When it comes to MS people throw out numbers, but has anyone tested to see how different 100MS, 250MS, and 350MS are? The only thing I would agree with is if they added a max MS. Example: like 400-500 and up, just throwing out a number as I don't know at what point it starts becoming a joke.
|
Real world example ftw!
If we were to extrapolate what this guy uses from that last link:
- 175ms from the game client to the game server is considered slow
- 175ms from the game server to the game client is considered slow
- We can extrapolate that this to mean 350ms ping (i.e. Graal's known measure of lag).
I say that 400-450 is a good high limit, and kids with 56k internet service would probably get a chance to play when the server load is low. This guy implements a limit of 350ms on his server. However, this isn't a rule that means if you measure lag at 350ms, you're automatically kicked.
- the latency of 125 packets are sampled being sent from the game client to the game server
- the latency of 125 packets are sampled being sent from the game server to the game client
- If 9% of either of those 125 packets are above 175ms, the player is kicked
- If we were to simplify to round-trip-time, then if 9% of 125 ping samples are over 350ms, the player is kicked
I.e. having high lag for one ping check isn't going to get you kicked.
Onto packet loss: this is a measure of the number of packets that never reach the server or the client. Ever see someone warp from one side of the arena to the other? That's packet loss.
This guy says if 8% of your packets are lost on transmitting data from the server to the client or vice versa, you're kicked. There's also a flag for disabling weapons if packet loss is high enough, but the documentation doesn't really surmise what that means in comparison to the amount of packet loss that gets you kicked.