Tracer gap issue

  • This will be a short feedback post regarding gaps in the attack tracers.
    I often feel that attacks which appear to be within range of my target, and with near facehug contact, will still sometimes miss / ghost through that player. There are a few possible reasons why, including player movement speed / dodges, desync or packet loss dropping the attack. General things that can be attributed to online play.

    However the final issue I believe there might be some more control over, which is the frequency at which tracers draw. Here is an example of the kind of gaps can be found in the tracers when dragged to their maximum potential while covering large distances.

    Below is a more realistic example

    I do not know to what extent these gaps might have on the combat, they do seem rather small. At the same time however, small gaps could distinguish between a hit and a complete miss. One need only look at the flail from Chivalry to remember the facehug swings in which the tracers would barely miss in only some specific circumstances, but enough to where the whole feeling of the weapon was that it was broken.

    My query is this: Is it possible to increase the frequency at which tracers draw? Will this have adverse effects for player computer / server performance?

  • I feel like this would be too small of a difference, but I suppose it is a possibility.
    Definitely agree with a lot of my attacks feeling like they should hit and then don’t. The most notable case of this for me is the Vypress stab, but I don’t think there would be gaps for that. In any case, this is something good to look into, I never really noticed this before.

  • Ok there appears to be several issues here. One is that I also have had a lot of issues with ghost attacks even at face hug range. Because we only have one or two servers there is really no way of testing the effects of the differences of ping between two players.

    A big problem with Chiv was the large ping gaps between players plus packet loss issues. Because the netcode and the real time swing trace system was bloated, the lower ping player would suffer from ghost swings if the ping gaps were 60+ points of difference between two players. AKA Player 1- low ping 23 vs Player 2 - higher ping 85, the higher ping has the advantage due to client side hit detection and lag prediction algorithms because the lower ping player is waiting for the server to confirm Player 2s actual location and swing trace. Yes it is in MS but 60ms+ differential makes a world of difference. Now add in someone like me who has Google fiber with 1gb up/down, my position is the most up to date to the server and the most accurate. Thus in Chivalry, I am constantly getting hit through parries, ghost swings, late hits out of range and projectiles registering as a hit, a half a sec after I see them fly no where near me or I see lots of desyncs.

    With that said. that could be an issue since there are no ping limits or packet loss limits on these servers and not enough players to test it but from what I have see, Mirage for me has the same issues like chiv.

    The second issue, is that I have not seen these gaps that you are seeing in your swing trace. I will try to grab some pics. Maybe it is just that class or how your PC is rendering it. Where you see gaps I have seen overlapping boxes with no gaps. Again will look later.

    The third issue I noticed with the swingtrace in Mirage is that they are not accurate to the visual range of the weapon. I tested this by going up to a flat wall with the swingtracer on and trying to find the actual hit detection range of many of the weapons. What I found was that many weapons would penetrate the wall and not register at a hit until I got close enough so that Red tracer boxes would hit the wall with half of the weapon penetrating the wall. This is a huge issue as it looks like may of the weapon’s swing traces look like they have the exact same range as all of the colors are also the same. If they are indeed different, it is very subtle and not accurate to the visual weapon model in length. This has been one of my biggest complaints thus far but havent had the time to post up or test further. Again I will try and get this confirmed.

Log in to reply