134 replies to this topic

[Clideb50]

Nikolai Lubkiewicz, on 20 May 2014 - 04:32 PM, said:

This x1000

If I may make a suggestion: Someone earlier suggested that you post a "Heads up, we're going to test a new matchmaker system on *insert date here* to collect data. Premium time will be paused etc etc etc." I have to agree that this would be a great way to test match maker updates instead of the test server. In my opinion (and I'm not an expert by any means), test servers will never see normal server numbers, even in a perfect world.

The end result would ideally be:
-You get plenty of telemetry data to adjust/tweak match making systems
-Players know in advance that servers will be sluggish, and won't go into a complete and utterly pointless ra- Oh who am I kidding, rage will happen even if you roll everything out perfectly.

[Degalus]

I think MWO needs one thing that maaany other mmos have.
Backgroundsearching!
World of Tanks, Warthunder, Dota2, Loadout and so on and so on.
They all have backgroundsearching where you can do something while you wait for a Match.
Let me mod my Mechs while im searching.
Its just boring to look at the searching greyscreen...

[Fooby]

Nikolai Lubkiewicz, on 20 May 2014 - 04:32 PM, said:

This x1000

I guess my only "question" to that would be the fact you had many weeks for internal testing since the April 29th initial attempt. If you didn't take at least 5 days of "peak player hours" match launch data and feed that into a test server with the new matchmaker to see how it handled, I really don't know what you believe is acceptable testing. Load testing in a virtual setting with fake launches can be done simulating far below and above what the peak that the live server deals with, though it cannot account for "everything" it certainly would show these issues, and using actual live server data certainly should. Though that kind of testing does require hardware and not just software to pull off I see no reason why it hasn't been done.

Edited by Fooby, 20 May 2014 - 05:07 PM.

[Verkhne]

Took a heavy mech as lowest percentage at 22% but WOW wait times are bad over 10 minutes so far! 15 MINUTES!!

[Bartholomew bartholomew]

I have a feeling the wait time is too long because the MM is parsing too much data and being too particular about it. Y'all didn't have enough players for two seperate buckets before for 3PV and 1PV. Now you have four weight based ones, plus ELO splitting.

It really does not look to be like it is going to work. All considering..

[L Y N X]

Hot fix seemed to INCREASE "Searching" times for me from 1-3 minutes to 5-10+ minutes...

[GoManGo]

You all are able to play? its been 2 hours now and im still trying to download the patch to even play? OW boy im not looking forward to 10-15 minute wait times for battles.Why can PGI make mechs out the aassss every week but cannot fix the game itself?

[Koniving]

Bartholomew bartholomew, on 20 May 2014 - 04:55 PM, said:

I have a feeling the wait time is too long because the MM is parsing too much data and being too particular about it. Y'all didn't have enough players for two seperate buckets before for 3PV and 1PV. Now you have four weight based ones, plus ELO splitting.

It really does not look to be like it is going to work. All considering..

I agree. The general concept and rules with contingencies of rules and varying skill levels, etc., it's generally extremely complicated.

I do recall that at one point long before this, players weren't so concerned about 'class based match making' but wanted even tonnage.

If team A has 660 tons, team B should have within 650 to 670 tons too.

Though what makes up that tonnage could be anything and everything; creating something possibly exciting since you can't look at your own force and go "Oh, lots of heavies and mediums The other side must have heavies and mediums too," when in fact the other side has 2 assaults and a boatload of lights.

"ZOMG! I had fun!"

Edited by Koniving, 20 May 2014 - 05:03 PM.

[Malleus011]

Just so Niko knows: the player base seems to still be experiencing long wait times even post-hotfix.

I gave up a little while ago myself. Is it possible the server population has crashed due to frustration?

[Rebas Kradd]

Bartholomew bartholomew, on 20 May 2014 - 04:55 PM, said:

I have a feeling the wait time is too long because the MM is parsing too much data and being too particular about it. Y'all didn't have enough players for two seperate buckets before for 3PV and 1PV. Now you have four weight based ones, plus ELO splitting.

And each one is split into seven further buckets for game modes.

This is a niche game and not populated enough to support this kind of matchmaking. PGI needs to go back to its old matchmaker, just learn to tolerate the whiners and their screenshots, and focus on SRM fixes and CW.

[N0MAD]

And you guys are paying these people for this, just LOL.

[GoManGo]

OMG EMERGENCY STAT REWIND!!!!!!!!!!!!!!!

SEARCHING!!!!!!!!!!!!!!!!!!

GONE TO PLAY WARTHUNDER TANKS?PLANES!!!!!!!!!

SEEYA LATER MAYBE!!!!!!

SEARCHING!!! AGAIN!!!!!!!

BYE!!!!!

Edited by GoManGo, 20 May 2014 - 05:20 PM.

[CykloneJack]

how bout fixing the lrm's and other missiles please?

[Monsoon]

Didn't start playing until after 8:30 EDT, haven't had much problem with waits, but missiles (LRM & SRM) seem borked, I've also heard others say the same about Streaks but I haven't checked on those yet.

Edited by Monsoon, 20 May 2014 - 05:52 PM.

[Oni74]

Patch > Hotfix > Rage appears to be the norm nowadays.

Seriously guys - I'm baffled as to how we can have a string of failures that result in worse service quality than what was being delivered before each patch.

Someone in a previous thread mentioned the use of battle values + ELO for MM. I think it would be far better method of MM than the twice-now-failed 3/3/3/3.

As much as I'd like to play through the 3+ minute wait times and serve as a guinea pig to feed you guys MM telemetry data, I remembered that I'm a paying customer with other things to do and other games to play. See you (PGI) after you hotfix the hotfix!

[Donny Donowitz]

ill just add that the downtime for this has granted me a wonderful bicycle ride this evening. i have to be optimistic and patient; this is my all time favorite video game. i suggest all of you do the same. i dont know of any other mechwarrior games out there..

[Trevor Belmont]

3 matches so far:

First (medium) took 6 minutes
Second (heavy) took 3 minutes
Third (medium) at 10 minutes and counting.

[wanderer]

Given the current status of the game, my "telemetry" is:

I'd like a day of premium time returned, please. Miscoding on missiles has rendered most of my favorites no-go for play, and the massive match delays means I'd barely get to play anything else while watching those who don't know about said missile screwup getting punted in matches due to being unable to contribute.

Ironically, meta-laden AC/PPC boats will function nominally- when they can actually drop.

Disappointing beyond belief, except for PGI. This patch wasn't even "minimally viable" anything.

[Jonathan Paine]

Does average weight time based on mech class help at all as long as it does not take the players Elo level into consideration? The matchmaker might have immediate place for, lets say, an Elite medium mech, but no place for any cadet medium mechs.

Of course, this could sneakily be put into the code without telling the players directly. It is not like we are sneaky and will use this information to guesstimate our Elo's. No way!

Just a thought...

[Canadian Kodiak]

Donny Donowitz, on 20 May 2014 - 05:59 PM, said:

ill just add that the downtime for this has granted me a wonderful bicycle ride this evening. i have to be optimistic and patient; this is my all time favorite video game. i suggest all of you do the same. i dont know of any other mechwarrior games out there..

Seem to recall that's because they lawyer-nuked the only other one around (and it was a free community build, IIRC). But you're right, I should go ride my bike wash my car.