183 replies to this topic

[FerretGR]

Yokaiko, on 02 April 2013 - 07:09 AM, said:

No idea, but my w/l dropped from 3:1 pre ELO ******** and is down to 1.14:1.

From the outside, I have to tell you that this appears to be Elo working. If you face competition who is at your level, Elo or no Elo, you should be challenged, ie. you shouldn't be winning 75% of your games. Having a field of evenly matched competitors should mean that you win half the time, and lose half the time. You're approaching that point, so you're approaching your proper Eo rating. My diagnosis is that you're losing a lot because your Elo is higher (based on pre-seeding, team play, etc) than it should be, and you're facing stiffer competition than you can handle, and the losses are slowly but surely bringing your Elo to its proper point.

As your level of competition increases, your win rate will inevitably decrease, since you're facing people who know what is necessary to compete with you instead of the beginners or unorganized PUGs that you're used to facing. You had a 3:1 ratio before because you were facing a random draw of players, and those players were consistently worse than you (or your team, at least). Now you're getting closer and closer to where you belong. That's a good thing.

TexAss, on 02 April 2013 - 07:27 AM, said:

If ELO really just adds the rankings together (without any borders) to get an average and then tries to find a team with same average, it's not a good design.

Some folks will swear that this is how the matchmaker is doing things, but that's not my interpretation of the command chair post that explains how matchmaking is done. If someone could provide a quote where the devs state that the team average is used in making the match, that'd be great. The quote Mischief has been posting with regard to the target number does not say this so further evidence/clarification is obviously needed.

Edited by FerretGR, 02 April 2013 - 07:40 AM.

[Corwin Vickers]

Whatever the devs say, there are some real r-tards that you sometimes get. Now that could be a good player who happened to have a head injury earlier in the day but I doubt that is likely.

And I see it happen more when I'm dropping with 3 other pretty good friends. Could be a time zone thing since I solo late Pacific and group during aussie time.

If the teams were very evenly distributed I would expect everyone on my team to have around 300-400 damage. Not me with 800 damage, and 2 guys with 300, and the rest sub 100 with a 0 or two thrown in. How do you get 0 damage with a K2? Tell me that guy isn't drooling onto his bib...


Not to say that I don't win with team mates like that because I'm pretty sure the same thing is happening on the other side. How many light mechs standing still behind an assault mech do I have to 1 shot before they figure out stopping in a light mech is a bad idea?

[Yokaiko]

This is my experience 3 out of four games.

Or water walkers on forest colony.....no ECM, no snipers, just slow *** brawlers.

Or you are capping in conquest and notice that 5 mechs are down.

Or today, two assaults four heavies two ECM lights against a Jenner two heavies and 5 mediums.

[FerretGR]

Corwin Vickers, on 02 April 2013 - 07:53 AM, said:

If the teams were very evenly distributed I would expect everyone on my team to have around 300-400 damage. Not me with 800 damage, and 2 guys with 300, and the rest sub 100 with a 0 or two thrown in. How do you get 0 damage with a K2? Tell me that guy isn't drooling onto his bib...

I came over the ridge of the Caldera just the other day and got headshot before my first volley of missiles had left the launcher. I'm not great, but I'm not bad, ie. my Elo is good enough that I've seen all of the folks complaining in this thread in games.

**** happens... sometimes folks have a bad game, and it's the nature of this game that when one person gets killed quick it increases the chances of that happening to other folks. Bad scores can and do happen to folks with good Elos, especially when they're facing real competition. That's all you're seeing IMHO.

[RussianWolf]

Olivia Maybach, on 02 April 2013 - 07:35 AM, said:

The more I think about it, the more I think the issue is with the number of players available at any given time (as others have stated). I suspect that in the early evening, with fewer online, the MM is pulling in people with a wider range of ELO, essentially giving me a higher number of NEW players to shoot at (no offense intended). That said, it SHOULD also pull in a larger number of better-than-me players in... why doesn't it seem to be doing that? I know I'm not that good... in theory the high and low should balance out.

Later in the evening, when there are more of each ELO to draw from, the games should be "tighter", which they appear to be, at least to me.

because new players start out in ELO (all caps to annoy) as average players, not bad players. If ELO is based on 1-9, then new players are all 5s.

[Iron Frost]

I get a giggle from reading people whine about Elo. It's a perfect study in human psychology. Everyone assumes they're great players and should be able to stomp 90% of the population. When a system is implemented that shows them the truth - it must be a broken system.

To everyone saying, "I've started losing because Elo keeps putting ****** players on my team": Which players exactly do you think are ending up on the other team? Does Elo have a vendetta against you personally to make sure you get stuck with all the crappy players?

The truth is to evaluate the system we would need to know exactly how the algorithm works. Here's my guess: It progressively fills up each team with the available players and with each new player or group addition tries to keep the average skill across both sides equal.

First it snags team [A] and finds that you have an average skill of 9/10. Next it tries to find a team [B] or a combination of players which have an average skill level of 9/10 and fills in the spots on the other team. If it cant find that, it will try to get as close as possible with the available pool of "launched" players. Next it moves on to the remaining spots, adding people 1/2 at a time until all slots are full. As it adds each new player, it will attempt to best balance against the all-up average of each team. If one team came up short in the previous round, it will try to find a better player for that team to rebalance the average.

Something else to consider: Depending on the number of players who are launching at any given moment, there should be a reasonable bell curve of skill levels available. If that's the case, then Elo should be able to match you up against another team with an equivalent skill mix, or at least close. When you launch within 5-10 seconds, it's highly likely that Elo was able to find a good match for you. If you're consistently waiting 30 seconds for rounds, then it's more likely that you may be getting screwed in some way.

[RussianWolf]

Corwin Vickers, on 02 April 2013 - 07:53 AM, said:

Whatever the devs say, there are some real r-tards that you sometimes get. Now that could be a good player who happened to have a head injury earlier in the day but I doubt that is likely.

And I see it happen more when I'm dropping with 3 other pretty good friends. Could be a time zone thing since I solo late Pacific and group during aussie time.

If the teams were very evenly distributed I would expect everyone on my team to have around 300-400 damage. Not me with 800 damage, and 2 guys with 300, and the rest sub 100 with a 0 or two thrown in. How do you get 0 damage with a K2? Tell me that guy isn't drooling onto his bib...


Not to say that I don't win with team mates like that because I'm pretty sure the same thing is happening on the other side. How many light mechs standing still behind an assault mech do I have to 1 shot before they figure out stopping in a light mech is a bad idea?

I've headshot K2s (and other cats) plenty of times. Not too hard. If they haven't gotten a shot off yet, then 0 dmg.

Some of the others on your team may be driving spiders and such, getting high dmg values with them is more difficult. Maybe, as has happened to me even in my Atlas. I blunder into their team while trying to gain position. I can fall to focused fire before I can deal less than 200 dmg in that case. doesn't mean I'm a bad player, just bad luck. Assuming that they are low ELO is just that.

As I've said before, ELO is a skill rating system for individuals and its being used here in a Team game. It won't be accurate for this. It can't take into account many variables that it simply doesn't have access to. (voice comm usage, Mech loadout, etc.)

But all that being said, if you really believe you are high ELO (and I'd like to see you prove it), why not drop solo for a while and see what you are really made of.

And yeah, I'm one of those that ignores yellow comms because I'm a LoneWolf and doesn't take orders.

[Twisted Power]

Caleb Lee, on 31 March 2013 - 05:19 PM, said:

as my unit is barely playing anymore.

You and half the units in the game. Mine included.

[FerretGR]

All those units who can't maintain their 90% win rate... that must be very difficult for them. I can see why they'd just give up.

[silentD11]

RussianWolf, on 02 April 2013 - 08:45 AM, said:

But all that being said, if you really believe you are high ELO (and I'd like to see you prove it), why not drop solo for a while and see what you are really made of.

Dropping solo proves nothing. Your win and loss is still a function of teamwork over all and highly influenced by what mechs you drop with and against. So you aren't proving anything. Until they make a 1v1 game mode, solo dropping will not be that reflective of your individual ELO.

Advocating solo drops is just nuts.

[Carnivoris]

I sure am glad that I'm not the only one being paired with ******** cockroaches in every match. I'm in a heated discussion on reddit about this and everyone's calling me a liar and that elo's working perfectly and they're winning half their matches while I'm losing 90% of mine. Unfortunately, I don't have the time nor the desire to get in a lance and find people to play 8 mans with. The only reason MWO is still installed is that I haven't had the chance to uninstall it yet.

Edit: Really? r e t a r d e d is censored?

Edited by Carnivoris, 02 April 2013 - 09:46 AM.

[FerretGR]

You're losing 90%? Is that an estimate or do your stats reflect this? What about your pre-Elo stats?

[Yokaiko]

FerretGR, on 02 April 2013 - 07:35 AM, said:

From the outside, I have to tell you that this appears to be Elo working. If you face competition who is at your level, Elo or no Elo, you should be challenged, ie. you shouldn't be winning 75% of your games. Having a field of evenly matched competitors should mean that you win half the time, and lose half the time. You're approaching that point, so you're approaching your proper Eo rating. My diagnosis is that you're losing a lot because your Elo is higher (based on pre-seeding, team play, etc) than it should be, and you're facing stiffer competition than you can handle, and the losses are slowly but surely bringing your Elo to its proper point.

As your level of competition increases, your win rate will inevitably decrease, since you're facing people who know what is necessary to compete with you instead of the beginners or unorganized PUGs that you're used to facing. You had a 3:1 ratio before because you were facing a random draw of players, and those players were consistently worse than you (or your team, at least). Now you're getting closer and closer to where you belong. That's a good thing.

No its not.

During the seeding period I played basically two modes, solo and 8 man. I was 66th overall despite sleeping three times and working a Sunday for the Be a Poopsoc.....err Hero tournament.

I'm fairly sure that the issue in my case is timezone, I don't play anywhere near US prime, I go to bed at 2:30pm eastern (usually).

There is no way that I inflated my Elo with seeding.

....but I have killed a number of trial mechs over the weekend.

I'm glad you have a warm and fuzzy, but after the first day or so of Elo, the game has been sucking for me. Just this morning there was a GL HF, my reply was "not likely" sure enough, there was a cap attempt the other light and I took off to cover it, killed the offending base capper.....

.....to be the only two left. 6-2, with predictable results, rinse and repeat, my drops solo are resulting in a 4 or 5 to LOSS to win ratio.

....and I'm not playing any different.

Edited by Yokaiko, 02 April 2013 - 09:59 AM.

[Danish Fury]

Pug stompers mad because the game is challenging. Cool story

[Yokaiko]

Danish Fury, on 02 April 2013 - 09:59 AM, said:

Pug stompers mad because the game is challenging. Cool story

I've always pugged more than grouped, 50/50 was about as close of a ratio as I have ever been.

[FerretGR]

Yokaiko, on 02 April 2013 - 09:59 AM, said:

my drops solo are resulting in a 4 or 5 to LOSS to win ratio.

....and I'm not playing any different.

You are the one constant in all of these games. I'm not sure why you're not open to the possibility that perhaps your Elo rating was simply too high and your losses reflect that.

Yokaiko, on 02 April 2013 - 09:59 AM, said:

I'm fairly sure that the issue in my case is timezone, I don't play anywhere near US prime, I go to bed at 2:30pm eastern (usually).

How could the timezone be responsible for you getting the scrubs and the other team getting pros over and over as you seem to think is occurring? Again, what is the mechanism by which the matchmaker always does this to you and never to the other team?

[Yokaiko]

FerretGR, on 02 April 2013 - 10:11 AM, said:

How could the timezone be responsible for you getting the scrubs and the other team getting pros over and over as you seem to think is occurring? Again, what is the mechanism by which the matchmaker always does this to you and never to the other team?

Maybe a 1/4 of the population? Maybe? Who am I playing against I see the same names, including people that I know from 6 months in the public TS servers that I always play with.

Remember the matchmaker requires that there be X people IN THE QUEUE during a whopping two minute window.

Figure the average game with Tourmaline in heavy rotation is what 8-9 minutes, more on conquest. How many people at a time do you think are queuing? Out of a population of maybe 2500 optimistically, you are talking like 120 people

Edited by Yokaiko, 02 April 2013 - 10:15 AM.

[FerretGR]

You say you're always saddled with the terrible players and that's why you're losing. I get that you're saying the population is low at the times you play. The question I'm asking isn't why "bads" end up in your game, which I'll accept could happen with a low population (average player is just as likely to get "goods" in this situation, but I digress), but why you're convinced your team is the only one that gets them. Why always your team and never the other team? By what mechanism would this behavior occur?

[Elfman]

O come on ferret stop being so aggressive in defence of Elo and blaming people for what's a sucky system. I don't rem any of us saying the other team is always uber just in fact most have said the other team made of off average to good players and they get 4 new or bad players. I consider my self an average ish player and I know and have had the honour of playing with some really good players in closed and now open beta and my Elo should not be high enough to end up when pugging in matches with them.

So how come I end up in matches either with or against them and on either side come people who can't hit the side of a barn at 90m and run off and get killed lemming style.

its either

a) Its not working how you think it is
Its having to stretch the Elo bands so wide there practically useless
c) I am a much better player than I think I am or have a much higher Elo than I should do

PGI have been mistaken quiet often with how there code is working - missile splash damage as a small example and the code for Elo is complex and they have never explained what or how there implementation works just a brief few post's on the matter.

It could be like with weight matching before it picks one team then tries to match the average Elo or in fact just being completely random

[Yokaiko]

FerretGR, on 02 April 2013 - 10:19 AM, said:

You say you're always saddled with the terrible players and that's why you're losing. I get that you're saying the population is low at the times you play. The question I'm asking isn't why "bads" end up in your game, which I'll accept could happen with a low population (average player is just as likely to get "goods" in this situation, but I digress), but why you're convinced your team is the only one that gets them. Why always your team and never the other team? By what mechanism would this behavior occur?

I'm convinced my experience ******* sucks, no more no less.

Everything else is being zerged, not having a light against 3 lights on a big map, having a 250+ ton weight mismatch on and on and on.

Christ the other day I was top damage 3 times ....in a ******* spider.

I can't tell you how many times I've been running around with another light and we killed two lights in the process of capping and THOSE WERE THE ONLY KILLS.

Over and over and over. It's ceased to be even slightly amusing.

Hell of they were even close games, I would be fine with that.

They aren't its *** stomp after *** stomp and if blows.