Combined Queues - Final Discoveries
#181
Posted 25 May 2020 - 11:50 PM
#182
Posted 26 May 2020 - 12:10 AM
That's like suggesting an occasional game of soccer against Manchester U. (insert whatever sport/team you prefer to think of there, of course).
Nobody likes stomps (on either side), playing in a tier that represents your skill and gear level should be the obvious goal.
If you can climb the ranks as a weekend warrior and slap Salami with the best of them, that's awesome; More power to you.
At the end of the day though, even such talented people want the same thing - to play challenging matches, against people of similar skill (and gear/builds).
Edited by Kynesis, 26 May 2020 - 12:10 AM.
#183
Posted 26 May 2020 - 12:10 AM
AndiMech, on 25 May 2020 - 11:50 PM, said:
With that you would help all future teams you would be a part of more than sticking around, soooo...
Well that's kinda rude, isn't it? Let everyone have their opinion,no matter how bad they may be at the game. Maybe they like boating LRMs like I did back when I first started playing. Not getting many kills that way, nevertheless can be quite helpful to the team if done right.
On topic: I was looking forward to the day PSR got reset. Imho they should have implemented a reset cycle every 6 months or so from the very beginning since they knew almost everyone was bound to end up in T1 given enough time.
#184
Posted 26 May 2020 - 12:22 AM
Brightsoul, on 25 May 2020 - 04:23 PM, said:
i think the way to do this is to ensure that each tier has precisely 20% of the population. this will make the tier system a relative placement compared to other active players, so you might get some float near the thresholds without any player interaction as the others move about. but i dont see this as much of a problem and might actually be preferred behavior.
alternatively you might use a bell curve distribution, instead of flat 20% slices, with most of the players being t3. i think populations distributions for would look something like this: 2%, 14%, 68%, 14%, 2%. that way if you are in t1 you actually have to belong there.
Edited by LordNothing, 26 May 2020 - 12:25 AM.
#185
Posted 26 May 2020 - 12:28 AM
Base ranking calculations on how many people from other tiers you encounter in a match.
Make a ranking just big monthly event with prices like decals for best players.
#186
Posted 26 May 2020 - 12:29 AM
#187
Posted 26 May 2020 - 12:31 AM
#188
Posted 26 May 2020 - 12:45 AM
#189
Posted 26 May 2020 - 12:49 AM
#190
Posted 26 May 2020 - 01:10 AM
CompleteFailure, on 25 May 2020 - 06:01 PM, said:
EDIT: Resetting and making it "zero sum" is probably a step in the right direction, I guess I'd like to see a more exact definition of "zero sum."
Perhaps your PSR change should be determined by your "raw match score" / team "raw match score" in some way?
I got 50, my team average was 45, I go up (10/9 - 1) = 1/9 of a point. Doesn't matter if we won or lost.
Perhaps if you are in a group, you use the higher of the team average and the group average to determine your PSR change, or count the group scores twice.
#191
Posted 26 May 2020 - 01:12 AM
#192
Posted 26 May 2020 - 01:15 AM
In order for PSR to be efficient it must deal with a) individual player (like in Solaris) or B ) permanent team (like in Championship).
When it calculates the average performance of 12 different people grouped together by chance (picked up by...the same PSR) it meaningless.
Would it be fair to evaluate the performance of single student by the average marks of the class? Especially if the class reassembled every 15 minutes.
So it won't make any difference whether you restart PSR or not, in the end it will produce the same poor results.
Edited by drunkblackstar, 26 May 2020 - 01:16 AM.
#193
Posted 26 May 2020 - 01:16 AM
SO DO IT!!
Although the way any kind of reward system works in this game needs to be sorted. You can smash through the enemy personally but the rest of your team sucks and you get penalised for it.
Edited by Bowelhacker, 26 May 2020 - 01:18 AM.
#194
Posted 26 May 2020 - 01:23 AM
One think from me: In any multi game new players play with experienced players, even at the beginning. Whatever it gonna be new players can't be left playing only with themselves. Playing with vets is the way to learn something. Maybe some random formula which sometimes will drop few vets to newbie matches? I like the idea to mix between closest tiers, but then it can be problem with finding a match. Some prizes for top players at the end of the month is good idea imo. Also there should be some topic at the forum, like list, with best 10 or best 100 from any month of playing quick play.
#195
Posted 26 May 2020 - 01:24 AM
#196
Posted 26 May 2020 - 01:27 AM
#197
Posted 26 May 2020 - 01:32 AM
Is the PSR really a thing, that must be shown / indicated to the player, so a reset feels like losing something which was hard earned / grinded?
So how about just removing the PSR indicator? Do not show it to the players anymore, but only use it "secretly" to have a better matchmaking. So you are able to do some finetuning without upsetting the playerbase.
Of course be transparent about it, clarify how it works, note the changes in the backround, if you do some, but do not make the Tier something, which might appear as "earned" and nobody wants to lose.
So my suggestion: remove any indicator and then you are fine to reset and adjust as you wish.
Edited by Ebelon, 26 May 2020 - 01:34 AM.
#198
Posted 26 May 2020 - 01:33 AM
technopredator, on 26 May 2020 - 12:58 AM, said:
YES; MAKE IT HAPPEN, ASAP, LET'S FIX THIS GAME ONCE AND FOR ALL.
ALSO GO WITH UNREAL ENGINE AND C++, AND LEAVE CryEngine and .NET/C# Slowness; Intermediate Languages Suck!
t3 makes more sense as the best/worst players would have less distance to cover. starting at a low tier makes more sense when you bubble up, but with a symmetric system you probably want them to fan out from the center instead. the bulk of players are going to end up in t3 anyway.
im not sure the problem with pgi is that they write slow code, its that they write little code. even when they were actually doing stuff with the game they didnt write much code. im not sure that vms are really that much of a bottleneck anymore. any code that is performance critical can always be compiled to a dll and dynamically linked in the vm (for the record cryengine uses c++, c#, and lua). also i could see pgi using c++, seg faults, access violations and memory leaks oh my. ue is c++ and unreal script, but did you play mw5, lol.
#199
Posted 26 May 2020 - 01:38 AM
#200
Posted 26 May 2020 - 01:39 AM
technopredator, on 26 May 2020 - 12:58 AM, said:
Not going to happen because CryEngine licence agreement prohibits switching to a different engine. MW5 is legally a different game, so they did it there, but with MWO they can only update to the recent version of CryEngine. Classic case for reading before signing.
8 user(s) are reading this topic
0 members, 8 guests, 0 anonymous users