168 replies to this topic

[Kyle Polulak]

Greetings MechWarriors,

Please respond to the above question, and let us know what your thoughts and concerns regarding this latest Command Chair post by Paul.

Spoiler

Hey folks,

Touching on Elo again. Recently the reduction in wait times that you have noticed was due to the Elo threshold for searching players being increased from 1000 to 2300. After reviewing metrics, this was not needed to be this wide. We are going to run a test (which goes live without having to patch) at setting this to 1400. As of 11:30AM Pacific time, we set it to 1400 and are letting the numbers adjust so we can look at as broad of a spectrum as possible.

What I ask is if you have noticed a change between now and that time, what was the change you noted.

Please click one of the options in the following poll.

Thanks for your time.

-Paul



Poll Feedback thread: http://mwomercs.com/...djustment-poll/

[Kaldor]

Shorter

But ELO is still a poor way to balance this game.

Weight balance please

[Iqfish]

Shorter! GG Paul

[Heffay]

Will check when I get home tonight, but I like tightening it up a bit. Even 1400 seems high, but for the 5 guys who play in the 2500+ Elo bracket, the game must have been not fun at 1000 spread.

I've had some fantastic matches tonight. I want to blame the changes, but...

Edited by Heffay, 23 January 2014 - 07:07 PM.

[Sandpit]

I have also noticed that my games both grouped and pug are closer. Last night was a steady stream of good close games without stomps

[Skyfaller]

What about ' Remove Elo' option?

Its USELESS to match people on win:loss ratio. As a scout I can tag/spot/uav every enemy mech and keep them lit on the map plus do 400 damage and kill 3 of them so I'm doing my job skilfully. Yet if the rest of my team is full of lemmings and they all die without scoring a single kill then tell me, WHY is that hurting *MY* matchmaker rating?

Its absurd.

On top of that, it does not match by tonnage. This means a team of VERY good medium pilots can end up facing a team of very good assaults. Its not uncommon of late to see teams of 8 assaults 4 others facing teams that don't even have an assault mech.

Finally, premade teams exploit this silly matchmaker system by loading full assaults (usually ECM Atlai) knowing full well they'll get matched against teams that are likely to not have as many assaults and definitely not as many ECMs.

Wake up PGI. Ditch ELO. Match by tonnage. Limit each team to 1 light/medium ECM and 1 heavy/assault ECM. Do not allow premades in PUG teams.

[Heffay]

Testing in prod....

Edited by Heffay, 23 January 2014 - 02:24 PM.

[Heffay]

Skyfaller, on 23 January 2014 - 02:19 PM, said:

What about ' Remove Elo' option?

Its USELESS to match people on win:loss ratio. As a scout I can tag/spot/uav every enemy mech and keep them lit on the map plus do 400 damage and kill 3 of them so I'm doing my job skilfully. Yet if the rest of my team is full of lemmings and they all die without scoring a single kill then tell me, WHY is that hurting *MY* matchmaker rating?

It probably isn't. If your team Elo is significantly below your opponent's Elo (like say 175), a loss won't move your Elo at all. Or a point or two. And it'll still get balanced in the long run.

[Ghogiel]

I don't get failed to find any more.

I do get get more stomps in if I premade with good players. Losses end up with me or my group with 800+ dam and every kill the team makes while the other players bar 1 or 2 all did sub 200dam.

So slightly worse than before but had 2 hilarious matches in the past couple days due to the derps on either team hahaha

[Prezimonto]

So I was right. You could get matched with as broad a selection of players on your team... as long as the two team average ELO's were the same.

So YES, you were playing with people from a very wide ELO before. And NOW you'll be playing with people from a total range of ±1400 ELO. That's still pretty large(50% larger still than when games were created at as close to your target ELO as possible).

I greatly hope you guys can still tweak this smaller.

I also greatly wish you'd use a much more granular method of assigning ELO... instead of weight class based on loadout, range at which you take and give damage, ELO for any combination of players, ELO for specific mechs with specific loadouts. I might be AMAZING in a dual gauss K2, and awful with 2xPPC and 2xLL.... but my ELO counts the same for both builds (which play very different).

Edited by Prezimonto, 23 January 2014 - 02:44 PM.

[Helmer]

Kaldor, on 23 January 2014 - 02:07 PM, said:

Shorter

But ELO is still a poor way to balance this game.

Weight balance please

I really, REALLY, look forward to tonnage limitations.


In case you where unaware, it appears tonnage limits are coming 'around April':

Twitter User Jan 13th
I would really like to know roughly when are tonnage limits coming into the game? Only w/CW?
Bryan Ekman ‏@bryanekmanJan 14th
around April with the launch module.

Source: https://twitter.com/...197580185448448



Cheers

[C E Dwyer]

shorter, however I don't think the stomp issue is the ELO

In most cases the stomp has been caused by a big weight imbalance, I know there isn't much you can do about a 4 man all deciding to run locusts, but in most cases the imbalance isn't caused by this

[Ghogiel]

Good players are never easy to overcome. Tonnage can be delt with to a degree.

[RichAC]

Skyfaller, on 23 January 2014 - 02:19 PM, said:

What about ' Remove Elo' option?

Its USELESS to match people on win:loss ratio. As a scout I can tag/spot/uav every enemy mech and keep them lit on the map plus do 400 damage and kill 3 of them so I'm doing my job skilfully. Yet if the rest of my team is full of lemmings and they all die without scoring a single kill then tell me, WHY is that hurting *MY* matchmaker rating?

Its absurd.

On top of that, it does not match by tonnage. This means a team of VERY good medium pilots can end up facing a team of very good assaults. Its not uncommon of late to see teams of 8 assaults 4 others facing teams that don't even have an assault mech.

Finally, premade teams exploit this silly matchmaker system by loading full assaults (usually ECM Atlai) knowing full well they'll get matched against teams that are likely to not have as many assaults and definitely not as many ECMs.

Wake up PGI. Ditch ELO. Match by tonnage. Limit each team to 1 light/medium ECM and 1 heavy/assault ECM. Do not allow premades in PUG teams.

I agree with everything you said, except I feel they should still rank players for match maker. I think it should be done by match score and not win/loss alone. Maybe according to just the players on your own team. Quakelive has a good skill rating system.

But great insight, on all your points.

Most people aren't even playing to win in this game, unfortunately, they are playing for cbills. So they are always going to complain about being badly matched, even when they are on the winning team.

But win/loss is not as related to an individuals "skill" as the match score is. I thought this was common sense, so this is all a shocking revelation for me today.

Also to add, bring back the individual points for capping each base in conquest if you were on the base while it went from yellow to blue like you used to. Why was this removed?

Edited by RichAC, 23 January 2014 - 03:13 PM.

[White Bear 84]

Sandpit, on 23 January 2014 - 02:19 PM, said:

I have also noticed that my games both grouped and pug are closer. Last night was a steady stream of good close games without stomps

But the metrics show...

Id hope this change was in part due to people highlighting the level of stomps in the game - yet PGI were pretty keen to defend their position on ELO with examples showing that stomps are apparently not a function of skill level (although 4 is hardly a statistically significant sample...)

If the change in the range is leading to closer matches, then I think to say the community concern was right. I know id pick long waiting times over a poor ELO match..

[Jman5]

• Search time feels the same, or at least it doesn't feel noticeably worse.
  
• I also haven't had any blowout games yet, but I have only played 4 matches today.
  
• Tonnage variation between teams so far has been 45 tons, 65 tons, 110 tons, and 180 tons. So far it seems like it's just as uneven as it was before. I'll play more games and let you all know if things change.

[Chavette]

Solo it a bit longer bit its OK.

As you'd imagine, not too much of difference in match quality though.


Edit: 2 out of 5 times can't find match in a 4man... I'd say make it a bit wider, just a bit.

Edited by Chavette, 23 January 2014 - 04:34 PM.

[FuzzyLog1c]

Making the matchmaker more loose isn't going to help if there isn't anybody to match you with.

Be prepared for longer and longer waits, people.

[Ghogiel]

FuzzyLog1c, on 23 January 2014 - 03:21 PM, said:

Making the matchmaker more loose isn't going to help if there isn't anybody to match you with.

Be prepared for longer and longer waits, people.

Errr.. Before you would get failed to find consistently because the threshold never expanded wide enough to get you into the the middle range where most of the players are. Now it does, which opens up the bulk of entirety of the player base

[Darth Futuza]

As far as I can tell the wait time has been reduced. But I've gotten 7 matches on terra therma. o.O So it could just be the random gods screwing with me.