They’re not though. The devs have stated that he has the second smallest projectile hitbox in the game.
1 Like
What you see there is the projectile contacting the hitbox
the hitbox extends quite a bit out from the character
often times the visuals of the arrow will snap to the person it hit to avoid a floating arrow (though floating arrows still occur as the snapping only really happens with the head hitbox)
Here’s a video showing hanzo’s arrows vs widowmaker’s hitbox. It’s close, but I believe that you can see hanzo’s arrows need to be closer to get a headshot than they were in the replay linked earlier.
2 Likes
so…
you measured the replay and compared it against this hitbox test from what?
launch?
im pretty sure hanzo got a few small changes during the time between the replay and this hitbox test, like… idk…
an entire rework?
I hope you’re taking down notes on how to never end up in this situation again.
There is a reason everyone here is saying, no that’s not how it works. Because we know something you don’t. We told it to you as clearly as it can be explained and you still don’t understand.
Maybe it’s time to concede your pre-misconcieved thoughts and just start learning instead :o
Here’s a recent post showing hitbox testing https:// i.imgur. com/Z3gJnqr.jpg - you can see it lines up pretty cleanly with the video posted early.
Here’s a dev quote saying that hanzo’s hitboxes are small.
Is it possible that “smallest projectile in the game” actually means “huge log sized” because everything else is even bigger? Yeah I can’t disprove that. It’s also possible that the entire game and netcode has been changed since any of these sources, however unlikely that is.
Unless we get a blue post there’s not much more I can do to try and clarify how this works.
2 Likes
This is getting to be a chore
I know what I’m talking about
I’ve worked with people who work on netcode
While I can’t say anything about the people who have vehemently denied what I’ve been saying, I know what I’m saying
Take it how you will,
I’m not going to argue with people who post videos with devs saying exactly what I’m saying and then say that said video proves me wrong
I feel like I’m taking crazy pills
so goodnight Overwatch Forums
you are as mentally deficient as ever
Those who fall back on insults have no argument at all.
I’ve worked with people who know how the human circulatory system works. Does that make me a cardiologist?
I think I experienced the same problem last time I played Brig and when a Widowmaker was shooting at me.
I also thought I wasn’t quick enough to put my shield back up but it looks like I had the exact same problem you did.
I hope this is fixed soon… Brig’s shield is already a big joke… no need more to this.
The replay is the only thing matching the server reports 100% 
I don’t think the server vouches much, just keeps a log.
Brig lifted up shield earlier, logged.
Before the info reach window’s pc, she dinked it on her screen, logged.
If I understand what you are saying correctly then you are quite wrong.
The replay visually matches whatever client is requesting the information as it directly just uses the client information.
The server “vouches” for what is correct based on information being fed to it by each client.
Thus the reason VISUALLY the replay does not match what the SERVER has deemed the correct answer to be.
Oh wait, it stores the match logs of the last 10 matches locally in our pc? I thought it was on the server.
Think about the massive amount of resources it would require to hold 10 replays for each active player on a server.
Its logistically and monetarily suicide for Blizzard.
I try to tell people all the time who tell me “such and such character is fine in the OWL, so they must be fine on the ladder” that the OWL players are not playing the same game that we are. Playing over LAN inherently makes for fundamental differences. This here is a prime example of that.
Its not like they’re recording video, it is logs that are being reproduced in the replay player (which is actually regenerating an actual game inside the client). It can be compressed down to a very minimal amount. I believe it was from server due to the limitation of 10 - they keep game logs certainly for other purposes.
The limitation is likely for clients not being bogged with endless information.
TF2 replays once downloaded would remain on the pc until manually deleted. Causing issues down the road for anyone who did not know about them.
It is NOT regenerating anything. It is using the visual DIRECTLY from that of the client. Otherwise the replay would not be available upon match end. It would first need to be downloaded. This is not the case.
The only thing being downloaded in real time as the match is progressing is the server reports to the client. The visuals are entirely client generated and retained from the match. That is why the OP’s VISUAL does not match what the SERVER said what happened.
I just checked it, it works as I said. It is just reproducing logs.
Try viewing a replay with taskmanager open/network consumption for ow.exe and move the video some minutes ahead so it needs to buffer. You’ll notice small spikes of 1-2mbps (which says that the overall size is rlly small as well).
Replay is from server side.
And I tell you again it costs a very minimal amount of money… Its not like they send you a “big” file to predownload. It buffers like a video does, while its not a video, they’re compressed logs that decompress in small chunks and roll in the player.
So yea conclusion, the replay is from server 
Replay logs are server side. not the visuals. Otherwise the visuals would reflect the server side. Not what the client viewed at the time of play.
What is the difference between a replay log and a visual?
What makes more sense is to have the full game data as they’re written on the server and use the same visual models you use to play the game to regenerate it as a replay. If you mean these as visuals. I can’t think of anything else.
If they did this it WOULD make sense. However, evidence suggests otherwise.
If the server reported that the hit was indeed registered then when using those reports and those reports were being used to regenerate the match via the clients side then then the newly generated replay from those reports would show a visual matching that of the server.
However, it is NOT doing that. It is using the clients visuals as seen by the client (not the server) to generate what we see.