141 replies to this topic

[Knownswift]

The irony is anyone expecting this games' token infinite kiteyman to be less biased than the player who has played like 60 matches in light mechs over the last several years.

Y'all both ****** biased.

[Magnus Santini]

I read through the patch notes twice and couldn't figure out the question: did the patch fix what the last one broke in the MM and resulting PSR? Forgive me for thinking that was the big issue of the moment and might have been addressed more specifically and visibly. Seems like changing the size of groups in quickplay does not address the problem at all.

[John Bronco]

Magnus Santini, on 22 July 2022 - 09:41 AM, said:

I read through the patch notes twice and couldn't figure out the question: did the patch fix what the last one broke in the MM and resulting PSR? Forgive me for thinking that was the big issue of the moment and might have been addressed more specifically and visibly. Seems like changing the size of groups in quickplay does not address the problem at all.

Nothing has been broken with PSR, the problem was matches were allowing many groups per side instead of the 2+2, 4, or 2+3 that it was supposed to be. That has been fixed, per Cauldron discord conversations.

[Capt Deadpool]

Heavy Money, on 21 July 2022 - 03:49 PM, said:

snip

There exists myriad better, more nuanced solutions to the issue at hand than blanket buffs to chassis that perform the highest whether in the hands of cadets, veterans, or top players. Potentially, some of the more imaginative cauldron members I am aware of could have suggested such alternatives and were out-voted.

Knownswift, on 22 July 2022 - 12:18 AM, said:

The irony is anyone expecting this games' token infinite kiteyman to be less biased than the player who has played like 60 matches in light mechs over the last several years.

Y'all both ****** biased.

Everyone is biased. That is why the gods gave us statistics.

Dear Zeus, let the holy numerals clear away the fog for your wayward children, amen XD

Edited by Capt Deadpool, 28 July 2022 - 10:27 AM.

[Steel Shanks]

Hey uh PGI, Cauldron, as of RIGHT now 3:34 Eastern Time, 7-22-22, in the last 5 Games I've played, Assaults have had 42-52% for Quick Play... Medium and Light were both down to 3% on Quick Play... Yeah... It's NOT a FUN game when half of every fecking Team is an ASSAULT MECH!! In those 5 Games My team has lost ALL 5 TIMES... Now usually it's pretty even... Not now... If I lose the first 2 games, then I usually win the next three... Even when the match maker was screwed, I wasn't losing 5 Games in a Row guys...

I'm going to play ONE MORE game... If I lose, I'm going to play HBS Battletech... a FUN Battletech game... Maybe You guys should try making a fun Battletech game... That'd be nice...

[Steel Shanks]

WELL WHAT DO YOU KNOW... THE 1st MINUTE of the next match 3 Players Died!!! IN A MINUTE!! Losing 6 in a row isn't fun guys... I'm playing HBS Battletech now... Forget this game...

[Knownswift]

Capt Deadpool, on 22 July 2022 - 11:10 AM, said:

Everyone is biased. That is why the gods gave us statistics.

Dear Zeus, let the holy numerals clear away the fog for your wayward children, amen XD

Except we can't see individual chassis performance, outlier performance is a big deal in a game this diverse.

With that said, if the few light outliers are close to assault performance, how bad must the rest of the class be to drag down the global average?

What is the difference between the worst performing assault and the worst performing light mech?

As far as bias on the part of the cauldron goes, most of those accusations stem from one very loud person who happens to be the games biggest assault cryer and the actual patron saint of lasers from zimbabwe.

[Magnus Santini]

John Bronco, on 22 July 2022 - 10:05 AM, said:

Nothing has been broken with PSR, the problem was matches were allowing many groups per side instead of the 2+2, 4, or 2+3 that it was supposed to be. That has been fixed, per Cauldron discord conversations.

My understanding was that the free tier-mixing was an error created at the same time as that. The PSR was broken by that because the match score is "garbage in" if it is comparing T1s and T5s based on their results from any game. To make it clear, regardless of how well a T1 scores in a game like that, he does not demonstrate a reason to improve his PSR ranking among T1s, and the same for poor performance by a lower-tier player in that scenario. If this is not getting fixed, sorry to interrupt but I think PGI made promises they were working on it.

[Heavy Money]

Magnus Santini, on 22 July 2022 - 01:04 PM, said:

My understanding was that the free tier-mixing was an error created at the same time as that. The PSR was broken by that because the match score is "garbage in" if it is comparing T1s and T5s based on their results from any game. To make it clear, regardless of how well a T1 scores in a game like that, he does not demonstrate a reason to improve his PSR ranking among T1s, and the same for poor performance by a lower-tier player in that scenario. If this is not getting fixed, sorry to interrupt but I think PGI made promises they were working on it.

Free tier mixing was a previous issue brought by Event Queue problems that has already been fixed for weeks.

[Magnus Santini]

Heavy Money, on 22 July 2022 - 02:07 PM, said:

Free tier mixing was a previous issue brought by Event Queue problems that has already been fixed for weeks.

Cool, thanks, I didn't know.

[dario03]

Figaro Jibaro, on 21 July 2022 - 04:43 PM, said:

I am curious as to what data led the cauldron to believe assaults, out of all the other classes, needed an armor/structure buff. Out of the top 150 players, data suggests most play the below classes:

21 play Light (14%)
26 Medium (17%)
45 Heavy (30%)
58 Assault (39%)

So 39% of the top players play predominantly assaults.

Out of the combined 445,247 matches the above represents the match breakdown is:

10% lights
19% medium
29% heavy
42% assault

Assaults already tend to score the highest match scores, can hold the most weapons, have the potential for most firepower and DPS, already have the most armor/structure.

By many metrics the assaults are already the top performers.

Even survival rates look comparable across the board.

You are using some kind of actual data right? Not just pulling it out of nowhere?

It was more so done to make assaults more tanky focused as a class with higher performers potentially having some adjustments done later if they need to be brought back down. Personally I wanted to just buff the underperforming assaults with more focused quirk changes but others preferred this way.

[KDK Prime]

Teenage Mutant Ninja Urbie, on 21 July 2022 - 06:29 AM, said:

and with no reason at all you're posting on an alt..
just sayin'.

You're saying nothing.
Since, you know, people are not psychics and can't read intent over the internet.
If you want to make a derailing post, then at least make it count and post a strawman of your choice, to which a person may or may not reply.
Rather than fail at being cheeky.

Capt Deadpool, on 21 July 2022 - 11:00 AM, said:

...He concludes it is everyone else who is confused about the methodology they have used, which is a classic symptom of Group Think.

Why are you speaking for everyone? There's quite a few in this very thread being ok with the changes.
Imagine talking about group-think, while lumping people into one side against Navid and Cauldron. A bit too ironic and strawman-ey.

Quote

When I could explain how to utilize the most basic of statistics and game balance to a group of second graders, and help make them aware of their own pre-existing biases, and they could output a more professional result than a group of ostensibly mostly college educated adults, it's likely that Group Think holds sway.

Well, look at you being so smart. Surely that means you have proposed quite a few well thought out ideas over the past year to be qualified for Cauldron, right? Or maybe you're some known great (insert class or mech here) player? No? I mean, even DATA is there, surely you have the chance = )
Guess not, have fun with your imaginary group of "professional second graders".

Quote

However, the net end result is the same, with a blanket buff to assaults being relativistically equivalent to a blanket nerf to lighter classes...

You have some formal logic issues. Every buff to (insert anything) is a nerf to A-B-C-etc? And the reverse is true? Well, i guess the game's balance should have stayed in limbo for years then, can't have people be upset over buffs to things they don't play, right? =)

Quote

...which are the least-played...

True. Since it's the hardest class to play successfully, naturally. Hope your understanding of the game is as high as you seem to think of yourself and i don't need to explain why.

Quote

...least damage-dealing

Wrong. According to what?
If you mean average stats, probably. If we talk potential damage compared to other classes, it's basically the same. Refer to any higher level light players.
Oh, what's that i hear? Stats of filthy compies ™ are irrelevant for the purposes of balancing for the quickplay? Well, i guess lights should be buffed even more then, until the average player can do 1000-1500 damage matches at the same rate as assaults can. Who cares about the different gameplay dynamic between classes, right?
It's only natural for the most "zero-bomb" prone class to have lower averages.

Quote

...least impactful to W/L

Wrong. Same thing as damage. With the sprinkle of teammates RNG factor added in. This is a team game after all.

Quote

...despite the dumbed-down machine gun spam mechanic that has been granted to them which of course was never the answer and is like a band-aid for a bullet hole.

Wait a minute...You think that semi-recent MG spam bonanza was given to lights as the main (only?) way for them to become viable? Weird then how different people were playing just fine without MGs this whole time, huh. Guess you already forgot about laser-based lights and light peeps, huh. Poor YTers and their clickbait videos...

[Figaro Jibaro]

dario03, on 22 July 2022 - 10:11 PM, said:

It was more so done to make assaults more tanky focused as a class with higher performers potentially having some adjustments done later if they need to be brought back down. Personally I wanted to just buff the underperforming assaults with more focused quirk changes but others preferred this way.

I understand, but the question is why would the cauldron think assaults need a buff in the first place, when they (arguably) already outperform the other classes based on data? Making assaults tankier keeps them alive longer which then allows them to shoot more, which translates into even higher match scores, kills etc… exponentially skewing the data even more in favor of assaults, which again, they already dominate stats in. Preference should not be used to make these changes as it leaves way too much room for bias. Your method sounds a lot more logical. I’m really hoping the Cauldron is using more than preference to figure out what changes need to be made.

[Ravni]

>looks on through the smoking wasteland of the light/pseudo snubpocalypse

Am I so out of touch?
No, it's the King Crabs who are wrong.

[pattonesque]

Steel Shanks, on 22 July 2022 - 12:00 PM, said:

WELL WHAT DO YOU KNOW... THE 1st MINUTE of the next match 3 Players Died!!! IN A MINUTE!! Losing 6 in a row isn't fun guys... I'm playing HBS Battletech now... Forget this game...

this unfortunately is a player skill issue. Occasionally one Gets Got early due to bad luck but in large part you can *decide* when you die in this game. Mechdads do not know this and believe it to be out of their hands.

[CrimsonPhantom6sg062]

Figaro Jibaro, on 23 July 2022 - 08:12 AM, said:

I understand, but the question is why would the cauldron think assaults need a buff in the first place, when they (arguably) already outperform the other classes based on data? Making assaults tankier keeps them alive longer which then allows them to shoot more, which translates into even higher match scores, kills etc… exponentially skewing the data even more in favor of assaults, which again, they already dominate stats in. Preference should not be used to make these changes as it leaves way too much room for bias. Your method sounds a lot more logical. I’m really hoping the Cauldron is using more than preference to figure out what changes need to be made.

Because assault brawlers are clearly quite short-lived *1? You need to see more than cross-class results, consider intra-class results.

Remember that armor isn't as useful on snipers than on brawlers. I played with you and I know you like light/ medium brawlers. How important is armor and ecm (i.e. getting the first shot off) on these mechs? Now picture yourself running an ERPPC SHC or PPC Cicada. How important is armor now? I heard Navid, dario, and the Cauldron are considering nerfing the top performers, especially sniper-focused builds, and buffing mechs such as Commandos so I am not worried even as a (ER)PPC light mech fan (meanwhile Capt Deadpool hasn't even played lights since early 2021).

Also, its quite clear that the skill tree for assaults isn't working as intended at all (nor was hill climb, etc.). Any argument that tries to justify its reversal is simply bad and wrong.

Finally, match score isn't the ultimate gauge of performance. It is currently undergoing tuning that has already removed the effect of AMS, and will probably decrease effect of damage and increase effect of objective-focused tasks. The best gauge is match impact. I admit and have always drummed the idea that lights and mediums are overperforming and assaults are underperforming since MWO came out, but trying to exaggerate the light-assault gap isn't going to lead to better balance than we already are seeing now. We need to know where the the gap is coming from, before fixing it.

Is it really because lights are underperforming as a class *2? Or is it the fact most people play them as brawlers relative to assaults? Or is it because of the people playing them? I mean a lot of lights are used as pure support platforms or purely for scouting/ objectives by their pilots after all, and you probably won't even break 200 MS anytime soon doing that. Sometimes they also go on kamikaze missions or get sent to Zimbabwe; I see at least 1 fast mech do one of these things virtually every QP match I play.

Getting over 1000 damage in a light is MUCH harder than in an assault. But getting 5 kills and/ or clutching a difficult (conquest) game? You and I both know it isn't that difficult in lights, relative to assaults.


*1 Meanwhile, mid to long-range assault builds are farming damage fairly easily for long periods of time, so they are the ones that should be targeted. Its not like long-range builds are benefiting much from the change anyway.

*2 For the record I am in support of buffing lights and mediums. I genuinely believe there is a performance gap across most builds.


Edited: Added some emojis to liven up my post!

Edited by CrimsonPhantom6sg062, 23 July 2022 - 09:52 AM.

[Knownswift]

CrimsonPhantom6sg062, on 23 July 2022 - 09:48 AM, said:

Remember that armor isn't as useful on snipers than on brawlers.

Completely disagree. Uptime is critical in long range trade, and an extra 10-15 health if you're shooting from beyond your opponents' optimal range isn't huge, but it is not insignificant. Tree buffs aren't spelling an F for lights and mediums, but they are definitely helping trade mechs stay up.

I expect to see some prominent big blue laser mechs lose a little health in the coming months.

Quote

How important is armor and ecm (i.e. getting the first shot off) on these mechs? Now picture yourself running an ERPPC SHC or PPC Cicada. (meanwhile Capt Deadpool hasn't even played lights since early 2021).

Deadpool literally does not play non-ecm mechs that often, or at all. He mained a stealth FLE to play infinite run-away while grouped. Switched to nothing but Lanner. Maybe they do sad viper stuff now, I haven't seen them since I no longer play those hours.

Their entire playstyle revolves around ECM buttplay. The tree buffs don't even really impact their gameplay.

Quote

It is currently undergoing tuning that has already removed the effect of AMS, and will probably decrease effect of damage and increase effect of objective-focused tasks.

plz no. Can't play anything with remote quickness and not get yelled at for not wanting to square-stand. The objectives in this game aren't even designed around that kind of play, and I don't see that changing.

Altho, I do think lights and mediums should get slightly bigger kickers for protect heavy/protect assault and maybe a few other things.

[Gas Guzzler]

Patch looks great and all but what happened to the next mech pack?

[CrimsonPhantom6sg062]

Knownswift, on 23 July 2022 - 12:47 PM, said:

Completely disagree. Uptime is critical in long range trade, and an extra 10-15 health if you're shooting from beyond your opponents' optimal range isn't huge, but it is not insignificant. Tree buffs aren't spelling an F for lights and mediums, but they are definitely helping trade mechs stay up.

I expect to see some prominent big blue laser mechs lose a little health in the coming months.

Agreed that some assaults are going to lose health to compensate.

Also agree that uptime is important for long-range builds - though you can argue that it is critical for any build.


As for your supposition that snipers benefit more from armor than brawlers:

Prove it



Upload some statistics or consecutive matches of snipers and brawlers where snipers used more of their armor than brawlers on average.

I speak from observation, not opinion. I don’t mind if someone criticizes my ideas, but please back it up with a logical explanation and/ or evidence.


I tried to upload my images of previous battles from my recordings (I don't like people looking at my recordings), but my image uploader has stopped working for good.

I am just going to post my survivability stats for ALL the previous QP battles that I recorded for snipers or brawlers. If anyone can provide proof for my battles, then feel free to upload the proof on your side.

Removed battles where I did less than 200 MS, because it could be biased by me being an idiot and playing too passively or too aggressively. Also, there are no results for assaults because I didn’t play many assaults at all during this time period, there are only results for fast and med-speed mechs:



26/12/2022 to 25/03/2022:

Results for snipers: Total: 23.

Good condition: 13 (56.5%). Poor condition: 2 (8.7%). Dead (win): 2 (8.7%).
Survived (loss): 0 (0%). Dead (late): 5 (21.7%). Dead (early): 1 (4.3%).

Results for brawlers: Total: 24.

Good condition: 6 (25.0%). Poor condition: 7 (29.2%). Dead (win): 6 (25%).
Survived (Loss): 1 (4.2%). Dead (late): 3 (12.5%). Dead (early): 1 (4.2%).


Summary: Looks like brawlers actually need the extra armor compared to snipers. There is no difference in terms of defeat statistics. Brawlers are more likely to end winning games with very little armor left provided the brawler/ sniper actually had impact in the match.



Actually, just realized that my detailed statistics will generate a MASSIVE wall of text. I can upload them in the forums if requested.

You can nitpick at my observations, but the fact is I personally observe and believe that brawlers rely more on armor.

Also, if you doubt the authenticity of my results, just ask me and I can give you even more details on a particular match. E.g. Why I specified a mech had good/ poor condition, which mechs died in the match and their death timestamp, etc. (I can just look back at my recordings)

[Knownswift]

CrimsonPhantom6sg062, on 24 July 2022 - 06:39 AM, said:

As for your supposition that snipers benefit more from armor than brawlers:

Prove it

I never said that long range trade mechs benefit more than brawlers. I countered your statement that it isn't as useful. I completely disagree.

The ultimate test though, would be to only record your results from boreal

Since we're here tho.. The king crab still massively sucks. Can we make like half of the top of it act like a missile door for damage mitigation?

Edited by Knownswift, 24 July 2022 - 08:00 AM.