36 replies to this topic

[LordNothing]

i think a more hard stats approach is probably better, im just biased in a lot of areas. like i gave the ice ferret a perfect 10, because like 3 of the mechs in my top 20 are ice ferrets, no other chassis has that much representation at the top of my stats. lots of other players say they are terrible, useless. but my stats, in accordance with my preferred play styles, says those cold beer fridge on legs are stronk.

i also put my dires and warhawks higher up because i always preferred those over more recent equivalents. also stronk variants of both are really tilting the stats towards those. i gave the supernova extra because it gave me my ace.

also the per chassis rating ignores the fact that for any given chassis in the game there are at most 2 variants that are good, sometimes only one and rarely 3 or more, and sometimes none at all. thats been my exeprience with mech packs, there is that one variant you bought the pack for, and then there is all this dead weight ballast surrounding it that could have potential to reinvigorate the game if they got quirked or buffed or anything other than fill up space in my hanger. most of the bad mechs sit in the shadows of the preferred variant.

back to the ice ferret thing, what if a mech is stronk, but everyone thinks its weak, and then it receives buffs? im running the b variants of everything, anything i gave up on out right, and then put 50 games on it such that they get statistical viability for accurate comparison. of course with 300+ mechs this takes time. im still convinced there are good variants hiding among the detritus. but a general querkening is whats required.

[My Lord and Saviour Jesus Christ]

LordNothing, on 15 March 2021 - 05:19 AM, said:

i think a more hard stats approach is probably better, im just biased in a lot of areas. like i gave the ice ferret a perfect 10, because like 3 of the mechs in my top 20 are ice ferrets, no other chassis has that much representation at the top of my stats. lots of other players say they are terrible, useless. but my stats, in accordance with my preferred play styles, says those cold beer fridge on legs are stronk.

You can play something, do well in it and still be able to look objectively and go "yeah this isn't great".

[The pessimistic optimist]

Monke-, on 15 March 2021 - 05:58 AM, said:

You can play something, do well in it and still be able to look objectively and go "yeah this isn't great".

Thing with the fridge is there are a few strong builds for them but that the thing a few builds. Its engine is too big and limited its load outs. But the ones that work work VERY well. It plays like over sized light even has profile and speed of a light but armor of a 45 ton mech. It can be played like a tanky light mech and in that it has value. This would be a good fridge build https://mwo.smurfy-n...7e86134b83dc4ff or this skip the support roll https://mwo.smurfy-n...7fd02d2ce7a51f0 note heavy meds are it's friend

Edited by SirSmokes, 15 March 2021 - 06:54 AM.

[Adette]

SirSmokes, on 15 March 2021 - 06:42 AM, said:

Thing with the fridge is there are a few strong builds for them but that the thing a few builds. Its engine is too big and limited its load outs. But the ones that work work VERY well. It plays like over sized light even has profile and speed of a light but armor of a 45 ton mech. It can be played like a tanky light mech and in that it has value. This would be a good fridge build https://mwo.smurfy-n...7e86134b83dc4ff or this skip the support roll https://mwo.smurfy-n...7fd02d2ce7a51f0 note heavy meds are it's friend

Comp players only ranked the fridge higher than casuals simply because of Bows3r going 2 ERPPC on a fridge and destroying people with it in comp play. Outside of that, it is a terrible mech.

[Heavy Money]

LordNothing, on 15 March 2021 - 05:19 AM, said:

i think a more hard stats approach is probably better, im just biased in a lot of areas. like i gave the ice ferret a perfect 10, because like 3 of the mechs in my top 20 are ice ferrets, no other chassis has that much representation at the top of my stats. lots of other players say they are terrible, useless. but my stats, in accordance with my preferred play styles, says those cold beer fridge on legs are stronk.

i also put my dires and warhawks higher up because i always preferred those over more recent equivalents. also stronk variants of both are really tilting the stats towards those. i gave the supernova extra because it gave me my ace.

Uh, are you really complaining that the survey isn't helpful because you filled it out based on your individual biases? Its not a survey of what's in your top 20.

If the Ice Ferret is a 10, what did you rate the Vapor Eagle?

[Brauer]

Heavy Money, on 15 March 2021 - 09:39 AM, said:

Uh, are you really complaining that the survey isn't helpful because you filled it out based on your individual biases? Its not a survey of what's in your top 20.

If the Ice Ferret is a 10, what did you rate the Vapor Eagle?

Well there's always some noise in survey data.

[The pessimistic optimist]

Adette, on 15 March 2021 - 07:49 AM, said:

Comp players only ranked the fridge higher than casuals simply because of Bows3r going 2 ERPPC on a fridge and destroying people with it in comp play. Outside of that, it is a terrible mech.

Like I said comp guys always lack imagination. Being mechanically gifted doesn't mean you have creative idea's or strategy and there a lot things they just never think off seriously no offense guys

[Adette]

SirSmokes, on 15 March 2021 - 11:12 AM, said:

Like I said comp guys always lack imagination. Being mechanically gifted doesn't mean you have creative idea's or strategy and there a lot things they just never think off seriously no offense guys

LOL, lack imagination, you have no idea how much hundreds to thousands of hours individual comp players have been spent tinkering in the mech lab and scrimming private matches to perfect mech builds and strategies. I can pretty much guarantee top end comp teams have tested just about everything (even random bracket builds like 2xLBX 20 + LRM 60s etc), far more than you have ever done.

Edited by Adette, 15 March 2021 - 06:58 PM.

[Tarogato]

LordNothing, on 15 March 2021 - 05:19 AM, said:

i think a more hard stats approach is probably better, im just biased in a lot of areas. like i gave the ice ferret a perfect 10, because like 3 of the mechs in my top 20 are ice ferrets, ...

When people make posts, they are complaining that Mech A is bad and needs a buff or Mech B is too OP and needs a nerf ... all based on their very subjective perceptions. This survey is about aggregating those perceptions and looking for trends among them.

I would love to look at hard mech stats, but PGI doesn't release mech stats publicly. Individual players' own mech stats, even hundreds of matches worth, are not representative of the true performance of a mech. You need a very wide range of players who have many games played in a mech, to begin to get an idea of what its baseline performance is. We as players aren't provided the tools to collect such data, and PGI as a developer has the data but not the proper knowledge to interpret it. So the best we can do is collectively shout the same thing and hope the data supports it and PGI reacts appropriately. Based on this months' upcoming patch notes ... that isn't possible.

[PocketYoda]

Fridge runs great with 4 heavy mediums.

[LordNothing]

Tarogato, on 15 March 2021 - 06:55 PM, said:

When people make posts, they are complaining that Mech A is bad and needs a buff or Mech B is too OP and needs a nerf ... all based on their very subjective perceptions. This survey is about aggregating those perceptions and looking for trends among them.

I would love to look at hard mech stats, but PGI doesn't release mech stats publicly. Individual players' own mech stats, even hundreds of matches worth, are not representative of the true performance of a mech. You need a very wide range of players who have many games played in a mech, to begin to get an idea of what its baseline performance is. We as players aren't provided the tools to collect such data, and PGI as a developer has the data but not the proper knowledge to interpret it. So the best we can do is collectively shout the same thing and hope the data supports it and PGI reacts appropriately. Based on this months' upcoming patch notes ... that isn't possible.

its like the opposite problem i have with the goulag. while thats theory seeking testing, this is anecdotes seeking theory.

ive had that problem myself keeping track of my own stats. id love to filter out what a run when and cross reference that with jarls to see how i was playing when those stats are recorded, like was i doing an event or was i playing for real. even just having post mastery stats separated from leveling stats. stuff i insti-mastered is going to have higher ratings that stuff i ground, just by virtue of not having to run them without nodes. its very hard to get good usable stats. to have stats based on the build+node hash would be exceptionally useful for finding good variants lost among the detritus of bad mechs. the only thing my spreadsheets tell me is how well i do in each variant, and that varies wildly with build choice.

a lot of the bad mechs are bad because they are unable to live up to the best variant in the pack. op mechs always make more waves than bad mechs. the most meta variant is the one pgi seems to use for their balancing yardstick. many mechs have variants that are spread out a lot in my personal stats (i got into that in more detail in the worst mech thread) and im sure other people and their stats have similar trends. blanket buffs/nerfs tend to miss that detail. its no doubt useful to identify which mechs have the widest spread among their variants so that they can get more scrutiny from the balance overlords.

[LordNothing]

Heavy Money, on 15 March 2021 - 09:39 AM, said:

Uh, are you really complaining that the survey isn't helpful because you filled it out based on your individual biases? Its not a survey of what's in your top 20.

If the Ice Ferret is a 10, what did you rate the Vapor Eagle?

there are mechs im fond of that can tilt the way i perceive them. thats the problem with personal stats, they are personal. i dont trust my gut in these matters, though a lot of the more meh mechs didnt get as much scrutiny. there are certain fun mechs that i really like to run even though their stats say i shouldnt.

veagles are good but they arent as good as the hype seems to indicate. maddog h still wins the atm game (and it was the top mech until the corsair 7a knocked it down by one due to insane ams capabilities). ifr-d comes in 3rd. the vgl3 comes in 7th (right under the fireball which everyone said was weak), the only other one i own is the ballistic variant. i think i gave it a 10 in the survey. its still top 10 material (among the 450+ variants i have stats for).

then we hit the problem of balancing variants based on the most meta variant, which is why we have so many bad variants.

Edited by LordNothing, 16 March 2021 - 12:26 AM.

[LordNothing]

Monke-, on 15 March 2021 - 05:58 AM, said:

You can play something, do well in it and still be able to look objectively and go "yeah this isn't great".

thats the bane of the fun mech. you usually have such a good time with them that you forget that you are actively gimping your team.

[Alexander of Macedon]

I leave for two years, come back, and see people are still underrating the Osiris.

[Brauer]

LordNothing, on 16 March 2021 - 12:20 AM, said:

there are mechs im fond of that can tilt the way i perceive them. thats the problem with personal stats, they are personal. i dont trust my gut in these matters, though a lot of the more meh mechs didnt get as much scrutiny. there are certain fun mechs that i really like to run even though their stats say i shouldnt.

veagles are good but they arent as good as the hype seems to indicate. maddog h still wins the atm game (and it was the top mech until the corsair 7a knocked it down by one due to insane ams capabilities). ifr-d comes in 3rd. the vgl3 comes in 7th (right under the fireball which everyone said was weak), the only other one i own is the ballistic variant. i think i gave it a 10 in the survey. its still top 10 material (among the 450+ variants i have stats for).

then we hit the problem of balancing variants based on the most meta variant, which is why we have so many bad variants.

The reason surveys seek to collect many responses is because individual responses may be biased by some personal opinions or lack of information, but once you start to see trends you can often filter through this noise and see a consensus. One individual response does not have to be perfect for a survey to overall yield helpful information.

Veagles are just as good as the hype indicates. They are far better at atm boating than the Mad Dog H. ATM mechs greatly benefit from having JJs to poptart with, and the Veagle can also carry more tubes at similar speeds to the MDD. It's not much of a contest between the two tbh.

[The pessimistic optimist]

Brauer, on 16 March 2021 - 04:57 AM, said:

The reason surveys seek to collect many responses is because individual responses may be biased by some personal opinions or lack of information, but once you start to see trends you can often filter through this noise and see a consensus. One individual response does not have to be perfect for a survey to overall yield helpful information.

Veagles are just as good as the hype indicates. They are far better at atm boating than the Mad Dog H. ATM mechs greatly benefit from having JJs to poptart with, and the Veagle can also carry more tubes at similar speeds to the MDD. It's not much of a contest between the two tbh.

MDD is a better LRM boat but VE leave it eating dust on ATMs. ATM+JJ=very good

Edited by SirSmokes, 16 March 2021 - 05:04 AM.

[LordNothing]

its not reflected in my stats but ok. right now both have been dethroned by an ams boat.

Brauer, on 16 March 2021 - 04:57 AM, said:

The reason surveys seek to collect many responses is because individual responses may be biased by some personal opinions or lack of information, but once you start to see trends you can often filter through this noise and see a consensus. One individual response does not have to be perfect for a survey to overall yield helpful information.

its a good place to start i suppose. but the variant spread needs to be delt with eventually.

Edited by LordNothing, 16 March 2021 - 05:15 AM.