11 replies to this topic

[Dread Paladin Esparza]

Coming back to the game after a long time. as i am getting back into the swing of things i have noticed a trend that bothers me and i am hoping there are some vets that can explain this.

I do pretty well solo pugging. Win about half the matches i take part in.

But when I use the LFG tool and join a group it seems the group looses every match I drop with them. Only winning the occasional match. i just made nine drops with a group and we lost every one.

i dont understand

[JC Daxion]

there is a group cue, and a solo cue.. While both use ELO to help balance the match to some degree, the solo cue's often work out a bit more balanced, and come down to about 5-6 players that really know what is going on, and newer players.

the group cue, while you are getting a group, you are going against other groups that play together much more often, and are typically on coms, and playing with group times, and mechs that compliment one another.

using looking for group, you are basically a PUG, going against organized teams

[Dread Paladin Esparza]

would it be better to get my ELO? (is that right) higher and then try LFG?

exactly what is ELO and how is it used?

[mailin]

ELO is a ranking system based on, of all things, chess. It gives you a number based on your wins and losses, and the matchmaker somehow uses this number, along with the ELO of your team mates and compares it to the ELO of your opponents. MM then predicts who will win. If the prediction is true, no one's ELO changes that much. If however the team that is supposed to win loses, their ELO goes down, while the team that won who was supposed to lose has their ELO go up.

That's my understanding of it at least. As far as I know, there is no way to determine what your current ELO is.

[Bhodi Li773]

The LFG is really a tool more for the events that require a certain number in a drop.We had one recently with the LFG 's Introduction. You needed a drop with, 2 - 10 or 12 players , with a separate reward for each drop.The matchmaker has it much easier when you are solo than when you LFG.Solo, you get 12 man groups which are matched basically on a 1 VS 1 basis.In LFG the matchmaker tries to balance mutiple groups in the same and opposing teams which is very hard to do, someone is obviuosly going to win, decidedly.I have had games where i scored over 1000( not piloting an assault) but my group still lost .Another thing with ELO is that new pl;ayers will have a low ranking but have played similar games and will outplay other new players or you may have a player on a 2nd new account that is an experienced MWO pilot. CONCLUSION : If you want a really good gaming experience then drop solo for a better matched game. If you are dropping in a group you should be in a guild/clan,on coms for a better match ,but still no guarantee. It only takes one or two players to go LEROY JENKINS to put your team at a disadvantage or have a person piloting a new mech thats unskilled to skew the matchmaker.

[Tiamat of the Sea]

Elo (it's only one capital because it's named after the man who created it- his last name was Elo) is a system that basically uses whether or not you won or lost a match against someone with a comparable Elo rating to determine whether or not you should be rated higher. It was originally designed for one-on-one competitive gaming, which means that any attempt to use it in a team game environment (such as this), especially one where you don't typically have the same team every single time you play (such as this) is kludgy at best.

The basic concept of Elo is as follows:

When you go up against an opponent, the system predicts that you will win (your Elo rating is higher than your opponent) or lose (your Elo rating is lower than your opponent). If you do as the system predicts, it assumes that your rating is accurate and you do not undergo a change in rating (or you undergo a small change in Elo rating). If you do not do as it predicts, then your rating drops (if you lost to someone with a lower rating) or rises (if you won against someone with a higher rating). This is intended to continue until you are breaking even overall- a roughly 1:1 win-loss ratio against people with a similar rating to yourself.

MWO has a number of serious breakdown points for the system that make it imperfect.

First of all, luck is a significant part of this game. The Elo system was conceived for chess, where the only luck at levels of play where ranking matters is 'does this opponent know the strategy I'm using?' Since there is considerably more luck in this game than that, the system breaks down a little.

Second of all, you are on a team. This means that your skill and the skill of your individual opponent are very much not the only factors of skill involved in a win or a loss- there are 22 other players in the match at the same time, and you're all interacting and cross-interfering in that competition of skill.

Thirdly, the Elo system as applied in this game doesn't match only players within a narrow tolerance against each other. Using a very wide-tolerance 'Elo bucket' (and one that gets wider if it cannot find enough players in the right Elo bracket), the game system assembles two teams with relatively close Elo totals. That means that you could have a team of six people at the top of your Elo bracket and the other six at the bottom. This could, theoretically, be a very big skill gap. The Elo ratings are not available for perusal nor does anyone outside PGI actually know where the ends of the brackets are, so it's impossible to say for certain.

Fourth, when the Elo system for this game rates you, it gives you four different ratings- one for each weight class of 'mech you can pilot. This means that if you are used to, for instance, close-combat high-tonnage assaults and then you buy a brand new Stalker that you kit out for long-range combat, you're dropping against the same grade of people as you do in your close-combat assaults, regardless of the fact that you're not driving a 'mech you've got efficiencies in, it's not put together to function like the other assaults you're used to piloting, and it's set up for a completely different method of function.

Fifth, the Elo system does not discriminate your rating solo from your rating in a group. This means that if you have a high rating within your bracket and your group members have a high rating within the same bracket, you can expect to have much of the rest of your team consist of low-bracket players. You can also expect that you might get bumped up into the next bracket up to fill an 'Elo hole' and wind up dropping alongside people theoretically much better than you against a whole team composed of people who are theoretically better than you. This also means that if you drop exclusively in a group for a while and thanks to coordination your team does well, when you go back to solo you could wind up facing people in a higher bracket than you're used to dropping against solo, which can screw you up royally, as their behaviors may well be entirely different, regardless of whether or not they are actually better players.

Finally, the Elo system cannot in any way account for the fact that the majority of those who drop in groups are members of mercenary or loyalist outfits or clan groups and are often expected by said groups to be competent on a competitive level, even if the 'mechs they're dropping in right now are 'nonserious' builds that aren't competitive-style.



This means that there are any number of things that could be responsible for your experience.

It's also possible that you've just had the bad luck to be dumped on the predicted losing side of a lot of matches and the Elo-based matchmaker happened to be right- this happens even when you're dropping solo at times. Since the matchmaker is completely opaque, there's no way to know what condition or aggregation of conditions is really responsible here.

[Gagis]

I blame the separation of queues. Since solo queue and group queues were separated, the group queue tends to have a higher power level, so taking newbies along with you to show them the ropes is generally a horrible idea, as that newbie will be pulled way above his level right away on the group queue.

Combine the queues!

[White Bear 84]

I blame the Hula Girl.

[Rogue Jedi]

Gagis, on 27 May 2015 - 12:14 AM, said:

I blame the separation of queues. Since solo queue and group queues were separated, the group queue tends to have a higher power level, so taking newbies along with you to show them the ropes is generally a horrible idea, as that newbie will be pulled way above his level right away on the group queue.

Combine the queues!

while I agree that on the whole group queue skill levels are higher than solo queue that is purely down to coordination, if half the team play together regularly using teamspeak to coordinate then obviously they will work together better than a bunch of individuals with little planning or coordination.

I am against merging the queues, have you ever had a group queue match where you have a bunch of 2/3/4 person groups against a 12 person group? while it is possible for the team with several small groups to win it is far from a fair fight, now just imagine if instead of several of smaller groups it was 12 individual players, they would not have a chance, it happened before that 12 man groups were dropping against solo players and it was not even remotely fair

Edited by Rogue Jedi, 27 May 2015 - 01:11 AM.

[Tim East]

Rogue Jedi, on 27 May 2015 - 01:08 AM, said:

while I agree that on the whole group queue skill levels are higher than solo queue that is purely down to coordination, if half the team play together regularly using teamspeak to coordinate then obviously they will work together better than a bunch of individuals with little planning or coordination.

I am against merging the queues, have you ever had a group queue match where you have a bunch of 2/3/4 person groups against a 12 person group? while it is possible for the team with several small groups to win it is far from a fair fight, now just imagine if instead of several of smaller groups it was 12 individual players, they would not have a chance, it happened before that 12 man groups were dropping against solo players and it was not even remotely fair

This actually used to be quite bad. I remember just sighing and resigning myself to getting beat down whenever I would see a four-man or higher on the enemy team, because that's pretty much all it was. Groups would go out and fight, and whoever had the largest group on their team would usually just roll their opposition. At least the solo queue is a lot more likely to produce a good, non-roll match for me now.

I'd rather have one broken queue and one good one, than just one broken queue.

[mailin]

Back in closed beta when I started playing there was a single queue. Pugs would often drop against a full 8 man group on comms and get absolutely rolled. It wasn't even close. Granted, we didn't have in game comms back then, but most still don't use them for whatever reason and now the groups are larger by a full 50%. There was a LOT of talk back then about people quitting if they didn't separate the queues. They will never (if PGI knows what is good for them) go back to that sad state.

[Vlad Striker]

LFG presumes high-coordinated action and tactical understanding. If your mates can not coordinate and play mech class role you lose. If someone plays beholder you lose. If some one hides you lose. And so on.