Subspace Hockey/Football Zone Forums

Please login or register.

Login with username, password and session length
Pages: [1] 2 3 4 ... 10
 1 
 on: Yesterday at 07:43:45 PM 
Started by Flew - Last post by idkidc1233
Let's get this thread back on topic. HZ tonight!

 2 
 on: Yesterday at 07:02:35 PM 
Started by Flew - Last post by The Boogieman
The % confidence (which should be interpreted qualitatively, it's not a meaningful computation) not being 100% indicates that the player might have appeared in the crease even if they weren't on their own screen.
IIRC, there might also be edge cases where it can actually get it wrong when the person is flying out of the cr. :/

they're accurate insofar as the client also reports exactly where he shot from (obviously since everyone sees the puck in the same place once everyone gets the new puck info)
but the problem is that because of lag you don't actually see the puck leave the ship as soon as the person fires it, thus you see the person shoot after the person enters the cr when they actually didn't or whatever. the module only knows what the person did, not what everyone saw him do. this is why i really don't like having high ping goaljudges, because that's where this problem gets worse the most easily.

Arnk has always been on the side of "you should just trust the molecule" and in fact has always kind of looked down on us for treating the molecule like it's wrong and voting against it with our own bias

Quote from: Arnk Kilo Dylie
The module confirmed that it was a non-crease shot by not reporting anything. Goaljudges have voted clean when module says 100% cr (when it accounts for reality, and extrapolating for lag,) so long as the rule of crease calling continues as it is, I don't see why 0% voted cr could never happen.
(from an old appeal)

 3 
 on: Yesterday at 05:12:16 PM 
Started by Flew - Last post by Fur of Fur
I thought it couldn't deal with rotated angled ships being in or out depending on rotation ship was at..

 4 
 on: Yesterday at 04:26:06 PM 
Started by Flew - Last post by Goldeye
So CR calls should be 100% server based? Haven't GJ'd in a long time, but I thought it reported % confidence?

It's just semantic whether the call is "cr" or "lag" -- my point is that the server is accurate about what happened on the shooter's screen, but that doesn't mean the goal was clean

The % confidence (which should be interpreted qualitatively, it's not a meaningful computation) not being 100% indicates that the player might have appeared in the crease even if they weren't on their own screen.
IIRC, there might also be edge cases where it can actually get it wrong when the person is flying out of the cr. :/

 5 
 on: Yesterday at 04:25:45 PM 
Started by Flew - Last post by Thrill HZ
94% is a pretty fuckin huge percent

 6 
 on: Yesterday at 04:22:10 PM 
Started by Flew - Last post by Kula
There was a recent blatent CR that the server was only 94% confident on. Not sure you can trust the server that cannot be 100% confident on BCR.

 7 
 on: Yesterday at 03:55:53 PM 
Started by Rage - Last post by Leafy
This still exists?

 8 
 on: Yesterday at 03:43:58 PM 
Started by Flew - Last post by zero seven
The server is accurate for whether the shooter was in the crease when he shot.

So CR calls should be 100% server based? Haven't GJ'd in a long time, but I thought it reported % confidence?

 9 
 on: Yesterday at 02:19:16 PM 
Started by Flew - Last post by Thrill HZ
I feel like brain lag is a bigger problem than server lag on some of these crease calls

 10 
 on: Yesterday at 02:17:18 PM 
Started by Rage - Last post by koroshi
Agreed. Not a fan of the posts either.

Pages: [1] 2 3 4 ... 10

Page created in 0.136 seconds with 14 queries.