I’ve had that happen to me. I suspect it’s an artifact of networking.
Communication between computers isn’t instantaneous, so players’ games are constantly predicting what’s about to happen based on things like momentum. The server gets all the client’s data, mashes it together into a coherent picture of what actually happened, and sends that updated game state back to everyone, who’s machines then resolve to try to match the reality the server decided on.
When it comes to POTG, it would make sense to record game positions from the perspective of each client and recreate the events, rather than trying to upload/download all that data to everyone. That would make it looks slightly different to everyone watching. In your local recording of the play, your computer may be approximating the widow’s aim based on her update info + predictions, which could be slightly off from where the widow saw her own reticle. Fast precise movements would definitely exacerbate the loss of precision between both screens.
That’s just an educated guess though.
going back to what i said a few minutes ago…thats how i interpret it as well…and if youre a fast flicker this can either be good or bad…like if youre someone who often benefited from this you might want to leave this off 
There is no benefit to leaving this option off besides the fact that having it on increases CPU usage.
It only makes the position of the cursor more recent/accurate.
I missed the sentence that said it only affects primary and secondary fire.
Would’ve been weird if it affected abilities, because it would give an advantage to players who bind their abilities to mouse buttons. Still sad that abilities will still have a big input delay. I hope we get the same feature for abilities one day.
Replays, whether killcam or post-game, often don’t reflect the player’s local perspective. You’re getting the server’s perspective. Sometimes the two don’t match and there are visual glitches as well.
I’ve had one where the Play of the Game showed me appearing to do a suspicious looking ~45 degree flick in a single frame, but I know to me it looked a lot slower than that. Probably either my upload of the data in between had some latency, or the server farted.
If someone shot you in Overwatch… they actually clicked on your model, regardless of what it looked like in Killcam/Replay.
There shouldn’t be any aiming benefit to leaving it off. You cannot build up any sort of muscle memory because as a users you don’t actually know when the game executes it’s tick. So at the time you click it’s impossible to know if it was 1ms ago or 15ms ago. As stated above, the only benefit would be performance.
2 Likes
well how i interpret this is…you can overshoot your target and still land the shot because of the native poling rate if youre a very fast flicker…im saying if this is you this might hurt…
but im guessing for most people in most situations itll help
this gets added alongside surprise Baptiste hitscan buffs.
1 Like
Hello Derek,
What do you think of this?
https : //www .reddit .com/r/GamingOptimisations/comments/cuc74q/how_sensitivities_actually_work_in_overwatch/
Is anyone from the team aware of it?
Short fragment from the post:
I. Brief description: Sensitivity in Overwatch
► Overwatch’s game engine has no concept of sensitivity multipliers other than integers. What you see in the settings as 3.05 and 12.54 doesn’t actually exist to it. Instead these numbers are used to “fake” the feeling of the game not using integers by applying Interpolation. It alternates between multipliers, based on distance from the sensitivities corresponding to the closest 2 (one “higher” and one “lower”).
► A simple example would be: instead of a constant multiplier of 2.5, the game does 2 half of the time, and 3 half of the time, alternating after each consecutive mouse count.
► And it looks like this: 2 3 2 3 2 3 2 3 2 3 instead of a constant stream of 2.5 2.5 2.5 2.5 2.5 2.5 as you might expect. This successfully makes your aim feel like you are using that 2.5 multiplier - however it also brings some positive and negative effects.
► You will also find that the 2, 2.5 and 3 in this example are Multipliers - what OW actually uses. Here is a partial list of approximate sensitivities where these multipliers occur for the majority of time:
Sensitivity 0.00 0.83 1.67 2.50 3.33
Multiplier: 0 ___ 1___ 2 ___ 3 ___ 4
3 Likes
Is there any plans in the works to continue tweaking and working on console aiming?
This post is fiction. There is no interpolation in Overwatch’s aiming system. The errors in that post were explained (it was a workshop error) but I think the guy deleted them (it’s his personal subreddit that he has mod-powers in.)
1 Like
Also, for amusement’s sake, here is an example of the old 60hz mouse input causing an extremely weird-looking kill:
streamable .com/k50h3
no wonder my aim was so bad!
so does your aim being calculated at the exact time you shot or pushed your mouse button, make the shot happen at the 16/1000 tick intervals, and also where other people were at that time put there were you shot at at the same calculated time too? but show all of this at 16/1000 intervals?
so the 16tick intervals are still in teh game but, my mouse movement is going to be recorded for every 1/1000 intervals, but only display an in game image at the 16 tick rate, but the shots will register as being shot when I shot them and where too? as in if someone was in front of that shot at lets say 2/1000 interval i shot and they were in front of my gun at that time the shot would register?
if i hold my mouse too another question, like say on anne, does each shot register at its shot tick and i just keep holding the button down and shooting and it registers were my mouse was pointing at at the interval she shoots her gun at, when she shoots it?
in game earlier it ‘felt’ a tiny bit better. was running maybe 1-2% higher average accuracy. could have just been playing a bit better than average too.
I noticed this feature mentioned in the patch notes for the console versions. Was this accidentally copied from the PC notes or does this feature somehow benefit consoles?
it doesnt? even the main post says it
Subframe aiming only takes place when a mouse button is either going down or up, and only when using primary or secondary fire. Ie: Holding down primary fire for tracking type weapons will not have any changes to how it behaves.
Rly tho, they need to fix his bugs, not buff him to offset said bugs, then fix the bugs and leave the buffs leaving him in a OP State
@Derek Mulder -
So how will this effect detecting tool assisted aim ie:aimbotting/aimassist?
