129 replies to this topic

[Hauser]

Sug, on 07 August 2013 - 08:18 AM, said:

The crazy variance in teammate skill we're seeing is because every new player is assigned an Elo of 1300, the average Elo among players. Because of how Elo is calculated, if you have close to a 50/50 win ratio (as all pugs should), your Elo is most likely not very far from 1300.

Actually they start at 1100.

http://mwomercs.com/...23-21-may-2013/

[travelbug]

For me the simple solution is to give us the option of 8 or 12

[Bongfu]

If you want to play a single player game, play a single player game. That or play Call of Duty.

This is a battlemech simulation game. In lore, just as in real life, a military unit operates as a team. So form a group with your buddies or join one of the existing units out there. It is really as simple as that. I fail to see how anyone could enjoy this game solo.

[Liquid Leopard]

I joined a team for when I want to take a break from the PUG life, but I've found that I get lower damage scores and kill rates when I'm playing with certain people. One possible reason: I'm not that good a player, and playing with friends can pull me up into matches based on their Elo range: I'm literally out of my league. ....AND I don't feel like I'm pulling my weight. So, what's the point, then? I often feel more successful playing by myself, with random PUGers. It's a really good day when I'm playing with someone I like and we're winning matches. Now that 12-v-12 is in effect, I'll reserve judgment for a week or two.

Koniving, on 07 August 2013 - 07:16 AM, said:

Actually, that's exactly how it happens in group matches: It snags teams and pugs of lower and higher skill to match an average. (That's actually how the system is described, too).

This also happens when the player count dropping at that moment is low. I sometimes find myself balancing a team of players that barely seem to know what they are doing, against a team that has okay to poor skill with one enemy that has superior skill to my own. Other times I'll find myself as the weak link on a team of cheese builds who each manage to get more than 500 damage.

This is exactly what I've observed, over and over again. It's the norm, not an anomaly. If I see the same players for several consecutive matches, I usually see the same ones consistently scoring well. If the Elo system was putting me with players of my own skill level, I should never have been in a pick-up game with Koniving, ZeProme, PFC Wodan, or a lot of other people I recognize who are actually good at this game. The first time I played with Koniving, my Elo score should have reflected the fact that I could hit the broad side of a barn if I walked into the barn.

[Hauser]

Koniving, on 07 August 2013 - 07:16 AM, said:

Padic, on 07 August 2013 - 07:11 AM, said:

I am pretty sure that this is not how Elo works. It is trying to match players of a similar Elo together, not match good players with bad players to meet an "average Elo" target.

Why would they ever code it to act that way? Just to troll you, specifically?

Actually, that's exactly how it happens in group matches, it snags teams and bugs of lower and higher skill to match an average. (That's actually how the system's described, too).

I'm sorry Koniving, but you are mistaken. You are thinking of the way a pre-mades Elo is calculated. For the purpose of being put into a match the members of the group are treated as having an average Elo value of the group.

http://mwomercs.com/...-making-update/

Quote

How does the match maker compose a teams Elo rating, is it average rating or closest to a target?

It's closest to a target value, so the match maker starts trying to make a match for an Elo of say 1300 and will pull in players to those teams closest to those values; however, as mentioned earlier within growing thresholds and those curves will be tuned. Currently it may be a bit 'sloppy' about how it's filling those buckets but over time it will be tuned to be much more precise.

We need to do this carefully over time as generally the cost of precision is time to find a match we want to slowly find a very nice balance between time to find a match and the number of matches that are correctly composed.

Koniving, on 07 August 2013 - 07:16 AM, said:

This also happens when the player count dropping at that moment is low. I sometimes find myself balancing a team of players that barely seem to know what they are doing, against a team that has okay to poor skill with one enemy that has superior skill to my own. Other times I'll find myself as the weak link on a team of cheese builds who each manage to get more than 500 damage.

I think you're over exaggerating the good players. Although you can end up with new players when they need some one to fill out their team.

Edited by Hauser, 07 August 2013 - 10:44 AM.

[Pliskkenn D]

I didn't choose the PUG life. The PUG life chose me.

Because I have no friends.

[Koniving]

Hauser, on 07 August 2013 - 10:40 AM, said:

I think you're over exaggerating the good players. Although you can end up with new players when they need someone to fill out their team.

Actually I don't think my goal was to exaggerate their skill, it was that ELO emphasized their weapons as meaning they had better skill.

But to what you said earlier, if ELO is comparing each player to each player... You and you match up, one here, one there. That's 1 to 1, for 12 to 12...
Does it not, then, have one player comparable to one player... for 24 players, and thus an overall average for both teams being approximately similar, just like it does for groups? Think about that for a minute. It then becomes an average for one team, against an average for the other team, exactly as it does for groups.

Edited by Koniving, 07 August 2013 - 10:59 AM.

[Sasha Volkova]

Padic, on 07 August 2013 - 07:11 AM, said:

I am pretty sure that this is not how Elo works. It is trying to match players of a similar Elo together, not match good players with bad players to meet an "average Elo" target.

Why would they ever code it to act that way? Just to troll you, specifically?

I am afraid you are in for a hard awakening then friend.
And it seems you never read the initial thread on ELO in this game.
ELO in MWO works by taking the TOTAL ELO of one side and dividing that with the amount of players, then matching it vs the other side which also has its TOTAL ELO devided by the amount of players on that side to reach the same AVERAGE ELO on both sides.

What this means is that if one side has 4 great players and 4 bad players, then the other side MUST consist of atleast 8 average players to reach the same average ELO on both sides.

If you still dont believe me then I suggest you seach for the official ELO thread here on the forum.
In there you will find EXACTLY what I just told you here.

The reason behind why it works like this was excused with the fact that this was supposed to make the average game more enjoyable.

Again I suggest you go look up the official thread were everything is explained if you simply dont believe me.

Edited by 0okami, 07 August 2013 - 11:02 AM.

[Sasha Volkova]

Pliskkenn D, on 07 August 2013 - 10:42 AM, said:

I didn't choose the PUG life. The PUG life chose me.

Because I have no friends.

You like the drakes :3 I can be your friend <3
<<INTERNET ROMANCE COMMENCE!!!>>

travelbug, on 07 August 2013 - 09:57 AM, said:

For me the simple solution is to give us the option of 8 or 12

yea great idea, I think we should listen to this bloke here.
Except for the one fact that this would split the que AGAIN.
Honestly this game cant even handle having the que being split when 3rd person arrives, so if we further split those 2 ques into 4 ques (2*2) then we might end up having 5-10 minutes ques for just one game xD
GmotherfernG

[Bongfu]

Simple solution is just join a unit. If you dont want to do that, I stand by my statement of "go play CoD."

[Padic]

0okami, on 07 August 2013 - 11:00 AM, said:

I am afraid you are in for a hard awakening then friend.
And it seems you never read the initial thread on ELO in this game.
ELO in MWO works by taking the TOTAL ELO of one side and dividing that with the amount of players, then matching it vs the other side which also has its TOTAL ELO devided by the amount of players on that side to reach the same AVERAGE ELO on both sides.


What this means is that if one side has 4 great players and 4 bad players, then the other side MUST consist of atleast 8 average players to reach the same average ELO on both sides.

If you still dont believe me then I suggest you seach for the official ELO thread here on the forum.
In there you will find EXACTLY what I just told you here.

The reason behind why it works like this was excused with the fact that this was supposed to make the average game more enjoyable.

Again I suggest you go look up the official thread were everything is explained if you simply dont believe me.

Matthew Craig said:

How does the match maker compose a teams Elo rating, is it average rating or closest to a target?

It's closest to a target value, so the match maker starts trying to make a match for an Elo of say 1300 and will pull in players to those teams closest to those values; however, as mentioned earlier within growing thresholds and those curves will be tuned. Currently it may be a bit 'sloppy' about how it's filling those buckets but over time it will be tuned to be much more precise.

We need to do this carefully over time as generally the cost of precision is time to find a match we want to slowly find a very nice balance between time to find a match and the number of matches that are correctly composed.

Here, I believe, is the quote you mention. From how I read it, it disagrees with you on every point.

Step 1. The matchmaker picks a target Elo.
Step 2. The matchmaker tries to pull in players close to that Elo. (Note: Not players who average to that Elo).
Step 3. The matchmaker loosens up its restrictions if enough players aren't found in a specific time period, until the match is filled.

It never mentions averaging.

The only time I've seen the devs mention averaging with regards to matchmaking is when there are premade teams involved. In that case, each member on that team is treated as having an Elo equal to that of their average Elo. Which is, you know, whatever. If you're in a group and you're dropping with a noob, how can you possibly complain about being grouped with a noob?

Is there a different quote your getting your averaging info from? Just because I don't see it doesn't mean that it doesn't exist.

[Yankee77]

Short answer: no. Did all my matches yesterday while pugging, and had a good experience all around.

Long answer: It can punish you more if you screw up, but ultimately this works for both teams. The game is not inherently harder, nor is the other team magically less vulnerable to pug issues than yours is. It's ultimately balanced.

The main difference is that the "metagame" has switched a little. Lonewolfing is even more dangerous, and capping is more prevalent. If your team members haven't adjusted to that, you will lose.

Most of my matches last night were losses, mostly because the teams I went up against had more lights, or their lights/fast mediums understood how capping was more important now (especially in conquest, where all my matches hit 750 ressources). We won the brawls, but lost the match.

Once everyone has grasped the implications of 12v12, things will stabilize and pugged won't be any less viable than it was before.

[Tesunie]

I PUG. I do just fine. One mech was never suppose to "always" change the tide of battle. If anything, I have found 12v12 to make it easier to find a group heading in a similar direction as you for you to stay together with. You can also use Lance Chat to talk to your lancemates. I had a couple drops where my lance rolled together, and it was a lot of fun. Just us PUGs working together.

Even then, I feel 12V12 has made the game feel better as you play. It's more exciting. Anything can happen. Even a random group has a better chance now I feel than before. Communication is still key.

Look at it on the other side, if one mech suicide charges the enemy and dies, it doesn't hurt as much as it did on 8v8, as less of your team is dieing there. If one person DC now, it's no longer just about a given loss for you. Tactics are more vital, and you have more mechs to be able to spread out more for a net when you have to hunt down the last few lights, or to protect your base.

12v12 has it's goods and it's bads. Take each as they come. However, saying that "PUGing is dead" is a gross exaggeration. It's about as bad of an exaggeration as those who post that "this game is already dead, let it die". PUGing is not dead, it has just changed. Adjust or find a team. Can't give any other advise than that.

[Sug]

Hauser, on 07 August 2013 - 09:57 AM, said:

Actually they start at 1100.

http://mwomercs.com/...23-21-may-2013/

And then they get a standard player Elo that might be a 100 points higher than they should be at for no reason.

[Sasha Volkova]

Padic, on 07 August 2013 - 12:00 PM, said:

Here, I believe, is the quote you mention. From how I read it, it disagrees with you on every point.

Step 1. The matchmaker picks a target Elo.
Step 2. The matchmaker tries to pull in players close to that Elo. (Note: Not players who average to that Elo).
Step 3. The matchmaker loosens up its restrictions if enough players aren't found in a specific time period, until the match is filled.

It never mentions averaging.

The only time I've seen the devs mention averaging with regards to matchmaking is when there are premade teams involved. In that case, each member on that team is treated as having an Elo equal to that of their average Elo. Which is, you know, whatever. If you're in a group and you're dropping with a noob, how can you possibly complain about being grouped with a noob?

Is there a different quote your getting your averaging info from? Just because I don't see it doesn't mean that it doesn't exist.

nooooooooooooope
Wrong quote.
Get the ones from Russ, Poul or Bryan.
One of those did the initial ELO post.
In that it is clearly explained how things work.
Nuff said.
Now go have fun looking for that post or stay hidden inside your own little bubble of safety.
I dont really care either way :b

[WVAnonymous]

Just waiting for my Elo to finish falling, not doing well in 12 v 12.

I figure after losing 6 in a row last night, may not take too long.

[Sug]

WVAnonymous, on 07 August 2013 - 02:31 PM, said:

Just waiting for my Elo to finish falling, not doing well in 12 v 12.

I figure after losing 6 in a row last night, may not take too long.

Man i'm 3 for 42 since the patch and still losing....


Also, if and when CW drops aren't we going to choose who we're dropping/defending against anyways? So what is the point of the Elo system once the game isn't Randomdeathmatchwarrior Online?

Edited by Sug, 07 August 2013 - 03:27 PM.

[Hauser]

Sug, on 07 August 2013 - 12:56 PM, said:

And then they get a standard player Elo that might be a 100 points higher than they should be at for no reason.

Nah. Both N-Elo and S-Elo are changed as a result of winning. Because S-Elo is much higher then the rating a new player should be on it will fall quickly.

Koniving, on 07 August 2013 - 10:55 AM, said:

Actually I don't think my goal was to exaggerate their skill, it was that ELO emphasized their weapons as meaning they had better skill.

You mentioned 500 damage p.p. that is a bit much.

Koniving, on 07 August 2013 - 10:55 AM, said:

But to what you said earlier, if ELO is comparing each player to each player... You and you match up, one here, one there. That's 1 to 1, for 12 to 12...
Does it not, then, have one player comparable to one player... for 24 players, and thus an overall average for both teams being approximately similar, just like it does for groups? Think about that for a minute. It then becomes an average for one team, against an average for the other team, exactly as it does for groups.

To assemble a team the match maker tries to find people near a given value. When it can't find people near that value it will look furthter away. Only once the teams are assembled, the average Elo of each team is used to compute the probability of the win.

Padic, on 07 August 2013 - 12:00 PM, said:

It never mentions averaging.

The only time I've seen the devs mention averaging with regards to matchmaking is when there are premade teams involved. In that case, each member on that team is treated as having an Elo equal to that of their average Elo. Which is, you know, whatever. If you're in a group and you're dropping with a noob, how can you possibly complain about being grouped with a noob?

Is there a different quote your getting your averaging info from? Just because I don't see it doesn't mean that it doesn't exist.

Here is the quote. It talks about computing the Elo after the team has been established. Not match making. It really is a common misunderstanding in the forum.

http://mwomercs.com/...79-matchmaking/

Quote

Here's how a team works:

At the start of a match, all player's on ONE team have their Elo ratings totaled and divided by 8 (max players). You may realized that this is simply the AVERAGE of a team's total Elo. Team 1's average and Team 2's average are then used to calculate the probability of win (as per the formulas above). If Team 1 beats Team 2, then the appropriate math as above is applied to each player using the probability score calculated by the team averages.

Edited by Hauser, 07 August 2013 - 03:48 PM.

[Gremlich Johns]

While I could play before the AMD/Avast issue, I most often single PUG'd (in an HBK-4G(F)). Great fun and sometimes I ended up on some pretty great teams that worked exceedingly well together.

[Sug]

Hauser, on 07 August 2013 - 03:48 PM, said:

Nah. Both N-Elo and S-Elo are changed as a result of winning. Because S-Elo is much higher then the rating a new player should be on it will fall quickly.

We don't know how they change. And I can't think of a scenario where there is even a point to having 2 separate Elo's when you start.