Posted: Mon Feb 21, 2022 7:01 pm Post subject: Server unfairness
I think I've finally figured out, how to predict who has the server advantage. So far I have not seen deviations from my theory, and I can pretty much tell the outcome of the duel before it even starts (taking skill and advantage into consideration).
I made a new thread because the other topic is a bit messy, and I think this one I can explain quite thoroughly with no additional mess.
Let's say 1on1 server has 7 players. You type "players" in console and you see this:
Let's now define Player1 as the guy with lower ID, Player2 as the guy with the higher ID. Now I can say that:
0: Seal is in advantage against everyone, because he is Player1 against everyone, AND the server has ODD number of players. Doesn't matter who he faces, he always is the Player1 and has super good SSG, BFG, plasma rifle and easy to dodge rockets along with inhuman reaction times.
1: Lizard is in advantage against everyone else except 0: Seal. 2: Donkey is advantage against everyone except Lizard and Seal, and so on.
Now, if 8th player connects the server, so the full list with 'players' command looks like this:
Now we have EVEN number of players. This time Player2 is in advantage against everyone, instead of Player1. This means that 0: Seal is in disadvantage against everyone all of a sudden, a HUGE contrast to the situation just before where he was in advantage. Meanwhile, 7: Mule is in advantage against everyone, because he is the Player2 against everyone (higher ID once again).
I have been observing this behaviour for a while now, and it's just too obvious when you see the SSG vs SSG fights, or how the guy in advantage seems to be able to see the guy in disadvantage super easily.
I believe there is one more thing to this whole advantage/disadvantage thing, it is not as severe as this (dis)advantage I just explained. It goes like this:
The guy who joins (not connects) first to the match, is in disadvantage. This 'join first' disadvantage is not as severe as the disadvantage you get from the console IDs, though. That's why you get 4 different outcomes:
Join first disadvantage + ID disadvantage (if you get this, you can feel slow, your mouse feels donkey shit, and dodging rockets/plasma seems to be next to impossible)
Join first disadvantage + ID advantage (somewhat fair match, but ID advantage is better than join first disadvantage)
Join second advantage + ID disadvantage (somewhat fair match, similar to previous one)
Join second advantage + ID advantage (if you get this, you feel like turbo, no joking, the game looks different)
There's a tiny exception to this: let's say you type "players" in console and it looks like this:
Even though the server has 3 players, it plays like it had 4 players. The empty slot of ID 2 is 'still in use' in some sense, or maybe it has been the Bot somehow reserving that ID slot, as I have seen 7 players in server with 6 player limit, and I believe it was the bot increasing the maximum allowed player limit.
When I see a player connect or disconnect from the server, I feel it somehow with my mouse and general movement behaviour. I've started to paid more attention to the feeling, and now I can tell just from the mouse behaviour/general feeling of the game, that someone disconnected/connected to the server _before_ there is even any message sent by the server (ie. Monkey has connected).
So in short: Player1 (lower ID) is in advantage with ODD number of players.
Player2 (higher ID) is in advantage with EVEN number of players. I don't know how to fix this.
One of the big problems here is that it is difficult to nail down, as a result it will require a lot of investigation over time and very likely a few intermediate steps in order to both potentially confirm and/or rule out certain cases.
Either way the first two possibilities are being investigated currently to either confirm or rule out any relation.
No additional information is available yet; regardless this is being worked on as possibilities arise.
Any detail that does become available will only occur after a long-round of blind tests have been performed (wherever possible) to avoid potentially prejudicing the outcome.
Leading upto 1.10.22 further research has been performed concerning various potential factors.
Certain internal adjustments have since been made (ultimately how noticeable they will be is an open question); I can only recommend continuing to play as normal at this stage.
There is also a new client opt-out variable for disabling your unlagged target calculations under lowping conditions; it can be adjusted via Player Setup in the client; there are also several threshold levels. It cannot be disabled entirely, nor will that ever be introduced.
Also a given player's choice is made available via the 'playerinfo N' console command (replace N with the ID of the given player - retrievable by executing 'players').
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum