Jump to content

The Matchmaker, The Elo, And The Cesspool


7 replies to this topic

#1 Kyynele

    Member

  • PipPipPipPipPipPipPip
  • The Vicious
  • The Vicious
  • 973 posts

Posted 19 February 2014 - 09:19 AM

TLDR; The difference between low and high level play is surprisingly large.

I'm afraid this might come out looking like a humblebrag, but decided not to care. :wacko: This was somewhat of an eye opener for me, and I just have to share it.

I consider myself a pretty average player. I started playing a year ago and have played about 6000 matches. My win/loss ratio is 1.22 and my kill/death ratio is at a pretty steady 1.71, dying 20 times in a row or getting 20 kills in a row without dying would likely not swing it in either direction. I solo about 50% of the time, play with 1 friend 40% of the time, and the rest 10% goes to 3-4 person lances. I get good matches, I get bad matches, sometimes I'm useless and sometimes I save the day for the team.

I recently introduced a friend of mine, a hardcore RTS/FPS/MOBA player to the game, and to get him off to a good start, provided to him some pretty meta friendly builds he shouldn't have too much trouble to get into the game with.

Sure enough, the results he was getting after 2 weeks were clearly above average. In some conversation it came up that he had scored some 9 kill games and was astonished I had never managed that. After that I played a bunch of games grouped up with him. Most of the games he was hysterically laughing about the absurdity of the massive focus fire situations he got into. He died in every match, did manage to get a couple kills here and there, but never scored much over 300 damage.

This got me into thinking - how big can the difference be? I've played long and regularly, and the game has likely gotten steadily harder, so slowly that I haven't been able to notice it myself. Apparently the game I experience every day might not be the same for everyone, perhaps there are enough players in the game to actually have different levels of play.

After pondering this for a couple of weeks, I decided to find out and made a new account to see what the entry level gameplay is.

I chose the trial Cicada and played 8 matches with it, without any efficiencies, naturally.

The match start screens looked worrisome, there were many phoenix (P) variants and founders on the list on both sides, maybe I was mistaken and this would be just the same as the usual games I see. Perhaps I had also just put myself in a major disadvantage for driving a hot energy based mech without even the basic efficiencies or modules.

The actual matches were mindboggling. There were solo mechs cruising in the weirdest places, there were light mechs that stood still while shooting, and there were assault mechs that you could go behind and shoot in the back repeatedly without them turning around. In the game I'm used to, if you stay behind an assault in a light, trying to take multiple shots in the back, that assault will do a JJ turn and put an instant pinpoint alpha to your leg. There were some meta poptart assault builds, but they weren't mass poptarting. They barely even jumped.

I took the first couple matches easy, but soon found out that for example dual gauss or AC40 jagers here weren't nearly as dangerous as I was accustomed to, and I could play pretty arrogant rambo game killing stuff pretty much as fast as the heat levels allowed.

After 8 matches, my KDR stood at 9.0, 27 kills for 3 deaths. Actual average damage 522 per match.

The difference in the gameplay is HUGE. And like said, I'm nowhere near a top-tier player. I get my ass handed over often enough to be very aware of that. ;)

However. The thing that surprised me most, was how NICE people were over there. There was friendly chitchat in nearly every match, and often at the end of a match, the losing side was the ones that went "good game!" The games I play on my regular account, on a good day the chat can go something like this:

opponent1: ---some german spam looking for lonely people without a group---
* fighting *
teammember1(dead): what a stupid noob team
teammember1(dead): you all suck balls have fun losing lol
teammember1 has disconnected
teammember2 has disconnected
* defeat 0-12 *
opponent2: ggclose
teammember3 has disconnected
opponent3: ggclose
opponent4: gg
opponent5: LORD OPPONENT5
teammember4(dead): what a shit team
opponent6: gg

I'm a bit sad that I can't play in that relaxed atmosphere on my main account, but am happy that the new player experience isn't nearly as bad as I expected it to be.

What I personally got out of this experiment:

1) A pretty strong faith that the matchmaker actually does something reasonable, even if it does sometimes provide a considerable amount of one sided stomps. Those are nowhere near what the skill differences in teams could be.

2) I have a much clearer idea of where I stand now, comparing to where I've likely started.

3) I shouldn't worry about my KDR too much. I've played enough and am average enough that the matchmaker manages to put me up against similar skill level people, and the games are supposed to be hard, not a means to stats padding. If I wanted to have stupid high ratings, I should just start that new account, stick to the meta and not try to level up every aw(e)ful mech in the game.

4) You can't tell much about other people's skill level or a build's efficiency from their screenshots of match results, kill counts, KDRs (especially after only a handful of matches in a mech) and so on. They can be playing a completely different game.

5) There are so big differences in the gameplay, that judging the effectiveness of weapon systems and builds is HARD. In the lower tiers, I'd imagine you can do just fine in a very sub-optimal weird build. Some may even consider a build OP, that would be laughed at in higher tier games.

6) Extra appreciation for truly good, close to even matches. :D

#2 Joseph Mallan

    ForumWarrior

  • PipPipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • FP Veteran - Beta 1
  • FP Veteran - Beta 1
  • 35,216 posts
  • Google+: Link
  • Facebook: Link
  • LocationMallanhold, Furillo

Posted 19 February 2014 - 09:29 AM

Yes. The ability curve CAN be that dramatic.

#3 Josef Nader

    Member

  • PipPipPipPipPipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 2,243 posts

Posted 19 February 2014 - 09:36 AM

It's something I say quite often when playing. I'm really not sure if my ELO is so low that I can bring pretty much whatever I want and do well, or if I'm skilled enough to make non-meta builds work well in the face of the meta. I certainly see the meta builds often enough, and they seem to do well score wise, but the fights between my goofy builds and the pop tarts seem to be pretty decent for the most part, and it's rare that I die without significant attrition or facing multiple enemies at once.

Frankly, I'm perfectly happy being a low ELO player, if that is indeed what it is. I'd much rather be able to bring whatever build I want and with enough skill make it do well then play the same thing for 200 hours and have stupidly high stats.

#4 Sandpit

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • Veteran Founder
  • Veteran Founder
  • 17,419 posts
  • Facebook: Link
  • Twitter: Link
  • LocationArkansas

Posted 19 February 2014 - 10:26 AM

+1

I'm sure someone will be along shortly to talk about how you're somehow wrong though. There's a LOT of factors that go into MM which gives almost every player a unique experience. A lot think their individual experience is representative of the entire population though.

#5 Blissful Negligence

    Member

  • PipPip
  • Ace Of Spades
  • Ace Of Spades
  • 39 posts
  • LocationSan Diego

Posted 19 February 2014 - 10:49 AM

Good post. It will be near impossible to make a flawlwess matchmaking system and I understand this. Something I have been experiencing lately though is winning 2-3 games in a row and cranking out significant damage outputs, then suddenly finding myself in the next game with 4-6 trial mechs on my team. Players that are no question new(er) to the game as the damage they put up is typically less than 100 pts. Then looking at the enemy on the other side, more often than not they have 2 or less trial mechs and outweigh my team somewhere in the 200-300 ton range. It makes me feel as though the MM is actually settting my team up to fail in order to "even out" the win loss ratio and give the illusion of balance. Is anyone experiencing similar situations as they play?

Edited by SLICKNIFTY, 19 February 2014 - 10:51 AM.


#6 BillyM

    Member

  • PipPipPipPipPipPipPip
  • 530 posts

Posted 19 February 2014 - 11:14 AM

What you are seeing is right on the money. ELO is actually working quite well on putting people in their skill range in 9 of 10 cases, but it gets hung up with tonnage and groups.

-Currently groups in 4 metabuild assaults will enjoy a 2.5w/l and average a 170ton weight advantage. Those same players in balanced classes of challenging builds (trials) instantly fall back to a 1.0w/l and either even or a weight disadvantage.

Pure solo queues would be EXTREMELY balanced games for all players.

...and Lords are known for creating alt accounts so they can dip further down into the pool of players. Deplorable stuff...

--billyM

#7 Sandpit

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • Veteran Founder
  • Veteran Founder
  • 17,419 posts
  • Facebook: Link
  • Twitter: Link
  • LocationArkansas

Posted 19 February 2014 - 11:35 AM

View PostSLICKNIFTY, on 19 February 2014 - 10:49 AM, said:

Good post. It will be near impossible to make a flawlwess matchmaking system and I understand this. Something I have been experiencing lately though is winning 2-3 games in a row and cranking out significant damage outputs, then suddenly finding myself in the next game with 4-6 trial mechs on my team. Players that are no question new(er) to the game as the damage they put up is typically less than 100 pts. Then looking at the enemy on the other side, more often than not they have 2 or less trial mechs and outweigh my team somewhere in the 200-300 ton range. It makes me feel as though the MM is actually settting my team up to fail in order to "even out" the win loss ratio and give the illusion of balance. Is anyone experiencing similar situations as they play?

I think one of the things that happens is "good" new(er) players win a lot of matches early on and quickly so their Elo jumps putting them into a higher bracket more quickly than the average player. So then you see stuff like this happening. At least from my observations and reading through all the "MM" threads this could be one of the "culprits"

#8 Joseph Mallan

    ForumWarrior

  • PipPipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • FP Veteran - Beta 1
  • FP Veteran - Beta 1
  • 35,216 posts
  • Google+: Link
  • Facebook: Link
  • LocationMallanhold, Furillo

Posted 19 February 2014 - 11:36 AM

View PostBillyM, on 19 February 2014 - 11:14 AM, said:

What you are seeing is right on the money. ELO is actually working quite well on putting people in their skill range in 9 of 10 cases, but it gets hung up with tonnage and groups.

-Currently groups in 4 metabuild assaults will enjoy a 2.5w/l and average a 170ton weight advantage. Those same players in balanced classes of challenging builds (trials) instantly fall back to a 1.0w/l and either even or a weight disadvantage.

Pure solo queues would be EXTREMELY balanced games for all players.

...and Lords are known for creating alt accounts so they can dip further down into the pool of players. Deplorable stuff...

--billyM
Are you kidding? I welcome the "Lords" in any drop I am in. If they are as good as they say are, it would be an interesting match to be part of.





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users