too long ago to be of relevance here.

tox also ususes m_yaw .18 but m_pitch -.014 if im not wrong .... i guess since he lifts up his arm he prefers a lower sens so there is more room for error ... would be my guess ... trying to pull down ur mouse ur much more accurate than going up i kinda understand why some ppl play inverted now ... but way too mindfuck for me to get used to it ..

So annoying that there are no mice that have prediction anymore..ive used mx500, mx518 and DA which all had prediction... i just cant get used to the new DA without pred. :(

isnt angle snapping or so correction ? im quite sure that command does the same thing

Stermy uses cg_filter_angles which technically "Alters how the player’s view camera changes in relation to aim position changes" but at low values (0.03) it acts similar to mouse prediction. Some players also use M_filter 1, but none of these cmd's act like sensor built angle snapping. There are still some good mice with predicition available, look around.

How can filter_angles have any similarity with mouse prediction? It just delays the view camera to produce a smoothing effect.

To be prediction it would have to actually keep the angle from changing unless so many counts preceded in a certain direction. Maybe this could be emulated in software, but it would have to be at a very low level and there would be a substantial risk of introducing latency imo.

Your absolutely right, technically speaking they do not work anything alike at all. Only thing that they share in common is that both mouse prediction and cg_filter_angles feel as if they have some sort of smoothing effect on your aim and thus help you not overshoot /over predict your mouse movement. It's most noticeable when your feeling twitchy or are somewhat nervous. That's the reason why almost always the people who use cg_filter_angles bind it only to hitscan weapons such as the LG or rail.

I know that certain mouse drivers allow you to enable/disable (and to what degree) the amount of angle snapping; and not just lazer based sensors such as the avago 9500, the CM spawn and a couple of other optical mice utilizing the avago a3090 sensor also allow you to enable/disable the level of angle snapping through the mouse drivers. I'm not sure if that works on the software of hardware level (firmware i guess), but i'm pretty retarded when to comes to computers. Just have a mouse fetish thing.

It works on the hardware level, the drivers just tell the hardware what level to use with sensors that support different levels.

I've never heard of prediction being implemented at the software level, although I suppose it would be possible. Like I said though latency could be a problem.

I can see how prediction could feel "smoother" since an insufficient number of counts away from the snapped line would be ignored. Not endorsing it though :-)