Saw this post and none of the replies are really covering the more important issues right now, so here's some of the things to pay attention to in 2023 (I might have to make a dedicated thread since there's even more issues to cover really):
First section - Cables
-----------------------
1) HDMI vs displayport - displayport has slightly better frame delivery (smoothness) but responsiveness feels worse than DVI and HDMI (which both feel more similar to each other while displayport feels alien compared to the two). I think HDMI and DVI send everything on the fly while displayport uses some type of packetization system. The guy that runs the website Blurbusters knows a lot of technical stuff about monitors and says HDMI should be faster but not noticeable to human reflexes. To me it just feels like the controls are more dead in general on displayport, but it's possible I've just not found a high quality enough cable like I have for HDMI which brings us to the next subject. (oh, and you also need to have "content type reported to display" set to "desktop programs" in Nvidia control panel desktop color settings)
2) The cable itself - It's not a meme. Both cable quality and thickness actually matter. There's a pretty large difference in using a thin (30 AWG) vs thick HDMI cable (24-26 AWG). If you look at any technical literature for things like HDMI, you can tell the design specifications are running the cables practically to the breaking point. Then you have things like each internal cable bleeds interference onto the others causing the signals to arrive at different times creating skew. Whether it's just a resistance difference or skew prevention, I don't know. I just know you can feel a difference in a 26 and 30 AWG cable.
3) HDMI 2.0 vs 2.1 - might be anecdotal, might not, but I purchased a high end HDMI 2.1 cable and for some reason it felt massively worse than my 26 AWG HDMI 2.0 cable. Not sure if they're doing something weird with 2.1 cables but I tend to avoid them now. I've also purchased exotic 2.0 cables made of silver instead of copper and it made my movements all feel like they had an overshoot and I didn't like it. So things like resistance actually seem to do things in display cables while ultra-casuals or people that don't play games at all will claim it's impossible.
4) Some people will claim this is BS, but I think I've noticed a 'directionality' to how HDMI cables are built as well. As in plugging them in one direction will give different results from the other. I don't know if it's a polarity phenomenon, manufacturing issue, or other, but I've switched cables back and forth numerous times with my current cable and it seems to give better results in one specific direction. Some expensive vendors like this "Audioquest" brand even have arrows on the cables indicating which end should go to the monitor and which to the GPU. Then they also make 'one way' cables that claim to reduce interference as well. All I'm saying is that there's a lot more than meets the eye with cables and it's not a simple matter:
Monitors
----------
1) Brand - I've used a decent amount of monitors but not every brand like Gigabyte and MSI. In my experience, I noticed brands like Samsung and Dell tended to feel more normal while I didn't seem to like brands like LG at all. I remember seeing some pro player who bought like 10 monitors and came to a similar conclusion as me. There's a pretty good amount of options (some hidden, some not) in gaming monitors now so if the vendor doesn't expose the right ones or set up whatever hidden ones properly, you can see how this would be an issue. I play well with a 144hz Samsung but can't hit the side of a barn with a 240hz LG for instance. We're talking mega absurd difference in how each of the two monitors controls which just should not be the case, but it is.
2) Overdrive - Probably one of the bigger industry scams that exists. I've never used a single LCD where this did not have a detrimental effect on cursor movement. Every LCD I've used from Samsung TN panels, PLS, IPS, and my current Samsung VA all have better cursor movement with overdrive off. Might be worth having worse cursor movement and better pixel response to some people, but not me, so I've had this setting off for the last 20 something years on every monitor. And yes, I always check to see if there's a difference and it's always the exact same result with each monitor. The motion clarity difference is noticeable (not enormous, but it's there), but I've never had a monitor that was usable with it on and not usable with it off.
3) Again, another setting people will claim is voodoo, but specifically with Nvidia cards, HDMI "limited" feels a bit snappier than HDMI "full" in the control panel. Whether limited sends less data or it's just the more native HDMI spec, I don't know what the mechanical cause is, but it feels a bit snappier. Only problem is it's very difficult to use this setting since you need a monitor that supports toggling between HDMI full/limited, and most don't. Not sure how this plays out on the AMD side.
Memory latency
------------------
Your memory settings have an enormous effect on cursor movement as well as there being a large difference in single rank vs dual rank. Single rank feels faster but less in control, while dual rank cursor movement feels slower and more controlled. In the DDR3 days I didn't like how single rank felt at all and felt it was too slippery, but in DDR4 and 5 it might be more of a preference thing. Timings also play a huge role where I always prefered how 7-7-7-20 1333mhz felt over something like 9-9-9-24 1600mhz in the DDR3 era. I've never since found a setting as good as 7-7-7-20 for DDR4 or 5. Something about that was always good for me where changing it to even something like 7-7-7-21 was worse.
A real shocker (or maybe not) - Networking effects on cursor movement
-------------------------------------------------------------------------------
Some casuals will claim this is voodoo, but I promise you it's not. If you have a combination modem and wifi router, all those billion settings on the router actually affect your mouse movement. Things like "readyshare" for media sharing and all the others. If you just want a single setting to test this theory, one of the more egregious ones is "Airtime Fairness." Toggle that on and off and you'll notice an immediate effect where having it on the cursor is slower.
So what is the solution to this network idiocy for a gaming PC? You can either attempt to figure out the golden settings for your router (of which none might exist), or you can buy something like an Arris modem at Walmart (that has no options at all and no wifi) with two ethernet ports on the back and run one of those cables to your PC and run the other cable to a wifi router, I guess. The only problem here is you need TWO simultaneous IP's to do this, and Comcast will likely refuse to give you another. I think it's possible to pay a small fee to get a 2nd one if you use a Comcast business internet plan (they're not even really more expensive), but they might refuse to do it still anyway.
Intel vs AMD
--------------
It's not big news that Ryzen 1+2 are not going to be good for latency orientated people. Ryzen 3 is supposedly much better and even beats Intel in inter-core latency now, but there's too many other external factors to flat out say one would be better or worse for gaming. Each one probably uses a different Microsoft provided USB chipset driver and things like that. All it takes is somebody not doing a good job on that to ruin the entire system for gaming. I'm curious to how Ryzen 3 feels compared to Intel nowadays, but I can't really justify setting $600-1000 on fire to find out it's worse than the systems I already have.
I'm surprised there's no reputable person in the entire universe with attention to detail who can say "I play better on Intel than Ryzen 3" or vice versa. Since Intel started to implement the same chiplet and latency problems AMD had of Ryzen 1+2, I imagine this is probably an objective thing now and not preference where one is going to be less worse than the other in terms of how moving a mouse around on the screen feels.
This one will piss you off if you spent a lot of money on an aftermarket card
-----------------------------------------------------------------------------------
I used to mine digital crapcoins and had like 30 GPUs at one time. I noticed the 'founder' edition, aka direct from the manufacturer had better mouse movement than the 3rd party cards. Some brands being worse than others, but the founder edition was always superior. I think they put the A-team engineers on the founder cards while giving them a year or two to design the thing while they have the fly by night engineers on aftermarket cards with dumber goals (infinite power phases), short time span, bad BIOS's, numerous fans creating harmonic issues, etc, etc.
Power supplies and power strips
-----------------------------------
I'm not really a big PSU guy. It's kind of a weird field where components are always changing even on the same model, but anecdotally the power supplies WITHOUT caps inside the cables feel better to me. I also noticed a similar thing with power strips where for some reason the cheaper power strip with no internal 'conditioning' seemed to provide a more direct cursor movement. Sounds good, right? Having a bunch of caps all over the place "conditioning" your power. Apparently it's not, and I'm guessing having these things all over the place creates resistance issues and you basically just want only the giant capacitors in your PSU and that's it.
Web browser - a ridiculous rabbit hole
------------------------------------------
Another setting people will claim is voodoo, but it's not. If I load up an old build of Chromium which I know is good (857942), then delete it and load up a newer version, you can tell cursor movement differs a lot from the last web browser you've had open, like the OS somehow gives browsers too much leeway to fool with system integration. It's the same thing if you load up say, Windows 8.1, open Internet Explorer and turn off "Smooth scrolling." It then feels like that browser setting alters system-wide mouse movement when whatever the browser is doing should be entirely contained and sandboxed into the browser itself instead of screwing with the operating system.
If you want to test this theory yourself, just delete your Chromium settings in users/usernamehere/AppData/local folder. Now open chromium (settings and bookmarks will all be default) and do something like delete that default "chrome webstore" link on the new tab page. You should notice mouse movement instantly becomes faster after deleting that link for no reason. I have no idea why or how browsers and their settings have this type of impact on global system performance when virtually no other apps do things like this. If I change a setting in "Borderlands 3" it does not alter my system-wide mouse movement. Browsers should be the same way. If I was king I would send whoever is responsible for this to prison.
More BS in PCs that makes you want to give it up and buy a console instead
------------------------------------------------------------------------------------
If you plug in your motherboard USB headers (i.e. for front panel USB access), usually you can feel your mouse movement stiffen up some potentially being a bit less responsive. Is it the cables not being shielded acting as antennas? Is it just the cables suck and create some type of grounding issue? Don't know. Don't care. I just notice they tend to be a problem so I stopped plugging them in.
Non-deterministic mouse sensitivity
---------------------------------------
When you click the DPI button on your mouse to toggle through settings it will feel completely different each time you arrive back at the same 800 DPI or whatever you started at. I first noticed this back in the Logitech Avago 3090 days where I attempted to utilize both 400 and 800 DPI in games and noticed if I toggled between the two settings the mouse would feel off until I did a system reboot. But once rebooted that 800 DPI tended to feel the same each time. That was on mice with no internal memory.
Now, fast forward to mice with internal memory and you have the same problem but instead of the mouse feeling normal after reboot, your 800 DPI now feels different than before as if you've gained or lost sensitivity. Some type of issue with bit flipping on crappy internal memory? No idea. I just notice toggling through DPI now is not deterministic and your mouse will feel different each iteration of 'toggling' DPI.
Ferrite chokes
----------------
These seem to actually do things. I had a day one launch edition Logitech G402 for instance and for some reason it controlled a lot better than the two other later units I had. After taking them apart I noticed the original had this ferrite choke built inside the mouse itself around the USB cord while they were omitted from later units. I suspect this is why the original felt better. Companies like Endgame mice seem to utilize a generous portion of ferrite chokes as well and their mice tend to behave in a very predictable way compared to many others.
MBR vs GPT booting
-------------------------
There's a difference between the two and you'll have to try both to see which you like better. I've always preferred MBR vs GPT - UEFI on Win 8.1 for instance. Then for some reason if you turn off CSM while booting 'pure' UEFI mouse control goes to utter crap. I believe Intel is supposedly removing CSM soon as well (if they haven't already).
-> 
.