Ties in spars don't seem to really be working that well.
 

I see a lot of the time where people spar, and the spar is clearly a tie but it seems to always just give the spar win to someone? I very rarely see ties. Mostly just tie hit spars ending with whoever has the longest delay getting awarded the win (aka the idev player in most cases)

I'm not sure if there would be a way to fix the Tie issues. I'm not a fan of sparring but it does seem like there are lots of flaws accompanied by an endless list of complaints about the sparring system. Delaying,lagging,ect. As far as I know there is no stat for it but I think there should be. If you get points or whatever for a win or loss why not a tie? Some major sports factor in ties in the points standings at the end of their seasons.

I don't have a lot experience in sparring but If I was you I would ask Mr. Thallen he is a spar master and expert I'm sure he can learn you some spar insight.

All I'm saying is that a lot of ties happen, where two people slash at the same time and the spar is not ended in a tie the win is awarded to someone. That doesn't make any sense. You seem to have over-analyzed it.

Perhaps they could rescript the hit detection so that if you both swing at the same time and both of your hit detectors go off, it could possibly not grant either of you the actual hit, therefore getting rid of ties and giving off the impression you 'blocked' eachother's swordswing.

It'd be much easier to determine whether both players died in a set amount of time. (I.e. 0.1 or 0.2 seconds) apart and use that to determine to grant a tie rather than a win.

Your suggestion could put a lot of unnecessary checks during a spar, and there could be increased delays as a result.

When I used to spar a lot, ties always ended in losses. What sucks is ties happened a lot for me so you can imagine how annoying that would be.

Your suggestion does the very same thing, causing the server's reaction to each sword input to be delayed in order to check with surrounding rendered players. It sounds like a load of strain on the server as well as an annoyance during a quick instance like a fast-moving spar.

Tyler is right, ties used to be marked down as losses on the record anyway. Even so, it doesn't make sense that a draw clear to the human eye should reward a player with higher latency because the signal takes longer to travel to his/her client telling the player to display the death gani. Which triggers a removal from the arena, and a loss in spars.

So Zid, when your client promptly responds to the server's direction in a spar, your opponent may not display those same actions as quickly, resulting in a delay between the moment a sword connects with a target and the moment the server accepts and reacts to the information the clients have provided. If you are also slashed at the same time as your opponent, your client will receive and send information back to the server quicker than your opponent's. So in lines of code, the server will read the opponent's death information after yours and award him/her with a win instead of the tie every human witness saw. It can get annoying if you know what you're doing and still get a loss. The only true remedy is to deal with all latency carefully and accordingly during a spar, regardless of its severity.

No it doesn't, what are you talking about?

His suggestion could potentially be very simple(and what came to mind for me as well). If the spar is considered at an end(ie you die), if the other player is also swinging their sword, wait for a moment(.1/.2 seconds) and if you die as well, consider it a draw.

I may be a mixing up the TCP vs UDP functions and purposes with this since I recently explained that somewhere else as well... I apologize.

I would find it very annoying if there were a constant delay in reaction time every time I hit someone with a sword, regardless of the latency for both players. Thinking quickly and timing my swords and movements, this could have a larger effect on some players than one might think. Granted, it definitely would effect some players (like myself) more than others.

However, I think this type of system could work if the default server latency was dynamic based on the highest latency of a competitor in the spar arena. This way both players experience the same visual delay for each sword slash, and the server would be impartial in terms of relative latency. I'm not sure exactly how the spars are scripted, but based off of roughly 40,000 spars I've fought on this server, I have a good understanding of the behavior. I think it may be a bit difficult to assign such a server value to each individual set of sparrers, including a few different spars on the server at a time. Not to mention spectators.

Which points back to the original static server latency number, which would be annoying to deal with from a couple angles.


I don't think anything needs to be done..

I have noticed this but I always thought it was just caused by an individual players lag. I thought it looks like you swung at the same time, but due to their lag they actually hit you 1 second earlier getting them the win. idk

Simplified:

Reece's suggestion: (assuming one match, 2 competitors) potentially tens if not hundreds of checks for sword slashing even if just random openness..

My suggestion: 1 check initiated upon an initial players death for x amount of seconds (or a few tenths of a second).

As Dusty noted, its feasible and simple.

pls fix this s0 i kan keep my 80-2 ratio

Yeah I definitely agree with you all that it's probably a latency thing, but it's also a problem that never occurred on the PC client v4-6 of graal for me, which is probably why it gets on my nerves so badly xD

That idea seems very logical.

Don't know what any of this means :D

just win every spar