104 replies to this topic

[Phoenix 72]

justcallme A S H, on 03 February 2019 - 10:13 PM, said:

Either that or it is some great conspiracy and the MM is just matching good players on the same team all the time

Well, of course it is not all the time... If my team wins, it is all my own personal skill... If my team loses, it is the MM stacking the odds against me...

[justcallme A S H]

Darakor Stormwind, on 03 February 2019 - 11:45 PM, said:

Well, of course it is not all the time... If my team wins, it is all my own personal skill... If my team loses, it is the MM stacking the odds against me...

Yeah that's the most common thing I see on YoloQ... Usually the guy saying it was a bad game or talking the most is the dude with the lowest score/damage. Literally every time.

[Phoenix 72]

I blame the match maker... I am 5:0 today... Still have not reached an even win/loss, but I think I will get there eventually. The month is still around for quite a bit.

[MischiefSC]

Playing a lot absolutely does NOT equal being good. Perfect practice makes perfect - 5,000 matches of making poor choices just makes people really, really efficient at playing poorly.

Your W/L reflects your average ability to drive wins. How you drive wins is irrelevant. Maybe you try to pug-wrangle every QP match, maybe you're just a really strong player, maybe you usually play good mechs and builds, maybe you're really good at reading the map. Even a the best players still lose but they lose less often.

That's why Jarls list is valuable as you see someones overall trends over thousands of matches. Match score is a **** indicator as it largely reflects damage farming. KDR can be gamed by absolutely terrible behaviors like running away and powering off when it doesn't look like your team will win or just being focused on stealing kills.

Above a 1.2 w/l is solid. It's a good contributor. Looking at matchscore, damage, KDR will help tell you how they are helping their team win.

If you're not winning as much as you'd like then re-evaluate how you play and what you play and look to get better. Watch streams and videos of good players, try to play with good players and learn how they get the results they do. We're all playing in the same matchmaker. You can't blame it - it's the same ocean of **** we're all swimming in. The only real difference is the degree of GIT GUD people have invested.

[MischiefSC]

justcallme A S H, on 03 February 2019 - 10:13 PM, said:

I play SoloQ 90-95% of my matches each season. My WLR is always over 2.0 on a global level and often higher. That's a hell of a lot more winning than losing.

1 player in 12 can and does have THAT much impact on a single game despite what the uneducated around here claim. More likely because they cannot accomplish this themselves. I will concede it is not always possible to carry but it is more common to do so than than it is to not.


Either that or it is some great conspiracy and the MM is just matching good players on the same team all the time

And for the last LOL... Burning in here trying to talk up his game again. I guess the burn from a few months back has healed and ready for more?

I used to believe that group queue skewed stats. However given the low GQ population and the huge SQ population GQ is really only skewing a tiny percentage of results both out of total matches and for a very limited group of players. I did the math a couple of months back and it ended up being a fraction of 1% of players who had a significant number of their matches influenced to any real degree by GQ. The idea of the 'this player isn't that good but his stats are inflated by playing in GQ' impacts a literal handful of people out of many thousands and their success only negatively impacts a small segment of people who play in GQ, which is in turn a small segment of the games population.

Hacking is very rare in MWO and still probably has a larger net impact on player stats than GQ. The reality is that the matchmaker isn't bad for the population size we have and number of queues. There's a few ways to improve it, sure, but the need to match tonnages, the fact that you pick a mech before map/mode resulting in otherwise good players/mechs ending up in a 'bad position' and such are going to mean that you're always going to have some big swing in QP matchmaking. We'd need 4x or 5x our current population to get a matchmaker that was really ladder viable for QP even with something super tweaked.

Flip a coin 1,024 times and at some point you'll likely get heads 10 times in a row. Yet periodically getting more tails than heads will balance out so you'll only vary by 1 or 2 in either direction from 512 heads results, 512 tails results. Streaks happen but averages are just that - averaged out results. GIT GUD is shaving your coin to make it come up heads more often. You'll still lose plenty but on average you'll be able to drive more wins.

[thievingmagpi]

sometimes you get stuck in a loop playing with the same people. take 30 seconds before the next drop so you can hopefully get sorted into a different bunch of people.

[Natred]

Yeah, group q match maker is a bit rough for the new crowd. Just got to bear with it and keep grinding learning. Work on map knowledge, battle awareness, find some optimized builds. Team up with some good teammates. Find builds that do not rely on lock on weapons cause 90 percent of the time they do not work on good players. Get better hardware.. avoid lrm groups.

[Variant1]

If you come to play quick play to win your gona have a bad time. Instead take off the meta build and try something new, the mechlab offers plenty of customization so go nuts. Play for fun and dont take quick play seriously. If you want to win then join a unit and play with premades. best advice i can give and im no pro

[Knight Captain Morgan]

I've only found a couple tactics that seem to work sometimes for breaking losing streaks.

1: change to a mech you don't care about. In the next match be extremely aggressive to the point of being the 1st or at least 1 of the 1st killed. Then leave match and switch to the mech you do care about, enter que with selected mech. If nothing else, it gets you away from the same synch dropping premade team cause they're all still clubbing seals in the match you just left.

2: exit game and exit steam. Log onto gog (or other steam competitor) and play another game for a bit. Then come back to MWO. Steam tracks your gaming activity. Rage quitting from one of their games and then playing a competitors' game causes them to send red flag warning to the rage quit game company in question, making your next gaming session "easy mode".

If neither of these work, try both in succession and see if that doesn't break the losing streak. If still nothing then report the wall climbing commando from that canyon network game for cheating because commandos are NOT supposed to get jump jets...

Edited by Andres Gomez, 08 February 2019 - 02:39 PM.

[Nightbird]

Andres Gomez, on 08 February 2019 - 02:39 PM, said:

Don't forget to rub the machine for good luck lol....

[Knight Captain Morgan]

Nightbird, on 08 February 2019 - 03:02 PM, said:

Don't forget to rub the machine for good luck lol....

Rub what machine with what?

[Angel of Annihilation]

justcallme A S H, on 03 February 2019 - 10:13 PM, said:

I play SoloQ 90-95% of my matches each season. My WLR is always over 2.0 on a global level and often higher. That's a hell of a lot more winning than losing. 1 player in 12 can and does have THAT much impact on a single game despite what the uneducated around here claim. More likely because they cannot accomplish this themselves. I will concede it is not always possible to carry but it is more common to do so than than it is to not. Either that or it is some great conspiracy and the MM is just matching good players on the same team all the time And for the last LOL... Burning in here trying to talk up his game again. I guess the burn from a few months back has healed and ready for more?

Nope. Not really. People seem to forget that while flipping a coin theoretically will result in 50% heads showing up, that just through random chance you might flip 10,000 coins and get something like 60-70% show up heads. Same happens for tails though, sometimes your just unlucky and 60-70% end up tails. My point is that it is very, very possible that high win or loss rates can just come about through random chance without skill being any factor what so ever. Try it some time. Use a random number generator to roll a 2 sided dice a couple thousand times. Count 1s as a win and 2s as a loss and total up the number of wins you get. Then repeat the experiment a couple dozen times. You will find all kinds of variations ranging from probably upwards of 75% coming up as wins to maybe as low as 25% coming up as wins. Aside from time consuming, it is easy and verifiable with your own eyes so you don't have to take my word for it.

Also people who subscribe to the theory that win rate is determined by skill fail to take into account that when your talking 24 total players divided randomly into two teams, chances are there is someone at least as good as you are on the other team. If that is the case, then your contribution to the win is cancelled out by their contribution to the win which leads right back to the fact that the match is mostly likely decided by random chance, not your personal skill level.

I will concede however that if you really are JUST THAT DAMN GOOD, say in the top 3-5% of all players in terms of skill, then sure, you are likely contributing to your win rate. At that point, the statistical chance of someone being your equal on the other team is very, very low, but for the other 95% of us, the vast, vast, majority, stastically speaking, there will always likely be at least one player on the enemies side that is as good or better than us.

[The Masked Luchador]

I feel you OP. I've played about 20 matches tonight and my team has won 2 of them. It's very disheartening and it's not a matter of "get gud" or builds. There has to be something wrong with matchmaking. I mean law of averages you should be somewhere from 40-60% win/loss not down around 0-10%. It wouldn't be such a big deal if there wasn't such a big variance between rewards in winning and losing and let's face it, people do not like to lose. I usually still manage to be one of the top 3 in damage done and I was just fine until I hit tier 3 but now the game is just not enjoyable to play. I have some mechs that I still want to skill up and we are still over 6 months away from MW5 so I'm not sure how (or why) I want to proceed other than the fact I like MechWarrior as a franchise and this is the only way to get my fix right now.

Edited by The Masked Luchador, 09 February 2019 - 12:04 AM.

[John Bronco]

Totally agree the exp gain for being on the losing side is terrible, if you play and contribute, say 100MS, you should get at minimum enough exp for a skill node.

[ShiverMeRivets]

Losing (and winning) streaks happen more often that they should statistically, because consecutive matches are not independent. Think of what happens after a game - most of the players immediately queue for another game, and get sync-dropped with the same players. The mixing between the players of the previous teams is limited because of chassis weight, and PSR, and whatever else is taken into account - the match-making solution will converge to something similar to the previous solution.

It is not a huge effect, but enough to create statistically improbable streaks against your assumption of a purely random and independent allocation of players to the teams. It is not biased, so the losing streaks should be balanced by winning streaks, excluding your personal performance effect on the results.

[adamts01]

MischiefSC, on 04 February 2019 - 09:38 AM, said:

That's why Jarls list is valuable as you see someones overall trends over thousands of matches.

I've never actually looked at that site till now. Pretty cool. Is there a way to single out the community of light pilots, on that site or another? I've never played competitively but I always used to think I was decent. It would be cool to see how I stack up against the pro light players.

[Acersecomic]

Loodacris, on 02 February 2019 - 09:56 AM, said:

There has to be something wrong with the MM...I've been on a losing streak the past few days and it's just plain old killing my desire to play. Once I moved up to tier 3 it just has been a never ending snowball of defeat after defeat.

Suppose it's a good thing...there are a few other projects i've been putting off for a while.

I don't think it's the MMs fault. There's just too many (snip) in the queue.

[adamts01]

Acersecomic, on 09 February 2019 - 11:55 AM, said:

I don't think it's the MMs fault. There's just too many (snip) in the queue.

There are some great threads criticizing the match maker algorithm. It seems that PGI treats all players in the same tier alike, which is obviously flawed. Dissecting 100 stomps, you'll find that the winners have obscenely stacked match scores and W/L ratios. Things where honestly better with elo.

All that said, you are the only constant, and over time, players end up having a pretty accurate W/L based on their skill. But yeah, the MM in this game is possibly the worst in the business.

[RickySpanish]

adamts01, on 09 February 2019 - 12:44 PM, said:

There are some great threads criticizing the match maker algorithm. It seems that PGI treats all players in the same tier alike, which is obviously flawed. Dissecting 100 stomps, you'll find that the winners have obscenely stacked match scores and W/L ratios. Things where honestly better with elo.

All that said, you are the only constant, and over time, players end up having a pretty accurate W/L based on their skill. But yeah, the MM in this game is possibly the worst in the business.

Any MM lives or dies by the game's population. Sadly MWO has a teeny tiny one, so the odds are not in its favour. This is of course not entirely due to bad luck - plenty of misteps have happened along the way with the game's design to end up with a niche audience. There are similarities between MWO and the other game I played frequently - Elite: Dangerous. Apparently it's easy to lose touch with what players want and to design features that turn out to be a waste of time in terms of player draw. Hopefully MW5 will attract more players to MWO but honestly, I think at this point we're caught in a spiral. Perhaps MWO will be rebuilt as a new game after MW5.

[Christophe Ivanov]

I just had enough and quit playing this game months ago. Just not worth my time, and especially my money. As for the lame "GIT GUD" moniker, that's an excuse for what is not wrong with this game. So many variables make this game less exciting than it used to be and MM is just one. My biggest nit pick is the constant micro managing this game that goes nowhere each and every month they bork the game.

I have seen many threads for pluses and minuses of just about everything in this game with most are negative, but some are positive too. I understand PGI is trying, but again I think too hard and not in the right way(s) to better this game for all. We all have different play styles, different preferences for which mech and type of weapons to use. As such, there is no easy way to fix this game as it is which is why in my opinion, they should have left it alone as it was before the engine desynch several years ago. Was it balanced? No, was it easier? Maybe, but it was much more fun to play than it is now.
I'm (or was Tier 2) and as you go up, it's less fun to play and spend my time on.
I know there are others who agree and disagree with any or all of this. But I can tell you now, just look at how many have moved onto something else to do. As with anything in life, things never are the same and folks move on. Same with this game and the way I see it, Mechwarrior's only saving grace is the upcoming MW5 which so far looks promising. Will it revive this genre? Only time can tell.

Good luck folks.

Edited by Christophe Ivanov, 09 February 2019 - 01:56 PM.