151 replies to this topic

[ImperialKnight]

[Veranova]

ELO is always balanced within a tight range between each team. Paul has said this and proven it in a recent post.
However the longer you wait, the higher the disparity, between highest and lowest between each team.

The games mechanics are such that a single lance or player, slightly out of position, can lose the game for their team.
It's just bad luck when this happens. Even games where I've recognised almost every player on each team, have ended in rolls.

What you have there is quite a balanced match, because everyone pulled their weight in damage, and almost half of your team actually died in the process.


edit: I fixed my phrasing which caused a misunderstanding lower down.

Edited by Veranova, 07 February 2014 - 11:25 AM.

[SniperCon]

Paul Inouye said:

The average deviation between team Elo ratings was approximately 175 points. The extreme maximum was around 750.

ELO predicts that a team favored by 175 points should win 73.3% of the time. A team favored by 750 should win 98.7% of the time.

From my chess experience vs a player 175 points lower, I can't afford careless mistakes but can usually capitalize on a weakness in their play. Against a player 750 points lower, I will crush them without trying.

[Xeno Phalcon]

ELO is a sadistic god that enjoys mingling in the affairs of mortals, spreading suffering and pitting bunnies against gladiators.


But seriously, this game dosn't coddle players so as a result the mass majority are terrible no matter what bracket you are in more than half your team are basically headbutting the keyboard and complaining their losing for it.

[Joseph Mallan]

knightsljx, on 07 February 2014 - 04:05 AM, said:

You are right, Spirit1941 & you had no business in that match.

Edited by Joseph Mallan, 07 February 2014 - 10:19 AM.

[Nicholas Carlyle]

So many crappy Stalker champion's...I feasted on them last night. They stink.

[Joseph Mallan]

Nicholas Carlyle, on 07 February 2014 - 10:20 AM, said:

So many crappy Stalker champion's...I feasted on them last night. They stink.

That is such a Team oriented build it boggles the mind! I am tempted to give it a whirl... I miss m days as fire support for the Law!

[Nicholas Carlyle]

Joseph Mallan, on 07 February 2014 - 10:22 AM, said:

That is such a Team oriented build it boggles the mind! I am tempted to give it a whirl... I miss m days as fire support for the Law!

It's the XL engine in a Stalker that kills it.

I mean, I don't mind playing LRM's (even though it can be super frustrating), but that build is just bad.

[Roland]

Veranova, on 07 February 2014 - 04:23 AM, said:

ELO is always balanced within a tight range on each team. Paul has said this and proven it in a recent post.

Actually, Paul said no such thing.

Paul said that the overall team Elo average tended to be even most of the time... But that average Elo could (and seemingly IS) achieved by balancing good players with paste-eaters. Which tends to be a terrible idea.

[IceSerpent]

Roland, on 07 February 2014 - 10:38 AM, said:

Actually, Paul said no such thing.

Paul said that the overall team Elo average tended to be even most of the time... But that average Elo could (and seemingly IS) achieved by balancing good players with paste-eaters. Which tends to be a terrible idea.

I don't recall him saying that either - iirc he said something about 175 point Elo difference being an average case, and that difference means that "underdog" is roughly 3 times worse than the "favorite" (75% / 25% chance to win).

[Veranova]

Roland, on 07 February 2014 - 10:38 AM, said:

Actually, Paul said no such thing.

Paul said that the overall team Elo average tended to be even most of the time... But that average Elo could (and seemingly IS) achieved by balancing good players with paste-eaters. Which tends to be a terrible idea.

Sorry I fluffed my phrasing. I was trying to refer to the disparity between each team, not ON each team.

In that respect Paul HAS said this.
http://mwomercs.com/...__fromsearch__1

Under "Matchmaking"

"As of yesterday (Jan 20), the average deviation between team Elo ratings was approximately 175 points. The extreme maximum was around 750."
"I used 4 screenshots that were submitted (and keep in mind, these were complete wipes/stomps) and compiled the team Elo rating for each team involved. In screenshot 1, the Elo difference between the two teams was approximately 10. In screenshot 2, the Elo difference was 47. In screenshot 3 the difference was 101. In screenshot 4, the Elo difference was 65. What means is, that even if a game ends in a "stomp", there is a high chance that the opposing team was just outplayed via communication and on field maneuvering."

Edited by Veranova, 07 February 2014 - 11:23 AM.

[Joseph Mallan]

Nicholas Carlyle, on 07 February 2014 - 10:29 AM, said:

It's the XL engine in a Stalker that kills it.

I mean, I don't mind playing LRM's (even though it can be super frustrating), but that build is just bad.

Didn't look at the engine... Gotta think about that then.

IceSerpent, on 07 February 2014 - 11:08 AM, said:

I don't recall him saying that either - iirc he said something about 175 point Elo difference being an average case, and that difference means that "underdog" is roughly 3 times worse than the "favorite" (75% / 25% chance to win).

Ok. Knowing that, I can understand how the pendulum swings so hard. Still I like not knowing I your team will be the Hussars or the Black Widow Company this time!

[Sandpit]

I'm lovign the Champ Stalker. It's strictly a fire support mech though. 4 LRM15+Artemis and Tag? That's a good build and 4MLs

[Nicholas Carlyle]

Sandpit, on 07 February 2014 - 11:31 AM, said:

I'm lovign the Champ Stalker. It's strictly a fire support mech though. 4 LRM15+Artemis and Tag? That's a good build and 4MLs

XL...Engine...In...A...Stalker.

But of course you love it.

Oh and the awesome part is it gets hit by Ghost Heat. L freaking O L.

[Joseph Mallan]

If he playes it wisely, and stays out at range, 4 15s are a beast of missiles!

[Nicholas Carlyle]

Joseph Mallan, on 07 February 2014 - 11:40 AM, said:

If he playes it wisely, and stays out at range, 4 15s are a beast of missiles!

Honestly, 15's aren't what they used to be. They spread too much now and track poorly.

If you shoot them as one volley it's especially bad.

5's and 10's are the best LRM's.

[IceSerpent]

Veranova, on 07 February 2014 - 11:22 AM, said:

Sorry I fluffed my phrasing. I was trying to refer to the disparity between each team, not ON each team.

In that respect Paul HAS said this.
http://mwomercs.com/...__fromsearch__1

Under "Matchmaking"

"As of yesterday (Jan 20), the average deviation between team Elo ratings was approximately 175 points. The extreme maximum was around 750."

Let me get this straight - you're saying that "the average deviation between team chances to win was approximately 26.7% / 73.3%. The extreme maximum was around 1.3% / 98.7%." qualifies as "ELO is always balanced within a tight range between each team.", correct?

Edited by IceSerpent, 07 February 2014 - 11:44 AM.

[Joseph Mallan]

Nicholas Carlyle, on 07 February 2014 - 11:41 AM, said:

Honestly, 15's aren't what they used to be. They spread too much now and track poorly.

If you shoot them as one volley it's especially bad.

5's and 10's are the best LRM's.

What ever happened to go large or go home Play style

[SI The Joker]

2 words.. tonnage matching. Hopefully they do away with this ELO system once and for all and let everyone play with everyone.

The tonnage matching (or so they say) is coming.

/thread

[Joseph Mallan]

SI The Joker, on 07 February 2014 - 11:49 AM, said:

2 words.. tonnage matching. Hopefully they do away with this ELO system once and for all and let everyone play with everyone.

The tonnage matching (or so they say) is coming.

/thread

Heck even weight class matching worked better than Elo. Sure your Cicada could have been matched with a Centurion... but they were both Mediums at least!

Edited by Joseph Mallan, 07 February 2014 - 11:59 AM.