

Tier Rank Very Unhealthy
#1
Posted 21 July 2017 - 06:47 AM
#2
Posted 21 July 2017 - 06:55 AM
#3
Posted 21 July 2017 - 07:01 AM
What to say? Perhaps the problems will go away when the harsh cold northern winter hits us again. I just feel sorry for those who live near the equator or heaven forbid southern hemisphere where things go upside down.
Bless ya peeps.
#4
Posted 21 July 2017 - 07:12 AM
Your team is composed of random individuals, but so's the enemy team, so they're in no better position. So it's fair, but there's a big random element. In the long term, if you play well, you'll be winning a bit more than losing, and go up tiers.
As far as random queue, I think you'll just have to accept that sometimes whatever you do for your team isn't going to be enough. Though on the flip side, you'll also get games where YOU made that awful mistake, but still won because of a handful of strong team mates.
Edited by jss78, 21 July 2017 - 07:13 AM.
#5
Posted 21 July 2017 - 07:30 AM
#6
Posted 21 July 2017 - 07:34 AM
#7
Posted 21 July 2017 - 07:44 AM
GuardDogg, on 21 July 2017 - 07:30 AM, said:
I've never been anything else than a tier 1 player. I remember one match in the last year where my 'raiting' went down. It's like impossible to go down in the tier ratings, and thats what is broken in this game.
One day I tried how it feel like to play in tier 5 and made a smurf account, and it was horrible. No one seem have any idea what to do. Movement, positioning, aiming, weapon loadouts all sucked balls. That said, game play on tier 1 also has gone down faster than the devs can fix bugs. Some times especially if you solo pug, the matches are just horrible. But, playing solo in a team multiplayer game is like asking for a headache.
#8
Posted 21 July 2017 - 07:58 AM
#9
Posted 21 July 2017 - 08:29 AM
GuardDogg, on 21 July 2017 - 06:47 AM, said:
This game requires a tactical mind to play right unlike many shooters thus mistakes are hyphened. But i will guarantee this, if you are far better than everyone else in your drop, you will rise in rank regardless. When i was in T3, every third game i did 1k+ dmg with IS mechs, then i got in T2 which has far more players in it then all the other combined(considering how long it took to get to T1) slowly but surely i get to T1 despite not performing as well because i still have positive stats.
If you cannot rise in rank mate, im sorry to break it to you but you deserve your rank especially at T3. I miss that rank most of all, games were so carefree and fun but doesnt last long unfortunately, when you hit T2 then you see a big difference in mentality. T1 is essentially same as T2, its just you see a little less mistakes.
Edited by Lordhammer, 21 July 2017 - 09:28 AM.
#10
Posted 21 July 2017 - 08:33 AM
#11
Posted 21 July 2017 - 08:39 AM
Edited by Natred, 21 July 2017 - 08:41 AM.
#12
Posted 21 July 2017 - 08:45 AM
Methanoid, on 21 July 2017 - 08:33 AM, said:
In solo QP I guess MM tries to match tiers something like this:
tier 5 matches have tiers 5-3
tier 4 matches have tiers 5-2
tier 3 matches have them all
tier 2 matches have tiers 4-1
tier 1 matches have tiers 3-1
It doesnt always feel like that though.
#13
Posted 21 July 2017 - 08:46 AM
There are always going to be a bunch of people who game the system in order to have their 'e-peen' be bigger than other people... just look at the "I made it to rank 1 in 3 days (etc.)" posts. Really the ranking system is pretty easy to 'game' if you know how scoring is done and have the time to play tons of matches.
IMHO PGI purposely rewards people who are "Play-time Whales".
#14
Posted 21 July 2017 - 10:00 AM
Ssamout, on 21 July 2017 - 08:45 AM, said:
tier 5 matches have tiers 5-3
tier 4 matches have tiers 5-2
tier 3 matches have them all
tier 2 matches have tiers 4-1
tier 1 matches have tiers 3-1
It doesnt always feel like that though.
Thats how its supposed to work but with trash in the upper tiers and low population the MM often opens all tiers so tier rating just doesn't matter.
#15
Posted 21 July 2017 - 10:02 AM
It takes no more match skill to advance in T1 than it does in T5, only more games played to shove that little exp bar forwards.
It can be done FASTER by good players, but only total tubers (ie, you fail at basic giant robot controls) will remain in the bottom of the tiering system.
That there is no hard prevention of a T4 ending up in a game with T1's seals the deal. Making seal clubbing incredibly easy and relentlessly dropping newbie chum in the shark tank is one of MWO's best tools for eliminating new players quickly and ruthlessly.
#16
Posted 21 July 2017 - 10:07 AM
I just maxed out the experience bar by doing mostly decent.
#17
Posted 21 July 2017 - 10:24 AM
#18
Posted 21 July 2017 - 10:29 AM
Christophe Ivanov, on 21 July 2017 - 10:24 AM, said:
PGI just needs to stop being lazy and actually use ALL the pilot stats that they have to create a true metric of pilots skill. Damage done, kdr, accuracy, wins, friendly fire etc. etc.-- there is so much data available that you could create a pretty accurate picture of a pilot's skill.
#19
Posted 21 July 2017 - 10:31 AM
nehebkau, on 21 July 2017 - 10:29 AM, said:
It's not laziness. It's just not a priority right now.
And what you're asking for is not a simple task. They have access to that data, but what do you do with it? Which stats are important to consider and which are not? Are wins more important than kills? How much more important? How much should raw damage factor in? How much higher should a solo kill be rated than a regular kill?
What about damage taken? Surely a player who takes a lot of damage bedore dying is better than one who doesn't. Should that be considered?
What about efficiency? It's certainly more time-efficient to core out an enemy ASAP, but disarming him first lets you kill him while taking less damage. How should efficiency factor in?
Construcing a player rating system is not a simple task. Heck, they've gone through 2 iterations of it already, one of which already takes a lot of the above into account.
Edited by Kaeb Odellas, 21 July 2017 - 10:42 AM.
#20
Posted 21 July 2017 - 10:33 AM
Edited by Natred, 21 July 2017 - 11:14 AM.
10 user(s) are reading this topic
0 members, 10 guests, 0 anonymous users