41 replies to this topic

[STEF_]

Stating many time, they are using stats and numbers to ......(tadaaaa)....BALANCE.

OK.

Now tell me how they can do this, if stats and numbers are fails.

This is my C-AC5 stat:


I think I've been very good to hit 77.59% and doing 164 dmg in zero matches!



It's not the first time I report this: server stats are useless because they report random numbers.
Here from a fresh bought marauder stats during a challenge:

marauder stats:


challenge stats:


(yep: they are referring to the same mech: I had 2 ghost defeats in the challenge)


Server "register" or not register at all.

[Y E O N N E]

Server stats != user front-end presentation of said stats.

[Medicine Man]

Yeonne Greene, on 25 June 2017 - 07:17 PM, said:

Server stats != user front-end presentation of said stats.

Well that's kind of ridiculous and lackluster. Which just about describes PGI perfectly so it's plausible.

[Deathlike]

Yeonne Greene, on 25 June 2017 - 07:17 PM, said:

Server stats != user front-end presentation of said stats.

While true, you have to remember that the bulk of players are not really "Tier 1" players (or true Tier 1 quality), so we're being dragged down to the mean (however low that may be) or have a crappy median to determine stats with.

Essentially, since the bar is naturally low, data is determined by below mediocrity.

[GenghisJr]

Stefka Kerensky, on 25 June 2017 - 07:13 PM, said:

Stating many time, they are using stats and numbers to ......(tadaaaa)....BALANCE.

OK.

Now tell me how they can do this, if stats and numbers are fails.

This is my C-AC5 stat:


I think I've been very good to hit 77.59% and doing 164 dmg in zero matches!



It's not the first time I report this: server stats are useless because they report random numbers.
Here from a fresh bought marauder stats during a challenge:

marauder stats:


challenge stats:


(yep: they are referring to the same mech: I had 2 ghost defeats in the challenge)


Server "register" or not register at all.

I'll see your C-UAC5 and raise you one - Ive only played a few games of FP but somehow have spent this much time on Emerald Taiga 277 137 136 1.01 1 day 04:27:31
...so Im guessing someone at PGI uses my account for FP?

[Karl Streiger]

No not this again.
Open you eyes. Open you mind.
You infidels are fools that fear what they can't unterstand.
See shots?
See hits?
See damage?

Consider that PGI can compare stats from patch to patch - this is everything you need.
Same for Mechs.
Absolute correct to use data instead of player XP.
Because otherwise you start to balance skill that would be stupid

[Kiiyor]

The stats and API's (or lack thereof) leave a lot to be desired. I'd love to get my hands on some backend stuff so I could SCIENCE without using web scraping tools.

Having said that, I have in the past married up data from my own stats page (kills, damage, match score etc) against what I recorded in EOM screens, and that part seems accurate.

The weapons themselves though, not so much.

[Tier5 Kerensky]

So Kodiak 3 was not too daam powerful when it got nerfed. And neither are Marauder IIC or Night Gyr pretty good mecs to receive any nerfs.

[Elizander]

Deathlike, on 25 June 2017 - 09:02 PM, said:

While true, you have to remember that the bulk of players are not really "Tier 1" players (or true Tier 1 quality), so we're being dragged down to the mean (however low that may be) or have a crappy median to determine stats with.

Essentially, since the bar is naturally low, data is determined by below mediocrity.

We don't even know if they are parsing it to compare stats of what tier players T1 players kill or if the stats are segregated by Tiers.

Tier 1 player stats bloated from killing Tier 5 potatoes? Tier 1 stats vs Tier 1 players? Group queue versus Solo queue? Stats of players in top units versus the entire pool?

Also what determines proper amount of usage for a weapon? Either way, I think the C.MPL and IS.ML changes are alright. The LPL changes don't really affect gameplay and I think some people are happy with the ER LL duration changes. Outside of the C.SPL nerf and the insignificant IS.SL changes, I can't say that it was a bad balance pass.

We also all know your C.SPL is coming back as the Heavy Smol Lazor.

[PhoenixFire55]

Yeonne Greene, on 25 June 2017 - 07:17 PM, said:

Server stats != user front-end presentation of said stats.

Yeah, server stats are probably fkd up even worse ...

[Luminis]

I'm not against balancing by stats in general, but I'm a little concerned about ease of use skewing the stats.

Even a mouthbreathing potatoe can hit a target with lock-on weapons. Or spray a hitscan weapon all over the enemy. But hitting a moving target with a projectile? Seriously, this might make ACs, PPCs and Gausses seem much weaker than they actually are when used well.

I do hope that PGI is gonna account for that and isn't just going by averages. Then again, going by averages might explain some of the balancing decisons we've seen...

[Aim64C]

Frankly, if my experience is anything... player tier is more a product of persistence than an actual ranking system. After almost 2 years of not playing... I was in Tier 3. Recently broke into Tier 2, where I do notice some difference in how people play, but have been rather consistently climbing toward Tier 1.

Whatever that means. I mean... I'm a dragon, so I like to think I'm a god among mortals... but because I'm a god among mortals, I know that my mortal form is expected to have limits and fall within statistical norms... so it would actually surprise me to find that I'm actually some kind of 1337 player just because I know how to point and shoot.

There again... I did survive against a firebrand in my damaged supernova that I had no business living against... but I've had plenty of matches where I get rolled before barely breaking 100 damage, too... so....

I take issue with the way they balance weapons with this game... always have...

Personally, I see the balance issue as a compound one. First, and foremost, is the reality that a fair number of the lighter 'mechs in the game were not designed to be used as anti-mech solaris champions. They were designed for infantry/armor support/interdiction and their weapon set is essentially being improvised when it is deployed against other battlemechs. You're shooting another 'mech with a locust not because it's a particularly good idea, but because other options are not presently on the table.

Something like the locust, however, was invaluable in dealing with infantry... which could put the hurt on 'mechs that lacked the proper weapons for fighting infantry. Since we don't have any of these other roles, we end up with this weird issue of trying to make the locust a valid Solaris contestant in the same arena with an Atlas.

The second problem is the actual mechanics of firing a weapon. I've gone over this a hundred times in the past. Too much emphasis on pulling the trigger and not enough emphasis on how each weapon can vary in its operation. This makes the game extremely twitchy for every weapon type. It's all about the single button press.

But... I guess we're all supposed to be Korean and doing the e-sports thing? ... I dunno... I just shoot things while saving up to try this thing out:

http://mwo.smurfy-ne...f32e6764c4a62f6

I'm thinking it would be amusing after going full ham on internal structure. In terms of heat dissipation, it actually out-performs doubles if smurfy's numbers are accurate.

[Pahrias]

ATLAS AS7-D 1,619 834 725 1.15 1,550 1,258 1.23 552,637 1,746,099 24855 days 03:14:07
notice anything odd here?

some stats are seemingly bugged in some way.

[kapusta11]

Even if those stats were correct, how do you interpret them? If you see cSPL outperforming other small lasers and even SRMs, what makes you think it's cSPL that needs to be nerfed and not everything else to be buffed? Do cSPLs win you as many games in team vs team environment (not solo puging) as Gauss/PPC/UAC/LPL/ERML? If not then you're just nerfing brawling and buffing long range loadouts.

Edited by kapusta11, 26 June 2017 - 07:35 AM.

[STEF_]

Karl Streiger, on 25 June 2017 - 10:06 PM, said:

Consider that PGI can compare stats from patch to patch - this is everything you need.

using wrong numbers?

Because if they nerf/buff from patch to patch with fake numbers, the result is the darthboard quiks, like we've seen

Kiiyor, on 25 June 2017 - 10:15 PM, said:

The weapons themselves though, not so much.

and sadly weapon numbers MUST be accurate, if they want to do a serious balance.
Otherwise, we have the darthboard.

[Y E O N N E]

Deathlike, on 25 June 2017 - 09:02 PM, said:

While true, you have to remember that the bulk of players are not really "Tier 1" players (or true Tier 1 quality), so we're being dragged down to the mean (however low that may be) or have a crappy median to determine stats with.

Essentially, since the bar is naturally low, data is determined by below mediocrity.

I know that.

For the record, if it's a mechanical/technical aspect of something, I am more than likely already aware of it (not directed at you so much as the forums in general) and I grow weary of people parroting the same **** at me ad nauseam.

All I was trying to say, though, is that what you see on your stats page has to go through some system to get there and that system is known to mess up the numbers. And I have no idea how perfect Mad Stats is, either. So the premise of the OP is pretty weak given that he has no idea what, exactly, PGI is seeing.

Edited by Yeonne Greene, 26 June 2017 - 05:27 AM.

[cazidin]

Ah, the Word of Cazidin's propaganda is great. The Balance Master title has stuck!

[Paigan]

Yeonne Greene, on 25 June 2017 - 07:17 PM, said:

Server stats != user front-end presentation of said stats.

That might be true, but you are making a similar assumption you insinuate the OP does.

The quering logic used to populate the UI MIGHT be buggy and thus falsify correct data.
It also MIGHT be correct and the error might reside in the collected/persisted data itself.

Either way, the OP still has a point in that the PGI stats business ist least somewhat fishy.
Their data structure might be overly complicated and prone to making erroneous queries.
The balancing decisions might be based on buggy queries (maybe even the same that is used for the UI)
etc.

Edited by Paigan, 26 June 2017 - 05:35 AM.

[Karl Streiger]

Paigan, on 26 June 2017 - 05:32 AM, said:

That might be true, but you are making a similar assumption you insinuate the OP does.

The quering logic used to populate the UI MIGHT be buggy and thus falsify correct data.
It also MIGHT be correct and the error might reside in the collected/persisted data itself.

Either way, the OP still has a point in that the PGI stats business ist least somewhat fishy.
Their data structure might be overly complicated and prone to making erroneous queries.
The balancing decisions might be based on buggy queries (maybe even the same that is used for the UI)
etc.

Shots / hits / damage are valid (damage is obstructed by critical hits and shots beyond long range)

[Skanderborg]

Deathlike, on 25 June 2017 - 09:02 PM, said:

While true, you have to remember that the bulk of players are not really "Tier 1" players (or true Tier 1 quality), so we're being dragged down to the mean (however low that may be) or have a crappy median to determine stats with.

Essentially, since the bar is naturally low, data is determined by below mediocrity.

I like to elevate my self importance by being good at video games. Oh wait , no one cares.