

Constantly Losing, Starting To Get Pissed Off
#41
Posted 20 May 2013 - 08:27 AM
Sorry for having to be the one to point it out. Not trying to be a ****, just pointing out the obvious.
#42
Posted 20 May 2013 - 08:48 AM
While I have only recently returned to MWO (from taking a break for awhile), I have been in 3-4 man groups on TS most of my time. But even this past week I seemed to have incurred a larger increase in losses.
I am not sure if it's because more new people are coming on, and grouping with them (it's one of the public TS servers) or if because more people are finding Clans/Merc groups and now playing more as a unit than just pick up TS groups (so the play is at a higher level).
I think it may be a bit of both, I have seen less of the familiar faces (names) on the TS server, so they either have stopped or went somewhere else and a lot of new faces (names).
Edited by Syrkres, 20 May 2013 - 08:49 AM.
#43
Posted 20 May 2013 - 08:49 AM
#44
Posted 20 May 2013 - 08:55 AM
sharkbait689, on 20 May 2013 - 05:52 AM, said:
...
if I have to change my style of play in order to consistently win then im not going to play anymore.
If you refuse to change how you play, how can you say that you play to win?
You keep losing, but you refuse to change how you play.
Have you ever heard the definition of Insanity?
#45
Posted 20 May 2013 - 11:17 AM
Hammerfinn, on 18 May 2013 - 10:11 PM, said:
From re-reading the Devs description it works like that one if the following is true. If the group of people you are will should theoretically win but instead you lose, then your ELO drops. If you are expected to lose but win then your ELO increases. If you are expected to win and you do win, then your ELO stays the same. Same if you are expected to lose and do.
I think there are a few details I am not remembering, but the gist is if you win and were expected to win, your ELO does not change.
#46
Posted 20 May 2013 - 11:37 AM
#47
Posted 20 May 2013 - 12:03 PM
What, are you four years old? Change your tactics!
Edited by Carcass23, 20 May 2013 - 12:04 PM.
#48
Posted 20 May 2013 - 02:30 PM
ICEFANG13, on 19 May 2013 - 09:40 PM, said:
So I die, and I killed someone, and then its 7v7
You die, and don't kill someone, and then its 7v8
See why I use the 1 kill as good rule?
Getting one kill means you did at minimum 1 point of damage to one internal component on a torso to a mech.
Your now fighting 7v8.
Two minutes later its 2v6 because your absorbed with getting your kill, your buddies are trying to get theres and your team loses
See what i did there?
Assists, Damage Done, Components destroyed are all far better metrics then kills are. Intangibles like Scouting, Capping, using Tag, Spotting enemy movement are all far better metrics then getting a kill is as well.
#49
Posted 20 May 2013 - 02:34 PM
Tarl Cabot, on 20 May 2013 - 11:17 AM, said:
From re-reading the Devs description it works like that one if the following is true. If the group of people you are will should theoretically win but instead you lose, then your ELO drops. If you are expected to lose but win then your ELO increases. If you are expected to win and you do win, then your ELO stays the same. Same if you are expected to lose and do.
I think there are a few details I am not remembering, but the gist is if you win and were expected to win, your ELO does not change.
Throw in there, that when you kill a higher ranked player, your ELO goes up; and if you are killed by a lower ranked ELO your ranking drops.
DaZur, on 20 May 2013 - 05:16 AM, said:
While not perfect, at least the current Elo keeps a player in the ballpark with out convincing them they are Babe Ruth.
And my point still stands... Blaming the game for losing before doing a little self-analysis and trying to figure out "why" is far more damaging to a players progress than the deficits in the game mechanics.
Look at it this way... "IF" it was as simple as the Elo and game mechanics being broken, everyone would struggle to succeed. The reality is many, many people have consistent success in spite of the MW:O's deficits.
I see where you are coming from, and I agree...to a point. Whilst they may be buffed higher than they are ready for in respect for skill, they will have had time to get the equipment and experience to be competitive against the big boys. Yes any mech can be dangerous to another; but (for example) my atlas purpose built, and mastered, a first time pilot in his shiny new stock (or trial) mech stands no chance; no matter what he is in. I just think they should have time before they are thrown into the shark pit.
I also think rank should be shown. People have no idea where they stand at the moment, and I really think this half the problem. If someone ranked at 1300, gets owned by someone ranked 2000; then at the moment they just see it as them loosing constantly. At least if ranking is shown, then they will understand.
Carcass23, on 20 May 2013 - 12:03 PM, said:
What, are you four years old? Change your tactics!
Why bother posting this crap? OP has every right to post what he did. He made a valid, reasonable point, you just made an arrogant one.
#50
Posted 20 May 2013 - 05:40 PM
Carcass23, on 20 May 2013 - 12:03 PM, said:
What, are you four years old? Change your tactics!
Thanks that was so helpful, glad you chimed in. Everyone loves current mechwarrioronline.
Blackadder, on 20 May 2013 - 02:30 PM, said:
Getting one kill means you did at minimum 1 point of damage to one internal component on a torso to a mech.
Your now fighting 7v8.
Two minutes later its 2v6 because your absorbed with getting your kill, your buddies are trying to get theres and your team loses
See what i did there?
Assists, Damage Done, Components destroyed are all far better metrics then kills are. Intangibles like Scouting, Capping, using Tag, Spotting enemy movement are all far better metrics then getting a kill is as well.
Assists?
If you stripped 100% of armor off a mech, and left all of them alive with 1 point in the head, leg, and CT, and no one killed a single mech, you get no assists.
Damage done?
You can kill mechs pretty easily if you're good enough to damage the spots that are weakest. Say someone really got 8 cockpit shots on mechs with no cockpit armor and deals 144 damage with 8 kills. Is he a bad player because he couldn't break 200 damage?
Component destruction?
Yeah shooting arms off is the best way to kill enemy mechs, that's why everyone does it and why its how the game is won. No killing a mech through CT is always better than making it lose arms unless it has ridiculously low armor in the arms.
You know what combines the advantages and disadvantages of all of those?
Kills.
Yeah I used some extreme examples, but so did you. When I say, get a kill, I mean you kill the mech completely or deal most of the damage to it to kill it. What is hard to understand about this concept?
#51
Posted 20 May 2013 - 06:03 PM
#52
Posted 20 May 2013 - 07:04 PM
Here's a game you're not likely to win. You lost? That's okay, we'll just keep your ELO about where it is, and continue putting you in games you're likely to lose.
#53
Posted 20 May 2013 - 07:06 PM
Lokust Davion, on 20 May 2013 - 06:03 PM, said:
In the days before HSR and the BAP buff, I would agree with you, but life is going to be difficult for many light pilots now.
#54
Posted 20 May 2013 - 09:18 PM
Maybe you aren't using the best chassis for you.
I was hellbent the CTF being my best and favorite chassis then tried a dragon and did very well with it.
K/D on CTF (4xchassis) is about 1.0. W/L is 1.35.
K/D on DRG (3xchassis) is 1.5. W/L is 1.15.
with this, it means DRG is a good hunter but the CTF is a good fire support mech and wins matches.
to hell with ELO, gauge how much you contributed by looking at the match-end scores. if you are on the top 3 of your team, well done even if you lost.
it should be your new target, the top 3.
Here are some tips to improve your game.
Intelligence - a good scout greatly helps the team. Info on enemy position is very vital to success.
Positioning - i have seen 3 win vs a 5 mech group because of a good firing line, so position yourself properly.
Timing - look carefully on the enemy arrow. time yourself when they are looking at a different position for free hits.
Focus Firing - fire on easy to hit mechs like assaults. shoot lights if your loadout is anti-light weapons. ML, streaks, GR (yeah)
Learn to run away - if you are being focus fired, run and change position
Try different loadouts - PPCs are overrated, as well as poptarts. ERLL is good against poppers. lasers are instant. shoot them in the air then head for cover asap. you should have hidden before their shot arrive.
Hope your situation improves.
Edited by PrismArc, 20 May 2013 - 09:18 PM.
#55
Posted 20 May 2013 - 10:26 PM
PrismArc, on 20 May 2013 - 09:18 PM, said:
Maybe you aren't using the best chassis for you.
I was hellbent the CTF being my best and favorite chassis then tried a dragon and did very well with it.
K/D on CTF (4xchassis) is about 1.0. W/L is 1.35.
K/D on DRG (3xchassis) is 1.5. W/L is 1.15.
with this, it means DRG is a good hunter but the CTF is a good fire support mech and wins matches.
to hell with ELO, gauge how much you contributed by looking at the match-end scores. if you are on the top 3 of your team, well done even if you lost.
it should be your new target, the top 3.
Here are some tips to improve your game.
Intelligence - a good scout greatly helps the team. Info on enemy position is very vital to success.
Positioning - i have seen 3 win vs a 5 mech group because of a good firing line, so position yourself properly.
Timing - look carefully on the enemy arrow. time yourself when they are looking at a different position for free hits.
Focus Firing - fire on easy to hit mechs like assaults. shoot lights if your loadout is anti-light weapons. ML, streaks, GR (yeah)
Learn to run away - if you are being focus fired, run and change position
Try different loadouts - PPCs are overrated, as well as poptarts. ERLL is good against poppers. lasers are instant. shoot them in the air then head for cover asap. you should have hidden before their shot arrive.
Hope your situation improves.
For some reason i did really well with the Dragon as well. No idea why. Once i get enough money im going to buy one and make it exactly like the trial one.
Anyway, it seems like what needs to happen is for Pirahna to implement more options with matchmaking. When i first started the game up, i thought "Launch" meant launch the actual game, because surely, i thought, it would give more options than, well, none at all. If they let people, for instance, pit groups against other groups, or allow you to make balanced weight matches, it feels more the players fault when they lose, because theyll know the matches were more even. It doesnt seem right to just make a simple program that doesnt seem to work half the time and not implement any other matchmaking to speak of. come on, pirahna.
#56
Posted 21 May 2013 - 03:24 AM
Additionally, I seem to do better when I pug than when I'm in a premade. The guys I drop with are generally much more competent players than me and so when I play with them, I'm assuming I'm put into a much higher ELO bracket than I would normally be.
#57
Posted 21 May 2013 - 05:21 AM
It's your job to carry the weight sometimes take pride in it and the challenge of facing a better team.
#58
Posted 23 May 2013 - 10:42 AM
#59
Posted 23 May 2013 - 11:23 AM
P.S. Don't play sick - I was watching my own game play later and was embarrassed for myself.
#60
Posted 24 May 2013 - 02:33 PM
6 user(s) are reading this topic
0 members, 6 guests, 0 anonymous users