52 replies to this topic

[Nicholas Carlyle]

Circle jerk post, go!

[SgtKinCaiD]

Nicholas Carlyle, on 12 February 2014 - 06:09 AM, said:

Circle jerk post, go!

What ?

[NRP]

I see this type of thing all the time. That's why I laugh at all those who cry for tonnage limits, as if it will be some type of magical savior for their sorry arses. If/when tonnage limits go in (and people are still getting ROFL stomped) I wonder what the next complaint will be?

[WhiteRabbit]

NRP, on 12 February 2014 - 07:55 AM, said:

I see this type of thing all the time. That's why I laugh at all those who cry for tonnage limits, as if it will be some type of magical savior for their sorry arses. If/when tonnage limits go in (and people are still getting ROFL stomped) I wonder what the next complaint will be?

to be fair the map prob played a big role in that match....e.g. the ddc have a really hard time dealing with enemies higher/lower than themselves--> that would be 2 assaults that are nearly worthless if the enemy is in the right/wrong position.

Edited by WhiteRabbit, 12 February 2014 - 08:06 AM.

[NRP]

Perhaps, but Bishop's team was much faster (and Ember's have a hard time bringing their torso weapons to bear on uneven terrain too). A group of faster mechs (who don't derp) will almost always beat a group of slower mechs. Tonnage is not nearly the overwhelming advantage it is purported to be. I've seen this same result time and again on many maps.

Edited by NRP, 12 February 2014 - 08:14 AM.

[meteorol]

Bishop Steiner, on 11 February 2014 - 06:24 PM, said:

...(meaning almost guaranteed that we faced at least one other premade)...
Did our skill level really matter? If Elo is remotely working, it should not, as there should have been a rough analogue of each of us on the other team...

Man, since i bought a jagermech for the lulz, i don't think this is even remotely true.

I never ever piloted any mech outside the assault class since closed beta, which means i should have base elo for heavies.
I really have the feeling the matchmaker is ALOT worse if you are on base elo, don't ask me why.

Since playing my Jager i dropped 2 or 3 times against you. This either means you are playing close to base elo (which i don't believe), or the matchmaker screws base elo players big time.

What have i seen in my time with the jager in the base elo so far? Alot more stomps compared to playing with my assaults (which is above base elo for sure). More brutal stomps, i'm talking about 10 of 12 guys below 100 damage stomps. If i'm playing with my assaults, there are stomps aswell. But those are "almost everyone did above 150-200 damage, and they somehow snowballed us" kind of stomps. Not "3/4 of my team failed to hit a powered down atlas from 30 meters" stomps.

I read that premades are assigned to the alpha and beta squad in first place. If this is true, well then i don't think a premade on one team really guarantees a premade on the other (atleast in base elo).

Just yesterday i was in alpha squad as lone pug. One guy of the opposite team announced his premades homepage. I watched my team an was like "seriously?". Almost the whole team outside the alpha squad was using trial mechs. 7 of 12 guys in trialmechs.
Enemy team (after a heartcrushing stomp within like 2 minutes)? Alpha: 4 Victors (the premade that announced their homepage), Beta: 2 HGN 733C , 2 DDC (all of them steiner, which alteast "could" mean they were premade awell). They had 2 trial mechs in charlie though.

Dunno why this is the case, but the matchmaker really seems considerably worse when close to base elo (atleast for me).
Upside is, it is alot faster when dropping in my jager. I get into games almost instantly. When dropping in my assault mechs, i often have to wait up to a minute or more for a game.

If i had the honor to drop against you, other base elo player probably had it too. Someone who plays this game for a week is not prepared to face someone like you or guys you play with.

On a side note:
This really isn't ment to talk down your match. Winning when seriously outnumbered and/or outgunned is always a cool thing which deserves respect.
It's just that the matchmaker is so much worse for me when playing my jager than with my assaults, that i'm about to drop the jager and just go on with my trustworthy assaults. Since i played against you twice (or so) with it, i borrowed your thread for my little rant.

[Bishop Steiner]

SgtKinCaiD, on 12 February 2014 - 03:05 AM, said:

Sorry it was not my first intention. It's just that, contrary to the numerous threads on the subject, the problem of the MM and stomps is not due to tonnage mismatch. And i'm sick to hear from PGI that the MM is working as intended and that tonnage limit coming in april will magicaly resolve stomps. It will not.


See, you nailed it : those players you mentionned, guess what ? They are new/bad players. Why are they facing regular veterans ? Why don't they have their own match with people of their level so that they can enjoin the game and not be crushed in match where they don't belong ? Because they are here to compensate for the rest of the team : to level down their team.
And talking about the rest of the loosing team, even though it was a one sided match, they still managed to get 5 kills with relatively low damage count and i can fairely bet that the ones who got the kills were the last to die. How can I say that ? Because i've experienced this situation countless times since the introduction of ELO ratings.

Actually..... again, you miss the part.... Elo is also taken into account.... the issue is You got 12 kids on each side. Lets say 4 are High Elo, and 8 are nooberNoobs on each team. How do you guarantee that the High Elo face off? You can't which is why Elo don't work when you are using it in teams.

That said, according to how the MM is SUPPOSED to work, you will be withing a very small window on tonnage and team average Elo. Said Elo has more to do with W/L than kills (contrary to what many of those touting their high elo seem to think... being solo rambo god and getting your team killed while you statpad doesn't boost your Elo).. That said...... I am willing to bet I have "been there" a little more than you have, and really don't think you can read into who did what on the match, sorry. Best Stalker pilot in the game gets wolf packed by 4 average but coordinated Lights, he will lose almost every time. Has nothing to do with his skill level. Simple fact is the Stalker in question is too gimped mobility wise to do more than get glancing hits on his opponents (which might kill one or two in some scenarios).

Some times it has nothing to do with tonnage, nothing to do with skill, and everything to do with who bothers to try playing as a team. This match was the latter.

[WhiteRabbit]

NRP, on 12 February 2014 - 08:11 AM, said:

Perhaps, but Bishop's team was much faster (and Ember's have a hard time bringing their torso weapons to bear on uneven terrain too). A group of faster mechs (who don't derp) will almost always beat a group of slower mechs. Tonnage is not nearly the overwhelming advantage it is purported to be. I've seen this same result time and again on many maps.

Well faster mechs have fewer problems getting out of a bad spot than slower mechs--> any problems the embers had couldn't have lasted that long. Faster teams only beat the slower team almost every time when the slower teams uses only close-range weapons and can't shoot worth shit. If both teams know what they are doing it comes down to the execution of the plan....which in pugs is mostly "ahhhhh shoot the ****** that shoots me..derp derp derp " That being said, with the right plan every team can beat anyone as long as their tactic is better than their enemie's.

[Bishop Steiner]

meteorol, on 12 February 2014 - 08:46 AM, said:

Man, since i bought a jagermech for the lulz, i don't think this is even remotely true.

I never ever piloted any mech outside the assault class since closed beta, which means i should have base elo for heavies.
I really have the feeling the matchmaker is ALOT worse if you are on base elo, don't ask me why.

Since playing my Jager i dropped 2 or 3 times against you. This either means you are playing close to base elo (which i don't believe), or the matchmaker screws base elo players big time.

What have i seen in my time with the jager in the base elo so far? Alot more stomps compared to playing with my assaults (which is above base elo for sure). More brutal stomps, i'm talking about 10 of 12 guys below 100 damage stomps. If i'm playing with my assaults, there are stomps aswell. But those are "almost everyone did above 150-200 damage, and they somehow snowballed us" kind of stomps. Not "3/4 of my team failed to hit a powered down atlas from 30 meters" stomps.

I read that premades are assigned to the alpha and beta squad in first place. If this is true, well then i don't think a premade on one team really guarantees a premade on the other (atleast in base elo).

Just yesterday i was in alpha squad as lone pug. One guy of the opposite team announced his premades homepage. I watched my team an was like "seriously?". Almost the whole team outside the alpha squad was using trial mechs. 7 of 12 guys in trialmechs.
Enemy team (after a heartcrushing stomp within like 2 minutes)? Alpha: 4 Victors (the premade that announced their homepage), Beta: 2 HGN 733C , 2 DDC (all of them steiner, which alteast "could" mean they were premade awell). They had 2 trial mechs in charlie though.

Dunno why this is the case, but the matchmaker really seems considerably worse when close to base elo (atleast for me).
Upside is, it is alot faster when dropping in my jager. I get into games almost instantly. When dropping in my assault mechs, i often have to wait up to a minute or more for a game.

If i had the honor to drop against you, other base elo player probably had it too. Someone who plays this game for a week is not prepared to face someone like you or guys you play with.

On a side note:
This really isn't ment to talk down your match. Winning when seriously outnumbered and/or outgunned is always a cool thing which deserves respect.
It's just that the matchmaker is so much worse for me when playing my jager than with my assaults, that i'm about to drop the jager and just go on with my trustworthy assaults. Since i played against you twice (or so) with it, i borrowed your thread for my little rant.

Oh, I'm sure it's broke...even if working as intended. As I said, Elo really does not work for a team score scenario. (Possibly in dedicated 1q2 man where everyone has similar skills and drops together every time, it MIGHT work somewhat)

Because Let's say....oh (RNG here) you Teams average Elo is........3000. And so you get Teamed up against a team (pretending it gets close here.,,,, suspension of disbelief, please) with a 3200 Elo average.

Sounds groovy..... except when averaging, you can get to that number in a number of ways. Everyone may indeed have about a 3000 score....... or 6 may have a 4000 and 6 a 1000, or any permutation. Even if it does attempt to find an Elo analogue for each player, there is no way to guarantee that those players will face each other in a team setting, as "good" players quickly target the weak and pick them off, enhancing their numeric advantage to better guarantee a win.

And if our stats weren't so easy to game and bias, it might mean a lot more. And if success in 12 man actually translated to success PUGing, or vice versa.

[Bishop Steiner]

WhiteRabbit, on 12 February 2014 - 09:04 AM, said:

Well faster mechs have fewer problems getting out of a bad spot than slower mechs--> any problems the embers had couldn't have lasted that long. Faster teams only beat the slower team almost every time when the slower teams uses only close-range weapons and can't shoot worth shit. If both teams know what they are doing it comes down to the execution of the plan....which in pugs is mostly "ahhhhh shoot the ****** that shoots me..derp derp derp " That being said, with the right plan every team can beat anyone as long as their tactic is better than their enemie's.

Both you and NRP are right. Speed is often a huge advantage. But that advantage can be negated by several factors, such as coordination (where this Opfor semi failed, as their individual lances did seem to stick roughly together...they just didn't coordinate between the lances) and terrain, or in this case, lack thereof. Had the OpFor congregated in almost any of the more open sections of the map, their overwhelming range and firepower advantage would have won with even middle of the road targeting skills.

They actually did start in a nice open bowl that punished the crap out of us if we got adventurous. But again, they lacked company cohesion, and allowed themselves to be goaded out of the killing field. Individually many of them fought well, but 1 mech vs 3-4 is hard going, period. They lost more on the tactical level than the individual twitch skill level.

WhiteRabbit, on 12 February 2014 - 08:01 AM, said:

to be fair the map prob played a big role in that match....e.g. the ddc have a really hard time dealing with enemies higher/lower than themselves--> that would be 2 assaults that are nearly worthless if the enemy is in the right/wrong position.

actually it would depend where you fought on the map, the southwestern areas have several spots where you have open terrain for 500+ meters in any direction. Our force mostly had popguns which were next to useless at that range. Had the engagement occurred near the old radio tower derp spot or the i9/h9 sector, it indeed would have made a difference though.
Next strawman?

Edited by Bishop Steiner, 12 February 2014 - 09:31 AM.

[Bishop Steiner]

NRP, on 12 February 2014 - 08:11 AM, said:

Perhaps, but Bishop's team was much faster (and Ember's have a hard time bringing their torso weapons to bear on uneven terrain too). A group of faster mechs (who don't derp) will almost always beat a group of slower mechs. Tonnage is not nearly the overwhelming advantage it is purported to be. I've seen this same result time and again on many maps.

I've also been on the other end of the Steiner Scout roll. Had a team with 7 Atlases and one mech below the heavy range absolutely crush my underweight medium/light company, because they parked in a nice wide open space and wouldn't be budged. Even when you tried to swarm into them, they didn't mind a little friendly fire because they had the armor to shrug it off.

One thing about Skirmish mode...... no cap is good most of the time, but it also means you can't use it to force the oppositions hand. (Then again, I have also had to try to dislodge a DDC and BoomJager company off our cap in 12 man..... that was just ugly)

[Trauglodyte]

23 damage between a D-DC, A1, and a Stalker Champion is hard to handle. Oof! I tell you, I'm happy that the high ELO guys are getting games faster now but it sure screws the rest of us by having our games filled with drooling neck beard bads. I'm getting games like this every other one now. It is awful. My worst to date was a 12-0 smashing last night where my side had a d/c and 6 Champion mechs of which none of them broke 95 damage.

[Fuerchtegott]

Read the dev postings about matchmaking again, maybe you get the riddle solved by yourself.
If not, write me a pm and I tell you what happend and why you got such a nice tonnage disadvantage and what you can do against it.

[WhiteRabbit]

Bishop Steiner, on 12 February 2014 - 09:30 AM, said:

Snip

Next strawman?

here you go :

Spoiler

i just cant finding my first one...did you nick it?

[Bad Andy]

congrats you beat bads

/golfclap

[NRP]

Bishop Steiner, on 12 February 2014 - 09:34 AM, said:

I've also been on the other end of the Steiner Scout roll. Had a team with 7 Atlases and one mech below the heavy range absolutely crush my underweight medium/light company, because they parked in a nice wide open space and wouldn't be budged.

Artillery strikes. And PPC Cicadas. And ERLL Ravens.

In my experience, Steiner Scout Lances are a big problem on brawler-oriented maps like HPG and Crimson. On larger open maps, they are actually at a disadvantage to a coordinated lighter unit. I actually relish this type of match up, as it provides a good test of the skills and decision making abilities of both me and my team. Sure we get crushed sometimes, but this type of match up really gets the adrenaline flowing. If you somehow yank a victory out of this situation, it is the most epic rush ever.

And it earns you respect. Whining about tonnage imbalances and how the awful matchmaker screwed you does not.

[Bishop Steiner]

Bad Andy, on 12 February 2014 - 10:12 AM, said:

congrats you beat bads

/golfclap

umadbro?

(yephemad)

Don'tBeMadBro.

If our collective Elo ever drops so bad we find ourselves in Terribad Territory, we'll be happy to give you an encore performance.







*****This Message has been brought to you by the "Don'tBeADoucheBro^tm" Council for Better Gaming.

Edited by Bishop Steiner, 12 February 2014 - 11:00 AM.

[SweetJackal]

All I can say on all this MM and ELO talk is that ELO doesn't work when it's an average of teams. Instead ELO needs to be used as a metric to balance each player.

If a match's average ELO is 1250 then the ELO of each player in the game -needs- to be within a small window of that amount. If +/- 100 ELO was an acceptable variance then that would allow a 1200 ELO player to be in a match with a 1300 ELO player and no greater variance.

What does happen with ELO averaging the team is that you get cases of Base-ELO players getting placed in matches with players on the Average Curve.

In the end the tighter you make ELO restrictions the larger the player pool you need to create matches. Using strict Player ELO for matches instead of Team ELO Average requires a much larger player pool to work. One might say that a poor reflection on Match Maker is a poor reflection on Player Pool size at any given moment.

How ELO is used to determine balanced matches matters much much more than any supposed allowable variance between teams.

ELO also has some supreme faults. First is that it is a curve. Being a curve means that the further you get from the average ELO value the less players there are that match your rating, the less players there are that are valid matches. These extreme ends of the curve are what I call the ELO Graveyard, where player accounts go to die.

There are a few ways this can be dealt with. First is dynamic banding to match ELO allowances, the longer a player tries to find a match the more loose the matchmaker becomes. This is bad as it cheapens the experience in High ELO by creating matches with extreme variance compared to the tight variance found at Average ELO. It also creates long wait time. This is the system used by games like League Of Legends and DOTA2.

I prefer the second means. Often ELO is capped at 0, no matter how bad a player is they will not get dropped down to such a level that they will be unable to find a match. This removes the lower section of the ELO Graveyard on that curve. What needs to happen is to give the same treatment against the upper end of the ELO Graveyard. As Follows:

You provide an upper tier cap to ELO. This creates a pool of players that are treated as the same tier and level even if they there is a variance in the skills. It makes matches easier to find for these players. Extra ELO earned is put into a Buffer Pool so when they lose ELO they are not knocked out of that bracket. This is the system used by Starcraft 2's Ladder System, this is the Masters and Grand-Masters.

Such a system allows strict matchmaking to continue to work and keeps searching times for games low. The cost is that those fresh into the bracket may be overvalued in ELO compared to the top players in that bracket but more tiers or brackets could be added as the player base grows and the ELO Cap begins to get bloated with players.

Anyways, my two digital cents (since Canada no longer uses Pennies)

[Bishop Steiner]

SuckyJack, on 12 February 2014 - 10:34 AM, said:

good stuff

might need a tweak here or there by people who know this stuff better n me, but pretty much no argument here!

[SweetJackal]

Bishop Steiner, on 12 February 2014 - 10:35 AM, said:

might need a tweak here or there by people who know this stuff better n me, but pretty much no argument here!

Thanks for the vote of confidence. Any values given are for demonstration purposes only, ELO only has a value compared against itself and as such are completely subjective. It's a floating number system, so the concept of how it is used is what is more important to get across