28 replies to this topic

[coe7]

Take a look of this edited pic,

http://imgur.com/LEskA2I

it's a collection of players who were in a match with unit friend of mine, Kossi. Kossi himself is pretty good in mwo, sporting lifetime KDR of 2.5+ while playing mechs from trash tier to actually good ones.

How is it even remotely plausible that matchmaker is literally hell bent on matching whatever it can find rather than extending match search time and finding actually fun matches? Current situation is absolutely garbage.

I personally think its following:

* As you can see on screenshot, these guys have played quite many matches since new stats where put in, but have done quite poorly. Its reasonable to suggest, since they face tier 1 that PGI's experience bar tier system has finally bumped people who have no business to be in tier 3 or 2 up so that they face competition out of their league.

* Instead of actually having matchmaker to try to put tier 1 or 2 players in their own matches, it takes first available players in any of the top 3 tiers and makes a match out of them. This leads to horrible inconsistency.

* Player like myself, is put in matches where lrm atlas and steering wheel aim is the norm, causing hackusations, rage and piss poor gameplay.

Considering that MWO's appeal is 100% tied to how fun the quick play matches are, more than any changes to the game I feel that quality and fun in the quickplay has gone down so much in last 5-4 months I have hard time justifying playing it. Situation is that dire.

It's not that we did not warn PGI, we told them their tier system needed to be Zero Sum and eventually tiers will become meaningless as zero contribution people will get bumped up to tiers they are not supposed to be.

I have no problems waiting 4-5 minutes for a match if it gurantees its built from solid tier 1 and 2 players, generating fun even matches, like it was when tiers where introduced. It was infact quite hard to do well in an average mech while levelling it when all you faced were good players. This was the first 4 months of tier system. After that... its been gradual decline in match quality.

I might not be the best player in MWO, nor I'm not tooting my own horn here. Everyone can check out my stats ingame. All I'm saying here that mudflation of tier system is finally done, and match quality is down the drain. This can not be fun to average players either. I'm also suggesting that lack of fun currently in quickplay is directly responsible for more than normal decline of players in summer period.

1. Rework the tier system to properly reflect skills of players. You have stats now to do so. Tiers must be zero sum based, period.

2. Increase wait times, good match is worth 3-5 minute wait at tier 1 and 2.

3. Do these fixes as number 1 priority, since quick play right now, is not fun.

4. Add UAV to show mech facing, play has pussified alot when aggressors advantage of seeing minimap facings has made gameplay a lil bit stale. Same deal with map voting, we rarely see maps that are interesting, but don't fall into generic mechbuilds people run. Either add dropship style "choose 1 mech out of 4 for this map" or drop the mapvote.

5. Get to bottom of this right now. Quickplay being problematic, equals people not wanting to buy new releases because platform to have fun with them is not actually fun. CW is done and implementation obviously does not attract us, so putting bit of effort to matchmaker should be high priority.

6. Stats being based on combined results of group and solo queue is very bad form. Stats should be either from solo or group, but not combined from both as playstyle, KDR and avg match score is totally different in solo and group queue.

7. TTK. There is no TTK problem when I face opponents of similar skill level. When I play these matches, I constantly tank my mech down to 30-45% before even being afraid of dying. The aim currently in "tier 1" match is often so poor, I can do stuff, I would never be able to pull off in a match against equals. People just make so bad moves and shoot so poorly. However, I alpha down opponents like they would be lawn, their reaction times are bad, their ability to twist properly is worse and their usage of cover is nonexistant. I'm firm believer that TTK would go up alot if people faced equally skilled opponents, instead of having a match where few good people do 800 to 1400 of damage, with 3-4 kills, carrying everyone. You can not make fun matches this way, and people will feel that TTK is very low, when they do not have the ability to keep themselves alive against the 1-2 actually good players who dropped into an average match.

Thanks for reading.

[Wintersdark]

You've got some wrong info here - MM makes a single tier match, so for example ONLY tier one, for 1m. Then 1+2 till the 2m mark, then 1+2+3 after.

Zero sum is complicated, because players enter with no ranking, and leave with higher rankings. Thus, points in the zero sum pool are consumed but not replaced.

So, while zero sum itself isnt a practical solution, the underlying concept is: have tiers split the player base into fifths, with the players with the top 20% average match score (could use better scoring!) In t1, then the next 20% in t2, etc.

You'd have to ignore players who haven't played in a given amount of time for that calculation though, or else you'd run into a similar problem..

[H I A S]

Wintersdark, on 13 July 2016 - 02:52 AM, said:

You've got some wrong info here - MM makes a single tier match, so for example ONLY tier one, for 1m. Then 1+2 till the 2m mark, then 1+2+3 after.

That makes it even more worse, right?

[MrMadguy]

MM is ELO-like crap and cares about W/L only. Their W/Ls are around 1? Yeah. Who cares, that their K/D = 0.0000000001, and other stats are crap too. PSR - is "team rating", i.e. once your team manages to win matches for you - your rating will rise, no matter, that you personally may be even AFK. "Carry or be carried" - is PSR's motto.

Edited by MrMadguy, 13 July 2016 - 03:14 AM.

[Wintersdark]

MrMadguy, on 13 July 2016 - 03:13 AM, said:

MM is ELO-like crap and cares about W/L only. Their W/Ls are around 1? Yeah. Who cares, that their K/D = 0.0000000001, and other stats are crap too. PSR - is "team rating", i.e. once your team manages to win matches for you - your rating will rise, no matter, that you personally may be even AFK. "Carry or be carried" - is PSR's motto.

No, it doesn't care about w/l only. It's just an XP bar; and primarily cares about DAMAGE. As long as you don't do appallingly bad, you'll still advance in PSR rating with a poor WLR.

arivio, on 13 July 2016 - 03:06 AM, said:

That makes it even more worse, right?

In a sense. PSR is only a good system for protecting newbies, nothing else. It's an XP bar.

The MM itself is fine - pretty good in fact - but it's stuck with people being classified by experience and also with very limited player counts. It'll never be very good due to player counts alone, but yeah: the current PSR system definitely decreases higher end match quality via stuffing T1 with everyone who plays a lot and isn't tremendously bad.

[Dingo Battler]

I disagree. Long wait time will just drive away more players, and create lower quality games.

The main issue is that the game is niche, not very good, and poorly designed.

It should have been 8v8, given the player population, and the fact that it's a niche game.

The problem will only go worse as time goes on, and our playerbase quickly marches towards 0.

I'm already waiting 3mins plus for the mm, because there are not enough players in tiers 1 to 3 to form a game.

One solution is for mwo to become mainstream, which is near impossible given the difficulty.

The other is for mwo to produce a good quality game to keep the people interested in the niche happy

[Tarogato]

MrMadguy, on 13 July 2016 - 03:13 AM, said:

MM is ELO-like crap and cares about W/L only. Their W/Ls are around 1? Yeah. Who cares, that their K/D = 0.0000000001, and other stats are crap too. PSR - is "team rating", i.e. once your team manages to win matches for you - your rating will rise, no matter, that you personally may be even AFK. "Carry or be carried" - is PSR's motto.

Incorrect. You can't advance in PSR if you don't break 100 match score with a 1.0 WLR.





That said, if you average exactly 251 match score and have a 1.0 WLR (or average slightly lower but have a higher WLR), you will eventually reach Tier 1. Players that average 401 match score will reach Tier 1 sooner.

Which almost sounds like it works, because theoretically as you go up in PSR, you face other people who also went up in PSR, thus stiffer competition hence your match scores should go down and your PSR should level out. But that's not what happens. Because people that don't deserve to increase in PSR........... increase in PSR. Thus when you tier up, you're not necessarily facing stiffer competition, hence your match scores stay the same, and you keep climbing, and so does everybody else with you.

Edited by Tarogato, 13 July 2016 - 06:07 AM.

[El Bandito]

KBurn85, on 13 July 2016 - 06:04 AM, said:

It should have been 8v8, given the player population, and the fact that it's a niche game.

^ This.

[MrMadguy]

Wintersdark, on 13 July 2016 - 05:03 AM, said:

No, it doesn't care about w/l only. It's just an XP bar; and primarily cares about DAMAGE. As long as you don't do appallingly bad, you'll still advance in PSR rating with a poor WLR.

This should be sticky post to prevent all such threads:

What many players don't understand, is simple rule: more skilled players, you play against - lesser your chances to win. That's why all tiers have exactly the same rating levels. That's why PSR - isn't XP bar. Because having W/L = 1 and earning 200 match score in Tier 5, while stomping complete noobs - isn't the same thing, as in Tier 1, when playing against skilled players.

See my sig for reference, if you're interested in how current MM works.

[Mudhutwarrior]

Played literally thousands of matches before and after Tiers came into effect. I started at 3 and 3/4 tier and remain there never moving up or down.

The idea that we will all reach tier one is ludicrous. I have no idea who started it but they are clearly an idiot spreading disinfo.

Tier 3 is mostly hell because you meet everyone on every tier all the time. You either stomp or get stomped and rarely see any close matches.

The system is junk unless you are at the top or bottom.

Edited by Mudhutwarrior, 13 July 2016 - 06:19 AM.

[Mawai]

MrMadguy, on 13 July 2016 - 03:13 AM, said:

MM is ELO-like crap and cares about W/L only. Their W/Ls are around 1? Yeah. Who cares, that their K/D = 0.0000000001, and other stats are crap too. PSR - is "team rating", i.e. once your team manages to win matches for you - your rating will rise, no matter, that you personally may be even AFK. "Carry or be carried" - is PSR's motto.

WRONG.

PGI removed ELO ages ago and replaced it with PSR. PSR is a "skill" rating calculated based on match score with a bias towards winning. If you win and do well you go up a lot. If you lose and do great you go up a little. If you win and don't do much you don't change. If you lose and don't do much you drop a lot. If you lose and do well you don't change.

Note: You can't always be "carried" ... you don't do much and your PSR goes down even if you win half your games. However, if you play "average" or a bit below (looks like an average match score of 150 to 200) then you will go up but it will take a lot of games.

PSR is not a conservative system. Points aren't gained and lost equally. Every match will likely generate MORE total PSR points so the pool and average PSR score over all players is constantly increasing (this is why many folks call PSR an experience system). As long as you don't play consistently badly, your PSR will go up.

Which brings us around to the OP. Every single one of those players in the list could well be "tier 1". They all have played lots of matches and are considered "equivalent" by the matchmaker since that is the data it is being given ... the matchmaker forms matches from players with equivalent PSR ratings.

Lots of folks have been saying from the beginning that PSR will eventually break as enough players reach tier 1 since the system is not conservative ... tier 1 does NOT reflect "skill" explicitly ... with the increase of complaints regarding matchmaking maybe we are finally reaching that point.

(Also note ... PSR does not have separate scores for weight classes or solo vs group queues. It needs this granularity because a good assault pilot can be a crap light pilot and vice versa ... each weight class mostly plays a bit differently and individuals have preferred playstyles. The rating system really needs more granularity so that at least skill in the mech class and solo vs group should have different values. Anyway, there are probably dozens of us in the community that could contribute to the design and construction of such a system ... maybe PGI should consider investing in contract work for specific projects like skill rating and matchmaking).

[KodiakGW]

KBurn85, on 13 July 2016 - 06:04 AM, said:

It should have been 8v8, given the player population, and the fact that it's a niche game.

+1 This..so much this. We had 8v8, but for monitary reasons they switched to 12v12.

Now we have a server queue for 4v4, FP Scouting. So now they are willing to give a server queue for half what the old server queue was, where matches take about the same amount of time. So can we have 8v8 back? I know a number of people who may come back if we do. Problem is, it is now "may" when a number of months ago it was "will". Time to make the change before it becomes "never".

KBurn85, on 13 July 2016 - 06:04 AM, said:

I'm already waiting 3mins plus for the mm, because there are not enough players in tiers 1 to 3 to form a game.

Now, with this I have a real issue. It NEVER takes me that long to find a match in solo QP, even when I take a mech with 40% plus queue. I usually notice my first game of the night takes up to 90 seconds to find a match. Most times after that, 5-15 seconds. Hell, a few times I clicked and immediately got a game. Almost every game is crap, though. Most times I'm trying to save my team from being stomped (sometimes succeeding). Fewer times my team is doing the stomping and I can barely keep up. Rarely, I actually have fun close game.

So, I'm not sure what is going on that you (KBurn) are waiting so long, and I get game so fast. I usually play a couple of hours anywhere from 4:30 to 1AM EST weekdays, and various times on the weekends when I do. If you are playing the same times, then there is something going on.

[Wintersdark]

I miss 8v8 so much

[Simbacca]

KBurn85, on 13 July 2016 - 06:04 AM, said:

I disagree. Long wait time will just drive away more players, and create lower quality games.

The main issue is that the game is niche, not very good, and poorly designed.

It should have been 8v8, given the player population, and the fact that it's a niche game.

The problem will only go worse as time goes on, and our playerbase quickly marches towards 0.

8vs8 (for solo or group queue) would help for faster MM. But the core problems of PSR also have to be addressed (along with several other game mechanics).

[Foxwalker]

Mudhutwarrior, on 13 July 2016 - 06:18 AM, said:

Played literally thousands of matches before and after Tiers came into effect. I started at 3 and 3/4 tier and remain there never moving up or down.

The idea that we will all reach tier one is ludicrous. I have no idea who started it but they are clearly an idiot spreading disinfo.

Tier 3 is mostly hell because you meet everyone on every tier all the time. You either stomp or get stomped and rarely see any close matches.

The system is junk unless you are at the top or bottom.

I am guessing you mostly PUG most of the time. If you group with a good team, you will most likely progress to the next tiers eventually, if your teammates are good, your group wins more games and your arrow will be up and green more often.

Pug matches are different and you will progress to your skill set and stay static, unless you figure out what you need to do to progress.
Some guys for no reason will run forward by themselves and die quickly. Some will run off by themselves and not contribute much. Often, players will be running bad builds. Checking metabuilds will help with that. Also mech chassis that just are not viable in the higher tiers.

The hardest thing for me and I guess most players is battlefield awareness. Did you notice the UAV above your head? Did you not check your six for the light taking out your back? The last thing I think in matches is communication. Often I have seen lights go off and scout, but not provide any of that intel back to the group. What good is spotting the enemy, but not putting up a UAV to show their location to the team? Calling targets for your team. Anyone can do this but almost no one seems to. Concentrated fire takes anything down fast.

Simple rules but it equals moving up in tiers.

S

[Yellonet]

The PSR change should be the same no matter if you won or lost, that way you couldn't get carried to higher tiers which is what happens to all the bads right now. The motivation for winning is more xp and cb.

[Spheroid]

Does the matchmaker try to balance the number of upper tier players between the two teams or does it just care about the average tier value? I suspect the later but have no evidence.

[Wintersdark]

Spheroid, on 13 July 2016 - 09:56 AM, said:

Does the matchmaker try to balance the number of upper tier players between the two teams or does it just care about the average tier value? I suspect the later but have no evidence.

It builds teams by:

1) seed match with oldest player in the queue (guy whos been in the queue longest)
2) add players at his rating, moving out from his rating but staying within his tier.
3) if 23 more players have not been found within 1m, stretch to allow +/-1 tier. Continue adding players with as little Delta from the seed player as possible le
4) if after 2 minutes the MM is STILL looking, stretch another tier.

So, the answer to your question is no, but with the caveat that matches weren't built exactly as you thought in the first place.

[Lily from animove]

the longer we play with the current psr settings the more derp non T1 material climbs up messing the system more and more.

[Deathlike]

I'm just going offer an explanation. People can take it or leave it, it doesn't matter, but it's not going to matter ultimately.

This isn't really just a phoneomenon... telling people to "come back" after a break from MWO. It has started long before launch.. including Beta (at least Open Beta for me), where I've seen people and their rant threads and why they quit MWO or at seeing people say "come back later, it'll be better". Well, I can tell you that it didn't get that much better, but people keep coming back, and draw the same conclusion a bit later... to go back on hiatus.

Now, this isn't something entirely random.. but you probably know someone else in your unit that feels this way, and is taking a break. Heck, I bet you'll get a similar response from the top tier comp units... while they aren't "hating" on MWO, they are not having the same level dedication than in other periods with this game.


What PGI has done, and what could arguably be to their own detriment.. is convert hardcore players of this game... to being casual.. assuming they haven't actually ragequit the game entirely. Think about it. I'm probably a loud posterchild of it.


I guess the more important question to really ask is - what does this mean to me (or this thread for that matter)?


As people that are more experienced/dedicated to this game become more "casual", the pool of players that you're pulling intelligence from really declines. Mind you, I don't think many players progress in this game due to the terrible NPE, but as currently designed - the stagnation of this game will get more people to play this game less.

Even if you argue it's "not a population issue", people that probably played this game with many hours a week have declined in the # of overall hours spent in the game. Sure, you'll see your buddy... but he could play 1 hour a week instead of something like 5 hours a week - just because the sheer monotony, boredom, or even lack of progress has caused them to not give a damn.


Steam numbers don't lie... while it can be "inaccurate" at best, we can surmise a few things from it. The Steam population has "normalized" a bit, and it doesn't look pretty. The NPE is definitely a factor, but if we go by some of the blatant false advertising from the FW video (Timberwolf in scouting? Must be some secret mode right?), the game should be as exciting as it looks.... only to find out that the Long Toms cull mechs on the field, the players, and the activity in the queues.


I'm telling you straight up - the upcoming "2 second mechlab save" should've been here ages ago... as QOL stuff should have appeared in the game long before we talk about the "circle of life", it's not going to magically excite the playerbase. I wish it could. Other games that have some QOL issues get praised to heavens when those changes actually come online... while the "Clan pack release" didn't give PGI any pause to rectify the situation from the start - only the delay such a change 3 years later (PGI did address the mechlab saving issue partially - because it became damned obvious) where it became "a built in feature" of the game for that long.

I don't know about you, but 3 years since the last major mechlab saving changing is not something to really celebrate about. It needed attention, but apparently it wasn't important enough - despite some people having a Pokemech collection of sorts. It gets them thinking sometime "do I really need all those mechs?" It doesn't help sales directly right?


TL;DR

The longer the game becomes stale (and it's rock hard from my POV, with quite a bit of mold), the more the MM reflects the players in the queues. There's enough people on some sort of MWO sabbatical that would be a real vacation for some.

Unless something huge happens soon (and decals are still cosmetic ultimately - they don't change gameplay), this is going to linger even longer than an unbearable summer.

Edited by Deathlike, 13 July 2016 - 11:52 AM.