Combined Queues - Final Discoveries
#161
Posted 25 May 2020 - 10:26 PM
#163
Posted 25 May 2020 - 10:37 PM
Nevertheless, I concur. Implement both measures.
#164
Posted 25 May 2020 - 10:40 PM
Grim Groundbreaker, on 25 May 2020 - 05:59 PM, said:
In game theory and economic theory, a zero-sum game is a mathematical representation of a situation in which each participant's gain or loss of utility is exactly balanced by the losses or gains of the utility of the other participants. If the total gains of the participants are added up and the total losses are subtracted, they will sum to zero.
Thorqemada, on 25 May 2020 - 06:13 PM, said:
It only means you will not face Tier 1 vs Tier 1 but also Tier 1 vs Tier 3 or in a low pop situation even with a wider spread.
It will simply sort the Players better but it will not magically cast a bigger pool of Players to your convenience.
ESC 907, on 25 May 2020 - 06:34 PM, said:
Efeljay, on 25 May 2020 - 08:21 PM, said:
MWO is one of my favorite games.
If you reset the Tier system, im out.
The issue is balanced matchs.
Bad PSR formula directly leads to imbalanced matches.
That Mech, on 25 May 2020 - 09:11 PM, said:
Edited by Horseman, 25 May 2020 - 10:41 PM.
#165
Posted 25 May 2020 - 10:40 PM
some other ideas that I had in my head as I was reading this:
- It would be nice to see PSR as a number, rather than a bar, I think that would make the "xp bar" mentality go away for some players.
- someone suggested balancing the PSR on the Jarl's list. Balancing on the Jarl's list would be redundant as Jarl's is based off of the in game match score system, which the zero sum PSR would be using. That being said, the match score system has a few inherent flaws, namely lrms, which can be used to pad match score without applying much skill, and AMS, which also does not apply much skill. That however, would require weapon balancing, which would be for another time. Taking time to look at the communities input on weapon balance, combined with a zero sum PSR system, pervasive bugs (ex. heat bug) and possibly a slightly restructured match score system, would allow for high quality matches for everyone across all skill groups, and could maybe even make disgruntled veterans take a look at the game again, if done right. That being said, most of this is wishful thinking, and I and many others would be very happy if a zero sum PSR reset goes through, as that would be one of the biggest changes to enable higher quality matches for everyone.
One last thing. It sounds like Worlds 2020 won't be a thing this year. With that out of the picture, and community run leagues likely sticking to 6v6 and under leagues, will 8v8, casual or competitive, ever be a thing? 8v8 has been a very popular and perhaps iconic way to play MWO, and before "soup" queue, causal 8v8 looked like it could have resurged.
#166
Posted 25 May 2020 - 10:44 PM
As for resetting PSR, yes this should have been done about 2-3 years ago...but please calculate PSR based on actually personal performance and not the luck of being on the winning team or we will see the same "experience bar" affect with PSR again. Why? Because a win or loss depends more heavily than what the other 23 people in the match do than what you do...because doing 100 damage in a win should not be rated higher than doing 500 damage in a loss like it currently is (yes I have had many matches where I have screwed up/got left behind and only done 100ish damage but when my team won despite this my PSR went up and I have had plenty losses where I have done 400-500 damage+kills,ect and my PSR has stayed even) . If PSR is intended to measure a person's skill rating (rather than win rate) then it would be far more accurate to establish what each person's average match performance and then have their PSR go up in matches where they exceed their average regardless of wins/losses and go do down when they perform less than their average again regardless of a win or loss, if they equal their average performance (or are say within 10-50 pts of their average depending on how much latitude you want to give) their PSR would remain unchanged. How to determine a person's average? Well, the easiest way would be to look at a players average match score for as long as match score has been used, or you could average out all factors (damage, kills, assists, scouting, uav detections, ect) for say the past 0-5000+ matches if you want to be more precise. Doing something like this will give you a better idea of what a players actual skill rating is...if you continue to base a persons kill rating on how often they are able to land on the winning team PSR will continue to be nothing more than an exp bar.
#167
Posted 25 May 2020 - 10:47 PM
#168
Posted 25 May 2020 - 10:49 PM
K4I 4LL4RD LI4O, on 25 May 2020 - 10:20 PM, said:
I don't think you understand the meaning of this metric and what it is supposed to do.
It's supposed to be a dynamic skill rating... not something you can gather or keep, lol.
you play good, your rating goes up. you stop playing or play bad, you go down in rating.
#169
Posted 25 May 2020 - 10:58 PM
As for the merge, can now actually group with my mates, before we would grow old and grey waiting for a match.
#170
Posted 25 May 2020 - 10:59 PM
What does matter at this point is getting a match as fast as possible. I don't want to spend my valuable gaming time looking at a matchmaking screen.
At this point, its easy to see that the very little of the remaining players in the game are mostly composed of either top-tier or medium tier players. There are no new players. There are no extremely bad players. And no matter what you do, T1s will drop (and stomp) T5's..
Especially in groups..
So all in all.. just leave it as is..
#171
Posted 25 May 2020 - 11:08 PM
I guess inevitably the balance will be found after enough games, and as long as communication from PGI is clear and timely the player base who actually care about any of this should be OK. Lets face it, the high level Jarls List players commenting on here will rise through the tiers quickly anyway!
Goodbye Tier 1 status, hello better matches!
PS I believe I should be Tier 2-3. In my mind, the Tiers should be a pyramid structure, with the greatest numbers in the lower tiers narrowing down to a much smaller number in Tier 1 (i.e. the elite). I am not sure if this is the current structure already, but thats what I want to see.
#172
Posted 25 May 2020 - 11:10 PM
Edited by dario03, 25 May 2020 - 11:11 PM.
#173
Posted 25 May 2020 - 11:10 PM
Go ahead do it.
#175
Posted 25 May 2020 - 11:17 PM
My only concern about the process is it should have been done years ago. You have too little population left for it to actually matter now.
#176
Posted 25 May 2020 - 11:21 PM
#177
Posted 25 May 2020 - 11:22 PM
It has also been said that if there is a reset back to T4 that players will move quite quickly up the ranks if they play well. Cannot see this is any sort of problem.
#178
Posted 25 May 2020 - 11:33 PM
Anomalocaris, on 25 May 2020 - 05:32 PM, said:
This adds up to a big list of variables. Groups of different people play different, and different combinations of mechs and load-outs also play different. Ideally every group should have its own collective PSR rating, which realistically is hardly possible to calculate with any sort of math-fu. Groups should stay in group queue, period.
D U N E, on 25 May 2020 - 05:34 PM, said:
That or tier 1 becomes near impossible to get into, and that becomes the game maker tier.
Here is some food for thought on that: https://starcraft.fa...e_(StarCraft_II)
ESC 907, on 25 May 2020 - 06:34 PM, said:
Legit concern. PGI already tracks individual player damage and damage ratio per kill as well as destroyed components, so it should be trivial to award players on that basis. That way people would get rewarded for efficiency and producing results that actually matter, and even that LRM-assault could make a contribution if it at least sandblasts one mech at a time. Ditto match score inflation by assists with a single medium laser.
#179
Posted 25 May 2020 - 11:35 PM
#180
Posted 25 May 2020 - 11:41 PM
8 user(s) are reading this topic
0 members, 8 guests, 0 anonymous users