State Of Match Making - Feedback/comments
#561
Posted 05 September 2015 - 05:52 AM
#562
Posted 05 September 2015 - 06:01 AM
#563
Posted 05 September 2015 - 06:07 AM
CainenEX, on 05 September 2015 - 05:52 AM, said:
I 100% understand your sentiments here. But for what it's worth, my two-man finds paring up in the group queue to be virtually impossible, at least if we want to enjoy ourselves. I understand that's our problem, but simply want to convey the sense of utter frustration we have. We cannot play together and have any semblance of fun in the group queue. I don't mind losing; not even a little. But I do mind 8/10 drops to be utter stomps in which he and I are the only ones with more than 200 damage. I don't know much about the match scores, but it used to be he and I would get 80-100 and 2/3 of our team would be below 20. Meanwhile, the 12 man group lost one mech. I get that stomps happen, but it's going to be pretty difficult to get five two man groups to be as coordinated as a 12 man, even with the integrated VOIP (which I've never got to work).
#564
Posted 05 September 2015 - 06:12 AM
Kjudoon, on 05 September 2015 - 01:51 AM, said:
how can these be at the expanse of everyone else? please explain.
rolkstomps. ah yes those were the reason: the groupqueue had been seperated from the solo queue and at one time gruops had been limited to 4 people. and?
SURPRISE:
As many people had predicted it had not changed anything, because people did not look themselfs or bad luck as the reason for roflstomps but no of course groups must have been the reason,
reality was completely faded out:
- looking at the Mechs of each side you saw often one side had no ECM the other 4 ECM. one side had several top notch mechs the other many mechs known as really bad performers.
-Builts on the losing side were terrible; LRM mechs without any backup weapons, plainly bad builts such as flamers or crabs with MG and so on
-people were running off solitary or in 2 man packs into the enemy, get wasted within the first 2 minutes
-people not paying heed to their surroundings if 3 mechs push, they simply stand and watch
-people not paying heed to orders, a prime target has been called, they shoot some light in 1 km distance with medium lasers...
-people not using torso twist, face tanking a dire in a griffin is a pretty bad idea they still do it, die within seconds and whine about broken mechanics
-people playing with joysticks, you see instantly if you watch a game who plays with joystick or gamepad. Any side having one of these has one person that cannot compete simply because his controller device is less precise AND less fast than a mouse. That is something that you have to blame PGI for, for even supporting vastly inferior input devices, or at least not telling people that they will never be able to compete meaningfully with mouse players
- people have no idea about basic tactics what tactics is and what it actually means. Apart from knowing you should join a push, you have to sustain a push and not turning around yammering the first time a light shoots at you.
-wrong Target Priorisation, since closed beta I keep seeing people who run with an assault after a light or who shoot a mech when directly to the left is a smoldering wreck that only needs a single Medium Laser hit but has still full weaponry and delivers heavy dmg.
- target switching, people shoot a mech then switch target. Even if the mech has cherry red torso and then switch the target instead of killing it, sometime that mechs stays alive until the end of the match and did 1k damage...
- people who completely ignore the map and who are oblivious to locking targets.
- people having not the slightest idea about basic mechanics of the game, shooting LRM without lock or withhin 100m, shooting MG at targets in 500m distance, shooting all weapons at once and shut down do that again instantly shutdown again.
Here PGI is of course responsible for not delivering a meaningful explanation to the mechanics for new players. But I must admit the tutorial I have seen on the test server has been done very nicely. PGI HEAR THIS YOU GOT A PRAISE!!!
-people not being able to hit the enemy, often directly combined to the fact they are using a bad input device.
All of those things contribute to a loss maybe a devastating loss, but of course groups have to be responsible for ones shortcomings... It was not true during the time of 4-Man groups it is not true now. Instead of looking at yourself the evil evil groups must have been responsible.
Strangely roflstomps did never cease simply because large groups at that time before 4-man goups were rather rare as these are now or these were matched to each other...
I rather take it the opposite to what PGI proposes: REINTRODUCE 2-4 man groups into solo PUG AND allow Solo players to join the group queue, compensate for the risk by giving solo players a Multiplicator to the monetary earning in the group queue or a flat amount of money (+50k maybe if matchscare is above 100 to avoid AFK abuse), to avoid new players being waxed give a XP lock you have to pay 100k XP for each weight class that you wish to play in the group queue solo, deductable from any mech and any mech combination in that weight class. Before you ask I have no problems joining as a solo player to the group queue. Also reduce the MM restriction to 12 players, if it finds only good matches for 9 players make it a 9 player match...
Groups were and are only a cheap excuse for losing. They might have contributed but the main advantage of group was and is that for each person you take with you, the possibility of people that can barely steer a mech on ones side or have a really bad built will drop significantly.
#565
Posted 05 September 2015 - 06:15 AM
Fire for Effect, on 05 September 2015 - 05:17 AM, said:
Of course you can try to game the system but its infinitely better than stuffing a cicada and a stormcrow into the same bucket and saying yes these are equal.
Use Mech BV times pilot ability to match; so what if one side has 2 heavy mechs more... Matches can finally be fascinating again if the enemy brings 5 lights or no lights to the battlefield.
It seems someone already posted my $.02 here.
I fully agree. In tabletop BattleTech (ducks quickly to avoid incoming barrage of rotten tomatoes), Battle Value and player experience had an impact on things.
If PSR is a mixture of KDR, assists, and other performance-based stats, then making it a multiplier for Mech Battle Value would be a good solution.
If the devs are tired of a certain type of "boat," be it a multi-LRM5 boat or a small pulse laser boat, have multiples of the same weapon stack the Battle Value higher and then add that in with a PSR multiplier. The result would be that players who "boat" weapons would be put into a higher play bracket than those who do not. ECM and Clan tech would need to be multipliers as well ... if a premade with 5 ECM mechs had to wait a longer time for a match, they might learn that trying to be invisible makes for a very boring night. (This would allow for greater separation in matches without nerfing the technology itself. It would also allow those who are "stock mech purists" a chance to match up and play with no problem)
This way, a player just starting out with basic 'Mechs would not be facing boated loadouts that are totally tricked out with max modules and pilot skill perks. This would make the current pilot skill tree more important without making a huge change to the game and help keep the lower and higher players separated.
For example, I am currently creating a new company, the Order of Asuncion [OA] ... I only have five active mechwarriors and we average three to four in attendance for our once per week practice (by Seraphim standards, [OA] is a Tier Four group). This two hour practice in the group queue is dominated by an opposing side full of 8-10 man pre-made groups. It is very hard to train our newest pilot (only 8 practices under her belt) even when she has three veterans for wingmen while facing those odds. The result? Practices that used to be two hours are shutting down at 1.5 hours or shorter due to frustration with over-matched OpFor composition.
Smaller groups cannot stand against large pre-made groups especially when everyone does not use in-game VOIP or text-chat. (Another point ... do not let users change VOIP or chat settings during a match. Make those who choose to opt out of chat and VOIP a negative factor in determining the BV since they are choosing to not coordinate).
For matchmaking purposes, more tiers would be beneficial (I am thinking the way Halo matchmaker used to work). With just a few tiers, there may be a large difference between the play of say tiers 3 and 2. More tiers would allow for more fine tuning of matches without a hard mech restriction (as Fire for Effect stated).
A broken CW has forced 12 mans down to group queue for matches. This is driving smaller groups down to sync dropping into the single player queue to have competitive yet fair matches. Fix CW and you fix this problem. Give 12 man groups something to fight for in CW and group queue will be a decent training ground for smaller units.
Why should smaller units have to train in private matches to effectively train their new players without having them totally demoralized due to ROFL stomps? This keeps new players from getting the XP and C-Bills needed to level their 'Mechs up and makes the problem worse, not better.
So, to re-inforce Fire for Effect's idea ... institute a PSR * Battle Value system with attention given to boated weapons, ECM, and Clan Tech as "unbalanced" multipliers. This put in effect with more tiers would make for more balanced match play and a better experience for our newer players .
TL; DR
- I second Fire for Effect's idea: PSR * Battle Value = matchmaking score
- Add multipliers to BV for things that add imbalances (ECM, Clan Tech, boated weapons)
- Make in-game VOIP and chat modifiers as well (players who opt out should be in lower tiers due to lack of coordination)
- Add more player tiers to allow for smaller deviations in player skill
- Fix CW and fix the problem
Edited by Alex Reed, 05 September 2015 - 06:26 AM.
#566
Posted 05 September 2015 - 06:16 AM
Which was a massive improvement from the 10+ minute waits.
Edited by DarthPeanut, 05 September 2015 - 08:10 AM.
#567
Posted 05 September 2015 - 06:16 AM
PhoenixFire55, on 05 September 2015 - 02:08 AM, said:
Those weren't even afk.
I did three matches yesterday before I gave up. Two of the three you can see here.
sry I dont see any matches.... but 3 is not staticticlly meaningful...
I have good solo matches since PSR has been introduced... not all but many many proper games, more than ever before.
#568
Posted 05 September 2015 - 06:39 AM
TheStrider, on 05 September 2015 - 05:41 AM, said:
yes that you did not read the thread is pretty obvious...
TheStrider, on 05 September 2015 - 05:41 AM, said:
so your circumstantial evidence is worth what? hearsay! Is it often? nope, cant be because PGI says that large groups are rare.
SO HOW CAN SUCH A SMALL AMOUNT OF PEOPLE BE A PROBLEM?
So maybe dont use your biased point of view as a reference but vastly exaggerating the amount of large groups and keep to the facts.
TheStrider, on 05 September 2015 - 05:41 AM, said:
Actually Pro Sports teams DO train against amateur teams (sometimes its even open for public to watch, its always great for the amateur team), Pro Boxer are sparringing against amateurs. And yes we are training, and no we are not "pro". Since VoIP you CAN organize if you dont use it thats your problem. But no you cannot train properly if you have no random enemy that you have no idea what he might do, or what setup he has. Against other units is a nice idea, you are volunteering?
TheStrider, on 05 September 2015 - 05:41 AM, said:
Again a tonnage centric view, Tonnage is not a useful measure for MM
USE COMBAT POTENTIAL TIMES PILOT SKILL FOR MM!
But i agree weight is potentially better than weight classes but still very bad (Orion = Mad Cat really? you serious?), still if you do it do it right go, for the combat potential of each mech instead of weight. We keep telling this since closed beta for over 3 years now...
#569
Posted 05 September 2015 - 06:42 AM
Josef Koba, on 05 September 2015 - 06:07 AM, said:
I 100% understand your sentiments here. But for what it's worth, my two-man finds paring up in the group queue to be virtually impossible, at least if we want to enjoy ourselves. I understand that's our problem, but simply want to convey the sense of utter frustration we have. We cannot play together and have any semblance of fun in the group queue. I don't mind losing; not even a little. But I do mind 8/10 drops to be utter stomps in which he and I are the only ones with more than 200 damage. I don't know much about the match scores, but it used to be he and I would get 80-100 and 2/3 of our team would be below 20. Meanwhile, the 12 man group lost one mech. I get that stomps happen, but it's going to be pretty difficult to get five two man groups to be as coordinated as a 12 man, even with the integrated VOIP (which I've never got to work).
I'm sorry to hear about that. My unit still feels the stomps as well, and against smaller teams too. It happens. Its very frustrating that a 2 man of friends encounter a 12 man of SJR or 228 to deal with. Back when I played league of legends such match ups did not exist due to how their MM was made. Mind you they also had the population to help with the issue. What should be happening is that the 12 mans need to be matched with 12 mans, and smaller groups be teamed with smaller groups within their given ELO/PSR/MM rating. However MWO does not have the population needs to make wait times sane.
I'm sure there can be more effective solutions if the MWO devs investigate further, or until the steam launch helps with population. Otherwise that other two mentioned solutions are going to create more problems, and you can still get stomped depending on the skill disparity
If anyone should be complaining it should have been me after THIS match ;p
Not sure what I can suggest in the mean time. Try solo queue, practice hardcore in group, or maybe try your luck in CW?
I'd like to help with your ingame VOIP issue though. Do you have a Microphone on your computer?
Edited by CainenEX, 05 September 2015 - 06:47 AM.
#570
Posted 05 September 2015 - 06:46 AM
Alex Reed, on 05 September 2015 - 06:15 AM, said:
I fear it is not my idea it was already brought up at least 3 years ago during closed beta. I am afraid I dont remember who bought it up. But it is ages old and was already in use during TT time. It was already described in the "Mercenary Handbook".
#571
Posted 05 September 2015 - 06:50 AM
Is the purpose of matchmaking in the group queue to protect the enjoyment and cohesiveness of the large 12 man groups or to create an environment where smaller groups (pugging with friends) can run and grow into these larger groups?
The more inclusive the solution is to all parties, the greater chance this game and game community will grow and prosper.
When it comes to e-sports ... were all the great e-sport games created to be e-sports games or were they great games with great multiplayer experiences that naturally translated to e-sport status?
I think that trying to create a great e-sport without creating a great game is a problem. If MWO caters to large competitive groups at the expense of new players and newly formed units, it will die before it truly gets out of the gates.
Big indicator ... check the Faction and Merc Corps forums. How many new units are being formed and are recruiting? How many of these are truly new groups and not splinters of former groups?
I think the world of both of you guys (Thanks to Wazan) ... we just have to focus on what is good for everyone so that we have a game to play next year.
Edited by Alex Reed, 05 September 2015 - 06:51 AM.
#572
Posted 05 September 2015 - 06:53 AM
CainenEX, on 05 September 2015 - 06:42 AM, said:
I'm sorry to hear about that. My unit still feels the stomps as well, and against smaller teams too. It happens. Its very frustrating that a 2 man of friends encounter a 12 man of SJR or 228 to deal with.
Do you really think it would have been any diffirent with max 4 steel jags or only 4 228?
Dream on, the pain just would have been longer... since the steel jags /228 would have taken longer to wax you and you would have killed one enemy PUG more, but that would be the only difference....
If it would be 4 max then THREE groups on the other side of 228/ Steel jags would have been waxed 12:0 or 12:1. So you would have trippled the pain. Congratulations you have spread the epidemic threefold. great solution...
Edited by Fire for Effect, 05 September 2015 - 07:00 AM.
#575
Posted 05 September 2015 - 07:24 AM
Fire for Effect, on 05 September 2015 - 06:53 AM, said:
Do you really think it would have been any diffirent with max 4 steel jags or only 4 228?
Dream on, the pain just would have been longer... since the steel jags /228 would have taken longer to wax you and you would have killed one enemy PUG more, but that would be the only difference....
If it would be 4 max then THREE groups on the other side of 228/ Steel jags would have been waxed 12:0 or 12:1. So you would have trippled the pain. Congratulations you have spread the epidemic threefold. great solution...
I'm sorry I'm not understanding the post? If its in reference to my original comment what I mean to say what that rolls happen, move along and play a great game
PS: the 228 and SRJ don't ALWAYS win. I've seen a defeat in their court
#576
Posted 05 September 2015 - 07:35 AM
We want to have a way for large, skilled groups to play together.
We want to have the ability to choose our server.
We want to have the ability to choose our game modes.
We want to have some form of skill-matching to create better matches.
We want to have a way for vet pilots to show their new friends the ropes, without getting mauled.
We want to have a way for small groups to be enjoyable.
We want to have the ability to play the mechs we want, when we want.
I could likely keep that list going for a bit.
The problem is that the solution for each of those problems is to subdivide the game population still further.
For instance, say you have a Tier 2, seven man team, that chooses conquest only, and is determined only to play on their local server. That means that the MM needs to find either a 5, or a 3 and a 2 to team with them, and those teams have to be comparable PSR, and they at minimum have to be willing to play conquest... all on the same server.
You split out the beans in that many different directions, and even at fairly busy times, the MM is going to have hissy fits. I don't know how many Tier 1 teams there are as a rule, but I suspect it's a small enough number that even being open to modes and servers won't alleviate the problem all that much.
But, what do you do? There are only really two answers. On the one hand, we could simply all hope that some mythical radical increase in pop with Steam will give the MM enough beans to make it all work. The downside to that is that the current situation is more likely to turn those new Steamites into fast exiting churn, than actually give a long term increase in the MM.
The other option? It's that something simply has to give. We can't have it all. One or more of our desires is going to have to get set aside. Which one is best? I'm not sure.
I'm highly in favor of letting solos opt into the group queue (and I'm not really inclined to PSR gate it either. Let low PSR solos work into low PSR group matches.) I think that will alleviate the problem, but not solve it. (Though, it would help with finding teams for odd size groups like 5 or 7).
Beyond that though, I think the bottom line is this. Especially with Steam coming, there must be a way for the new players to play in small groups (either together, or with a veteran mentor), and not simply get fed into the meatgrinder. That's got to be the first priority, because otherwise, the churn and chum will drive most to the solo queue, or off to whatever other game catches their fancy.
Edited by Enaris, 05 September 2015 - 07:38 AM.
#577
Posted 05 September 2015 - 08:16 AM
CainenEX, on 05 September 2015 - 07:24 AM, said:
PS: the 228 and SRJ don't ALWAYS win. I've seen a defeat in their court
It means the result will be the same if you have 12 or 4 of them on the other side. No they dont always win but unless you have something really capable on your side (which would have made it a challenge for them even if they have 12 man) forget about winning. so if you limit it to 4 man groups there will be 3 groups who will meet 228 or steel jags in the queue, so you tripple the whining since then 3 groups will be waxed with a high probability.
So the "solution" of 4 man groups just made it worse... great thinking...
We also won against them, but that s not the point. Probability is very high that they will win, even with only 4 man on a team.
Fact is as already stated its not the groups that are the problem...
#578
Posted 05 September 2015 - 08:18 AM
We've been saying for a long time that teams aren't a sum of their parts. It's not just average player contribution that contributes to match quality, it's also homogeneity vs heterogeneity.
#579
Posted 05 September 2015 - 08:42 AM
Enaris, on 05 September 2015 - 07:35 AM, said:
Here's a possible 3 queue solution:
1. Solo Queue - it stays exactly the way it is, no changes. PGI is unlikely to change it because it's MWO's best performing queue of all time, according to Russ Bullock. If it ain't broke, don't fix it.
2. Small Casual Group Queue - this is where new players group up. It's also where veterans can train newbies and where current players can introduce their friend/sibling/spouse/coworker/etc. to the game. Group size limited to 3. PGI could put a limit of only one Tier 1 or Tier 2 player per group (the Trainer) if they wanted.
3. Unit Queue - this for the minority of MWO players who like to organize, train and compete. Group size 4 and above. It could even be the CW queue after they fix CW.
#580
Posted 05 September 2015 - 08:58 AM
Kiriesani, on 04 September 2015 - 09:54 PM, said:
I say go back to Elo because on elo people didn't have to wait 6+ minutes for a match, much less 20+, like I know quite a few people have had to.
It's never been about Elo or PSR. It's always been about how matchmaker uses the rating, plus 3 game modes, plus 4x3, plus the population being split into 2 queues and CW, plus group size in group queue, and now 3 server regions. Elo worked fine, they should've just asked for tiering the rating out into buckets.
20 user(s) are reading this topic
0 members, 20 guests, 0 anonymous users