Combined Queues - Final Discoveries
#141
Posted 25 May 2020 - 09:14 PM
Why we cant get normal ranking system?
System with NUMBERS.
Solo Rating and Group Rating with NUMBERS.
For example:
Win +5 pts
Lose -10 pts
There is no incentive to play similar Qucik Play with TIERS.
#142
Posted 25 May 2020 - 09:16 PM
There are so many imbalances in the calculation of skill and match rewards that it isn't giving a clue at all at the moment:
1) Match reward and skill level have to be on seperate counters (if they are not already)
2) Match rewards should heavily take into consideration the contribution to match objective. At the moment it is very heavily geared to kills. You go on and capture the enemy base in assault? meh, the match rewards and match score doesn't care, you would be better off to kill things => not good
3) Match rewards should consider support Actions (more). Protecting other mechs with ECM, Narcing and Tagging enemies, especially ECM ones, AMS destroyed missiles have to count more, Scouting to give others a target (for more than a second) should be rewarded more, etc.
4) Skill level should also be influenced by the mentioned points above. It takes skill to support effectively, especially the skill to not blindly jump into action and get killed. And especially to have it rewarded to play the objective.
5) How much damage you TAKE in comparison to how much you COULD take would have to be a factor for skill as well.
6) And definitly, a main point as far as I could observe in my plays, the skill level would have to be calculated regarding the weight of the mech. In many matches I have observed how easy it is to get high damage - and thus high "skill level" with an assault mech, or at least do not go down when the match is a loss. Compared to that, it is harder with light and mediums mechs (although still doable). Its not at all rewarded to play the lighter mechs in regard to skill (and C-Bills).
At least I do not think its just me that easily can get high rewards and skill gains with Assaults and have a much harder time doing the same with light and medium mechs because they just can take less damage and usually can also dish out less.
(7) Also important regarding skill: There should be a way - and I realize this is hard to do - to determine skill and reward dependant on accuracy. I mean I get really high rewards when I just dish out high damage all over a mech, but if I shoot out just that one crucial component, for example the legs on a fast mech, the center torso from behind etc. without touching the rest, it shows up in the numbers that the game does not consider this as doing better than just damaging that mech all over and getting high damage numbers. One possible solution would be to reward damage done to the same part consecutively (by the same player). Or how much the shots stray from the center of the shots. That way doing much damage with LRMs all over the mech wouldn't be better than doing mainly centered damage with a gauss (for example)
I am sure more experienced pilots have more to say on this, for now this are my observations regarding the "skill" and reward mechanic.
#143
Posted 25 May 2020 - 09:19 PM
When the original Mechwarrior boxed games came out they were clearly meant to be fun for everyone to enjoy. This was just common sense because it was in fact the very best way to be able to aim at making a profit and proceeding on. The marketers understood the need for treating the customer well so that they would be willing to come back and spend more of their hard earned income for future products.
In fact MWO has seen a serious decrease in its player base.
We are not paid to play this game. The sole real criterion on sticking around and continuing to play it is whether or not it is fun. And that means that nearly everyone should be able to enjoy playing it or they will move on. If you are creating a game that is really only for "World Champion" players own up to it and the rest of us will move on to something else that is fun for the majority of the rest of us. If you want this game to continue you need to seriously rethink that it should be done in a way that nearly all players can enjoy the game. What has in the past been called "balancing" has in reality been nothing more than biasing for the satisfaction of the small percentage of top level elite players to the detriment of everyone else. Being consistently nothing more than cannon fodder is NOT fun. What is not fun is not worth playing.
Yes reset the tiers but make it a sliding 0-1000 meter (or some such) so that it will be easier to gauge how you are personally doing.
Fully publish the criteria, the rules, formulae and data for gauging it.
Get over the "I'm better than you so my opinion is worth something and yours is not" attitude.
A community feeling is needed not the ultra stratification in the player base that currently exists. (And this does not mean tier level).
Here is hoping the future holds a vast improvement for MWO. If not don't be surprised when it doesn't survive.
#144
Posted 25 May 2020 - 09:32 PM
#145
Posted 25 May 2020 - 09:32 PM
#146
Posted 25 May 2020 - 09:34 PM
So yeah, Tier doesnt really reflect your skill, probably more on experience and luck.
Resetting tier level will be a good idea (and hillarious )
#147
Posted 25 May 2020 - 09:35 PM
[Redacted]
#148
Posted 25 May 2020 - 09:44 PM
Paul Inouye, on 25 May 2020 - 03:41 PM, said:
[color=red]HOWEVER:[/color]
There's still an elephant in the room. PSR calculations.
Paul, I don't think that PSR reset is necessary.
In fact PGI already has all the data required to make precise calculation of every gamer skill level. You collected a lot of data (average dmg, kdr, w\l etc.), so its much better to use those stats instead of unreliable PSR method, that works good for individual game (like tennis or chess) but makes little sense in team sports.
I would strongly recommend to make matchmaking based on KDR and average dmg for the last 20 games played for each player. It would be more fair and more useful.
Maybe this thread on matchmaking will give you some insights:
https://mwomercs.com...__fromsearch__1
#149
Posted 25 May 2020 - 09:48 PM
And the circumvention of true elo and the possible 50 minute wait would be good. Sometimes even a bad match is a good thing while you wait for the good matches. I don't mind stomping and being stomped IF THE TREND IS BETTER MATCHES OVERALL.
I can stand getting bashed half the time if half the time the matches are nailbiters.
drunkblackstar, on 25 May 2020 - 09:44 PM, said:
In fact PGI already has all the data required to make precise calculation of every gamer skill level. You collected a lot of data (average dmg, kdr, w\l etc.), so its much better to use those stats instead of unreliable PSR method, that works good for individual game (like tennis or chess) but makes little sense in team sports.
I would strongly recommend to make matchmaking based on KDR and average dmg for the last 20 games played for each player. It would be more fair and more useful.
Maybe this thread on matchmaking will give you some insights:
https://mwomercs.com...__fromsearch__1
Agree with this.
Disagree with more points for objectives in psr though which another poster said a few pages back, we already have gone this way with ams. Capturing flags is actually not a test of mechwarrior skill, its a test of whether you lucked out and took a 165kph locust when the enemy has cougars.
#150
Posted 25 May 2020 - 09:50 PM
Regardless, the player base is smaller now, and the skill level of players would sift out in a relatively short period, I would think, as the ones who are still playing this are the hardcores, and a lot of players who have quit or play only occasionally would sink and fill tier 5 now. Active players would quickly fill tiers 4 and up.
Just my $0.02 CAD.
#151
Posted 25 May 2020 - 09:53 PM
#152
Posted 25 May 2020 - 09:58 PM
So their is no damage in me suggesting that the problem is how the Tier System calculates PSR, right?
So then I feel free to say that I am petty that the PSR Tier System does not appreciate my contributions to my team fight: my Warlock Builds (Summoner) had all the mad whack bells and whistles. TAG, and it was such a NARC. I had big targeting computers and a probe. I thought I made a Friggin Wizard...and I did...but for some reason, everyone forgot to bring an LRM 10; I dare not confabulate why.
All of this seemed moot, compared to the Glory of my Double AMS with even more bells and whistles: my god, it was a glorious addition to a fleeing mechs personal AMS. I felt like a hero.
But what does any of it matter if I cant just sit back, behind the enemy, and eat potato chips with the whole enemy team visible on my screen and rack up the points?
Are these not Just and Valuable contributions to the team fight?...or am I going to scrape the bottom of the barrel, and fall in Tier...Trying to help my team.
I hear that MWO has had a couple of issues that I was not here to see.
Edited by BogartTheStingy, 25 May 2020 - 10:01 PM.
#153
Posted 25 May 2020 - 10:03 PM
Paul Inouye, on 25 May 2020 - 03:41 PM, said:
I disagree. It's clear that the bias mainly comes from the formula saying that if your team WINS you move up, even if you personally do poorly. For example, if I do 200 dmg and get 0 kills, 0 kmdds, and 0 components destroyed but the team wins, I move up. Same performance and my team loses, I go down. To maintain your current rating on a loss takes around 500 damage, while on a win it only requires about 50-100. This is easily observed while playing.
My vote is to reset the PSR, and when you say "zero sum" you truly mean that win/loss has no effect on it. It should be based purely on individual performance.
#154
Posted 25 May 2020 - 10:08 PM
Tier 1 - top 1%.
Tier 2 - top 5%,
Tier 3 - top 10%,
Tier 4 - top 50%,
Tier 5 - buttom 50%
#155
Posted 25 May 2020 - 10:09 PM
Paul Inouye, on 25 May 2020 - 03:41 PM, said:
The major complaint to the PSR system is that the advancement is weighed in such a way that players who are well below global average (~225 score, they're in the ~170-180 range) in performance are able to advance unchecked to maximum of Tier 1. They most definitely aren't consistently doing well on losses and are rarely if ever doing well on wins. This may be related to the fact that after the Civil War tech update back in 2017, the average match score per player hopped up from ~190 to ~225 (https://leaderboard.isengrim.org/stats).
What actually happens is that the brackets for PSR advancement are set too low on a win. Players are often "progressing" despite doing extremely poorly while on the winning team and more due to being carried by more competent players, and we've seen empirical evidence of that. If you increase the minimum bar to advance in PSR so that a player has to actually perform well to progress (hardly a novel concept), this will be addressed.
Quote
Quote
I am not entitled to my PSR, much less my Tier. Nobody is. Anyone who truly deserves their current rank will be able to reclaim it after the reset.
-Verti-, on 25 May 2020 - 04:48 PM, said:
Edited by Horseman, 25 May 2020 - 10:10 PM.
#156
Posted 25 May 2020 - 10:11 PM
[Chancellor Palpatine face]
#157
Posted 25 May 2020 - 10:12 PM
Maybe you start different seasons (every year?) with a reset? That will give players a chance to get a higher tier if they improve.
Edited by Kat Baran, 25 May 2020 - 10:54 PM.
#158
Posted 25 May 2020 - 10:16 PM
#159
Posted 25 May 2020 - 10:20 PM
#160
Posted 25 May 2020 - 10:26 PM
Last three games in a row I played on this account he was on the enemy team which stomped my team horribly each time. 2 of the 3 were 12 - 0 stomps with his lance pretty much getting the vast majority of kills. He killed me twice with his Orion.
Needless to say I quit playing for the rest of the day as that was worse than bad.
Edit to include that those three games were all Quick Drops. I was simply pugging it. None of the people in his lance had the same unit tag but I assumed they were grouped up regardless.
Edited by Belacose, 25 May 2020 - 10:29 PM.
4 user(s) are reading this topic
0 members, 4 guests, 0 anonymous users