I am still unsure whether this is dependent on, or even related to, the "Responsive Scrolling" feature rolled out in Mavericks.

Responsive Scrolling is intended to provide smoothness of scroll by doing such things as pre-rendering the areas surrounding the scrollable buffer. This in theory promotes buttery smooth scroll animation. I am all for that and I think such an abstraction layer is essential to great user experience. In practice, however, reports of increased lag or sluggishness were common in the early days of 10.9.

However, this topic is not about that. For me, I have not seen issues of this sort, thankfully.

But what I have noticed (and which I am having trouble finding others who have noticed it) is an aggravating subtle change to the behavior of scroll. Before, in 10.8, the two finger scroll on the touchpad controlled the scrollable buffer in a one-to-one fashion. Now, for very low finger motion velocities, the scrolling motion is accelerated.

This means the page will scroll much faster (something like 5 to 20 pixels per second) even when I am barely moving my two fingers at all.

In addition to this, this acceleration appears to have a rubber-band effect, where for slightly faster finger velocities, it actually slows down the motion of the buffer. It's almost like Apple is imposing on me some sort of "optimal scroll rate" and correcting me even though I know damn well how fast I want to move through the page to match my skimming speed, thank you very much.

And then, there is a separate higher-velocity regime where if finger velocity exceeds some much higher threshold the rubber-band adjustment is taken away. This feels good and proper and the momentum scrolling works just great.

I have become a bit of a speed reader/skimmer over the years and my rate of about half a page a second is just at this threshold, between the top range of the aggravating rubber-band-slowing regime and the "free-spinning" proportional control regime.

So, basically... I don't like this. I want it to feel like my fingers are directly dragging the scrolling buffer. I don't want "help". The non-physical "help" breaks the proportional physics of control, and it basically feels like they took a perfectly tuned rolling tube that I was using to view my documents on and squirted a magic rubber banding lubricant that defies physics into the bearing.

Note: It's possible that I am experiencing some side effect due to not installing the new OS fresh and instead updating a rather heavily "broken in" 10.8 setup. I've got BetterTouchTool and other widgets of this sort that took the ride over that might possibly be interfering and generating this strange behavior.

So anyway. The actual question is.... try scrolling a page in Mavericks at this rate of speed. Maybe try to do a comparison with a Mac that's running 10.7 or 10.8. See if you can spot this difference. If not, please tell me I'm crazy and maybe that means I can get this changed back to the way it should be.


I arrived at this page because I also perceived a subtle change in the trackpad scrolling mechanics after upgrading to Mavericks. It does indeed feel that medium-speed scrolling is slowed down a bit from before (10.7 in my case). I don't have a machine running the older OS to test this on, though.

Anyway, I turned up the scrolling speed to max in System Preferences > Accessibility > Mouse & Trackpad and find the new mechanics tolerable on that setting. This is maybe not the solution you are looking for but perhaps other readers will be content as I am with this fix.

  • Honestly I got used to it in the space of a day. The redeeming factor is that slow movement makes the distance more "efficient", and even though it "drags" on you at medium velocities, well I think I tend to just avoid those velocities. My opinion of it now is that it's a perfectly usable system, it's just that I'd prefer for there to be conditional behavior which forces it never to compute a reduction in velocity. But the original behavior (which as far as i can tell, at least on the first swipe) is more faithful to reality and therefore strictly better. – Steven Lu Mar 31 '14 at 11:58

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .