

Why Do I Get Floated To The Minor Leagues More Than I Play In My Own League?
#61
Posted 03 April 2015 - 05:22 AM
Its partly ELO/matchmaking and partly the nature of the game. When teammates die, a teams ability to mitigate fire and output damage both decrease (geometrically?).
The best player in MWO can't make up (most of the time) for horrendous mistakes by 4 players.
#63
Posted 03 April 2015 - 05:32 AM
Darian DelFord, on 02 April 2015 - 01:26 PM, said:
Well I do find it funny that only 18k have viewed the Urbie Death Blossum trailer on You Tube. Tells me alot about the population of the game.
That has to be the most stupid metric I've ever heard as it was on APRIL FOOL'S DAY so only an idiot pays any attention to anything a gaming company says that day...... *facepalm*
But the rest of your bitter spamming on this thread isn't much better in fairness.
Ultimately we all get bad runs or days and then the reverse so try to not agonise over the moments when your team seems unable to spot the wall of 100 ton assault mechs sweeping in, or the idiots not listening to voip or the mediums and lights that just rotate away from the heavies and assaults as fast as they can... it happens - get over it.
PS: If you primarily group then you really should have a large win amount - like the sad 6-10 man group who were literally farming the group queue during the last event.
#64
Posted 03 April 2015 - 06:40 AM
Alistair Winter, on 03 April 2015 - 01:14 AM, said:
Clearly, PGI has a set of parameters it can manipulate to determine how the matchmaker works. If the goal was really to create even matches, you need players with relatively equal skills. A player who started his account yesterday should not normally be playing 3 year veterans, unless those veterans are abnormally bad players. But if you start a new account for MWO and play for a week, I guarantee you will notice the names of some good players in your matches. People who you perhaps never expected to encounter with your new account.
PGI mixes low and high Elo players to ensure that they maintain a weight balance and relatively similar Elo levels on both teams, while keeping the queue time down. What the OP seems to be suggesting - which I agree with - is that the matchmaker should not be allowed to mix Elo levels so freely. It makes the matches less fun for both low and high Elo players.
<snip>
No they don't 'mix high and low Elo players to ensure they...' It's strictly a function of what happens because of the player population and the release valves on the match maker.
Elo system or any other relative skill based system will have the same problem with a match maker. When a top 10% person is online to make a perfect match it needs 23 other top 10% players to make that match. Given the numbers in MWO even if all 24 were online at the same time, they likely didn't all hit 'play' at the same time and are in different matches. The same thing goes for the bottom 10%. Or top 20%, or bottom 20%. These are edge cases, and yes for the people that are IN the edge case it's what happens all the time. But for the other 60-80% of the players who are not in the edge cases it works fine.
And I say again - this will happen with *any and every* relative skill based system. Elo isn't the problem, lack of enough players at certain skill levels is the problem.
The *only* way to ensure perfect matches for the edge cases is to set the match making time out to infinity. And we know how everyone loves waiting 15 minutes for a match that lasts 6.
#65
Posted 03 April 2015 - 06:43 AM
Sometimes you just gotta deal with it.
#66
Posted 03 April 2015 - 06:48 AM
EgoSlayer, on 03 April 2015 - 06:40 AM, said:
Since we're not privy to look at the actual numbers, we're both just guessing and basing this off of personal experiences and what PGI is telling us, I suppose.
My experience is that the matchmaker is not much better during prime time, when there's thousands of people playing, as during "off-time" for both Europeans and Americans. So I'm not buying that argument.
#67
Posted 03 April 2015 - 07:01 AM
Darian DelFord, on 02 April 2015 - 12:30 PM, said:
This is frustating the every loving **** out of me and needs to be adressed even if it means an extra 30 seconds in the MM.
It happens to everyone who solo PUGs, evidently you don't do it a lot?
Try taking command by TALKING to the team, text or VOIP, give some basic orders and see what happens. I find it's often all that's required to turn a CharlieFrank into a victory.
Of course, sometimes, nothing you do will change the coming CharlieFrank, it's just a fact in online gaming with random players.
#68
Posted 03 April 2015 - 07:03 AM
Darian DelFord, on 02 April 2015 - 12:30 PM, said:
This is frustating the every loving **** out of me and needs to be adressed even if it means an extra 30 seconds in the MM.
I know the feeling. I keep getting these Assault pilots who waddle out to do 50dmg and then get wrekt.
Man I wish I liked piloting assaults more!
Edited by mogs01gt, 03 April 2015 - 07:03 AM.
#69
Posted 03 April 2015 - 07:17 AM
Alistair Winter, on 03 April 2015 - 06:48 AM, said:
My experience is that the matchmaker is not much better during prime time, when there's thousands of people playing, as during "off-time" for both Europeans and Americans. So I'm not buying that argument.
I am not guessing, I am basing it off information given by Russ and Karl Berg through the forums and twitter about how it works and the actual numbers. And Karl stated that 'nearly all matches' have a very small elo variance in solo queue. So I guess they are just lying/misinforming us.
https://www.youtube....BRzUvc_w#t=1403
Elo isn't the only factor - there are hard limits on weight classes and game modes as well that will eliminate players of the right skill to create that perfect match so people that have unchecked game modes or are in a high % usage class are not eligible to be in the match.
#70
Posted 03 April 2015 - 07:25 AM
EgoSlayer, on 03 April 2015 - 07:17 AM, said:
I am not guessing, I am basing it off information given by Russ and Karl Berg through the forums and twitter about how it works and the actual numbers. And Karl stated that 'nearly all matches' have a very small elo variance in solo queue. So I guess they are just lying/misinforming us.
https://www.youtube....BRzUvc_w#t=1403
Elo isn't the only factor - there are hard limits on weight classes and game modes as well that will eliminate players of the right skill to create that perfect match so people that have unchecked game modes or are in a high % usage class are not eligible to be in the match.
I've seen the same stuff that you have, so yeah. That changes nothing.
#72
Posted 03 April 2015 - 07:38 AM
Darian DelFord, on 02 April 2015 - 01:26 PM, said:
Well I do find it funny that only 18k have viewed the Urbie Death Blossum trailer on You Tube. Tells me alot about the population of the game.
I don't care to see it myself. Not an ubie fan however I play a ton... So it really DOESN'T tell you a lot...
Also... maybe it is time you pickup your mic and start leading. I get the heavy roll 50/50 and that is where you hear me chat the most. You can either carry or lead... make a choice.
Edited by Creovex, 03 April 2015 - 07:40 AM.
#73
Posted 03 April 2015 - 06:32 PM
EgoSlayer, on 03 April 2015 - 07:17 AM, said:
Wrong. They have this thing called "release valves" thats lets you stack up assaults and heavies in matches leaving you with only 3 to 4 total lights+mediums for the whole team.
Thats not the only place they have "release valves".
Release valves basically eliminate those nice game features like 3/3/3/3 and elo matching. Not enough people playing (which is the norm) and release valves almost shut off MM and says ... oh well, f*** it.
#74
Posted 03 April 2015 - 08:28 PM
Alistair Winter, on 03 April 2015 - 06:48 AM, said:
My experience is that the matchmaker is not much better during prime time, when there's thousands of people playing, as during "off-time" for both Europeans and Americans. So I'm not buying that argument.
My experience is that the matchmaker is worse during prime time some days, better during prime time on other days, and pretty hopeless on challenge weekends.
Your experience I can explain away if you are average, or at least close enough to average as makes no difference. (I don't believe that's true as I've seen you play, but it's an explanation that would make your experience make sense.) If you are average then the matchmaker is always going to work about the same because most players are within tolerance of your skill level.
My experience seems to imply that most people don't play every day, so the actual mix of skill levels on any given day is different. Some days I'm punching above my Elo because the others online aren't as good, so I have a series of good games. Then when prime time rolls around and more people get online, I get punished for that series of good games (and attendant Elo increase) in order to bring my Elo rating back down to where it belongs. Other days the opposite happens. And on weekends all hell breaks loose due to the influx of new (or infrequent) players.
All of this tells me that Elo is working just fine, and that - as I've been saying all along - the problems that people perceive come from the matchmaker's other inputs, it's algorithms, and the player population at any given time, and not from our Elo ratings.
#75
Posted 03 April 2015 - 08:31 PM
WarZ, on 03 April 2015 - 06:32 PM, said:
Only partly wrong.
3/3/3/3 is the goal, but the release valves will allow an extra Mech in each weight class if necessary. The hard limit is 4, though. You'll never see 5 assaults on one team, but you might see 4 assaults, 4 heavies, 2 mediums, and 2 lights like I saw today. Or 4/4/3/1 like I saw yesterday.
Even the Elo spread has a hard limit, but last we were told it was wide enough that it should basically never have to kick in.
#76
Posted 03 April 2015 - 08:32 PM
If the top 12 players in the game decide to take weaponless, armorless, 19kph Stalkers.....should the matchmaker take this into account?
Elo cannot account for this; the system is flawed. Period.
#77
Posted 03 April 2015 - 08:45 PM
Meeso Thorny, on 03 April 2015 - 08:32 PM, said:
It could, actually, but it would require a much more complex and layered system. And that added complexity is probably why no one's bothering to respond to your question - it's essentially rhetorical. Duh. Of course the matchmaker should take things like that into account.
#78
Posted 03 April 2015 - 09:31 PM
Roadkill, on 03 April 2015 - 08:45 PM, said:
Roadkill, on 03 April 2015 - 08:28 PM, said:
Added complexity?
It's complex, so just, gg. Don't even try.
combat value of equipment * pilot skill
Dare I say, combat value of equipment is the easier variable to solve.
#79
Posted 03 April 2015 - 09:34 PM
WarZ, on 03 April 2015 - 06:32 PM, said:
Wrong. They have this thing called "release valves" thats lets you stack up assaults and heavies in matches leaving you with only 3 to 4 total lights+mediums for the whole team.
Thats not the only place they have "release valves".
Release valves basically eliminate those nice game features like 3/3/3/3 and elo matching. Not enough people playing (which is the norm) and release valves almost shut off MM and says ... oh well, f*** it.
Sorry, but I am not wrong it's you that is misinformed. There is a hard limit on no more than 4 of any weight class, the release valve on 4x3 only allows it to go up to 4.. Elo's hard limit is so high as to not matter (+/- 1000 last I know was posted which is ~70% of the entire range) and will continue to expand the range until the match is made or the match time out value (4 or 5 minutes) is reached as elo is the last release valve and thus has no upper limit.
Edited by EgoSlayer, 03 April 2015 - 10:10 PM.
#80
Posted 04 April 2015 - 12:44 AM
Alistair Winter, on 03 April 2015 - 01:14 AM, said:
You've been around more than long enough to know better. If you don't like that tone, than too damn bad.
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users