338 replies to this topic

[Deathlike]

UnofficialOperator, on 27 October 2016 - 08:49 PM, said:

You like many others are missing the point that we are trying to make. We are showing that it is exactly illogical/irrational to base balancing on the stats of a few. You want to balance, then base it on all the data of all players in all tiers.

There are even more obvious reason why these outliers only prove how useless the stats you are trying to pick are.

First off, it doesn't even say what mech they are using primarily, but moreso you can have radical levels of K-D ratios where a player doesn't die very often.

When you're picking numbers, you need proper context... and you have none of that.

[Zergling]

Oh, because UnofficialOperator and others seem completely ignorant of statistical methodology, here's a definition for them:

Quote

A representative sample is a small quantity of something that accurately reflects the larger entity. An example is when a small number of people accurately reflect the members of an entire population.

See, a census of all players isn't required to determine if there is a statistical anomaly; a representative sample of the population is all that is needed.

El Bandito's + Mcgral18's plus the event leaderboards provides a representative sample, with further sample data coming from monthly leaderboards data of known KDK-3 players.

Go ahead and argue that those samples aren't representative if you want, but the burden of proof is on anyone making that claim.
The burden of proof means equal or greater samples are needed for refutation.

And no, you can't simply say 'but I disagree because raisans!'. That isn't how arguments work in the adult world.

[Dee Eight]

Zergling, on 27 October 2016 - 08:55 PM, said:

Pretty obvious you cherry-picked the mechs with good winrates, as apart from the Warhawk, none of those have sufficient damage/battle to produce much better W/L than 1.0 level.
Your leaderboard stats shows that these are the extreme outliers that only won more battles than they lost due to luck with random teams.

Ah I see...when I get good results its merely luck with random teams...when one of you tiertards does it, its got nothing at all to do with the luck of random teams. Ya know perhaps if you had better results one might listen to your fables of how things are, but...

sorted by W/L ratio...season 3 and 4 (you have no 1 or 2 games)

11036 Zergling 133 104 1.28 197 159 1.24 238 233

10107 Zergling 195 153 1.27 340 193 1.76 348 260

That's not particularly amazing results in my opinion. Maybe it impresses others here who wanna like your posts but it doesn't impress me.

Quote

If your leaderboard results are consistently bad month after month, then you are simply a bad player.

By total wins I finished season 1 in 98th, season 2 in 175th, season 3 in 160th and currently season 4 I'm 148th.

Quote

If you think a 15 battle sample in a single variant is proof of anything, then you are beyond reasoning with, because intelligence discussion is simply beyond you.

And incase you aren't intelligent enough to do a simple division calculation, that is only 312 damage/battle.

To sustain that winrate, you'd need to be doing at least 500 damage/battle.

Ah yes...insult me when the argument turns against you. Perhaps you should re-read the CoC before bringing intelligence into this discussion.

[Mystere]

Zergling, on 27 October 2016 - 09:05 PM, said:

And no, you can't simply say 'but I disagree because raisans!'. That isn't how arguments work in the adult world.

Well, not according to this guy:

[Mcgral18]

Dee Eight, on 27 October 2016 - 09:20 PM, said:

By total wins I finished season 1 in 98th, season 2 in 175th, season 3 in 160th and currently season 4 I'm 148th.

You do realise that one is simply playing lots of matches, right?
That isn't related to performance whatsoever.


Globally below 200 average match score (closer to 170)
Comparatively, aside from the current season (Levelling robots, low amounts of play), I am above 400


You have played a lot more than I, but your stats are not exactly stellar

[UnofficialOperator]

El Bandito, on 27 October 2016 - 08:42 PM, said:

Show me your Adder stats, I highly doubt your Adder average damage per match is anywhere near 800.

How are we to take anything you say seriously when you are comparing a light mech vs an assault mech for damage numbers?

[El Bandito]

UnofficialOperator, on 27 October 2016 - 09:28 PM, said:

How are we to take anything you say seriously when you are comparing a light mech vs an assault mech for damage numbers?

Cause that guy boasted that he gets 800 damage almost every game in an Adder? I hate liars.

Dee Eight, on 27 October 2016 - 09:20 PM, said:

By total wins I finished season 1 in 98th, season 2 in 175th, season 3 in 160th and currently season 4 I'm 148th.

Total wins alone doesn't mean much if your stats is average at best.

Edited by El Bandito, 27 October 2016 - 09:32 PM.

[Mcgral18]

UnofficialOperator, on 27 October 2016 - 09:28 PM, said:

How are we to take anything you say seriously when you are comparing a light mech vs an assault mech for damage numbers?

500 would be a fair score, in that case
An extra ordinarily powerful Light robot can sit there (see GodQuirk cheetah)

Let's see what he actually has.

[Dee Eight]

I never claimed they were stellar but I am satisfied with them. Again though I'm not a stat chasing tiertard which apparently you have to be to have any authority in a discussion. If I posted selected mech results...I'm cherry picking. When one of you others do it... you're not. Do you not grasp the hypocrisy on your own ?

[El Bandito]

Dee Eight, on 27 October 2016 - 09:33 PM, said:

I never claimed they were stellar but I am satisfied with them. Again though I'm not a stat chasing tiertard which apparently you have to be to have any authority in a discussion. If I posted selected mech results...I'm cherry picking. When one of you others do it... you're not. Do you not grasp the hypocrisy on your own ?

Just curious, which select mech stats have you posted about yourself in this thread? Cause your ARC-5W stats was nothing special.

Edited by El Bandito, 28 October 2016 - 02:37 AM.

[Mcgral18]

Dee Eight, on 27 October 2016 - 09:33 PM, said:

I never claimed they were stellar but I am satisfied with them. Again though I'm not a stat chasing tiertard which apparently you have to be to have any authority in a discussion. If I posted selected mech results...I'm cherry picking. When one of you others do it... you're not. Do you not grasp the hypocrisy on your own ?

No, I do not
Cherry picking a unicorn among heaps of trash (see 50 W/L in Lights) is frowned upon, for good reason.

Do you have a decent sample size and care to post the stats?
MadStats preferably, in image form (saves us from having to Math anything out)
Leaderboards are a lump amount, so you can give us a good robot, if you want.



The point of this thread is to point out what a farce the KDK-3 is, in that it's been the best robot in the game for nearly half a year.
I find that a touch ridiculous, as it's only had one pass.
That may change in November, but it should have been done many patches ago.

[Deathlike]

Dee Eight, on 27 October 2016 - 09:33 PM, said:

I never claimed they were stellar but I am satisfied with them. Again though I'm not a stat chasing tiertard which apparently you have to be to have any authority in a discussion. If I posted selected mech results...I'm cherry picking. When one of you others do it... you're not. Do you not grasp the hypocrisy on your own ?

If you're trying to say "Kodiak-3" is not out of whack with the competition, then you need to dig deeper for those stats.

The problem is that it can already be objectively found is you put your money where your mouth is and try your wonderful builds against the top tier builds. That way, you can find out how much you don't know about the game.

[SpectreHD]

Nerf its pitch and/or yaw arcs or remove one ballistic slot from each torso. Done and done.

[BCAW]

Mcgral18, on 27 October 2016 - 09:38 PM, said:

No, I do not
Cherry picking a unicorn among heaps of trash (see 50 W/L in Lights) is frowned upon, for good reason.

Do you have a decent sample size and care to post the stats?
MadStats preferably, in image form (saves us from having to Math anything out)
Leaderboards are a lump amount, so you can give us a good robot, if you want.



The point of this thread is to point out what a farce the KDK-3 is, in that it's been the best robot in the game for nearly half a year.
I find that a touch ridiculous, as it's only had one pass.
That may change in November, but it should have been done many patches ago.

Well, I'm relatively new here, but from what I've heard, it took over a year for the Wave 1 Clams (especially the TBRs and the SCRs) to be balanced. Seems like that is going to be the case again.

Powercreep is real. I've heard many horror stories of how ridiculous the Wave 1 Clams were, and how the TBR and SCR completely dominated their respective weight classes. Nowdays, in my personal opinion, the TBR almost feels like it is to the Clans what the Cataphract is to the IS. Out-dated, and gradually replaced in the "meta tier list". With dakka and ppfld becoming the meta, the TBR has been power-creeped by the newly arrived Night Gyr, a mech that does dakka and ppfld better. Not saying the TBR is a bad mech; heck, I don't regret buying them, I love those things. It is a flexible chassis with many builds, perfect for a newcomer like me who like to switch things up once in a while without having to buy entire new chassis, and its a good energy boat and brawler. Just that those two roles are not the meta at the moment, and the its lack of high-mounted direct fire hardpoints (excluding the A left torso) hurts its "meta-compatibility" in similar ways as the Cataphract's hardpoint configurations.

Edited by BCAW, 27 October 2016 - 10:01 PM.

[Mcgral18]

SpectreHD, on 27 October 2016 - 09:58 PM, said:

Nerf its pitch and/or yaw arcs or remove one ballistic slot from each torso. Done and done.

Remove all quirks, to start off

Then torso yaw (not pitch)
Affects firing arc and survivability, but I'd rather not touch pitch (yet)

[Dee Eight]

Deathlike, on 27 October 2016 - 09:41 PM, said:

If you're trying to say "Kodiak-3" is not out of whack with the competition, then you need to dig deeper for those stats.

The problem is that it can already be objectively found is you put your money where your mouth is and try your wonderful builds against the top tier builds. That way, you can find out how much you don't know about the game.

As has been pointed out...I play a LOT of games each month, A LOT more than any of you...thus I see a fair number of KDK-3s.. and the OP nature a lot of you are demanding must be real based on the performance of a handful of players and a small sample size of games...isn't actually how the mech performs in the really real world amongst the majority of players using them. Furthermore, and I and others have said this repeatedly... PGI also has access to all the real numbers for the mech, and they clearly do not think its OP as they haven't done anything to it since they removed its engine quirks and most of its structure ones three months ago. And typically the only times they even turn in these stellar results people seem to think is automatic, is when they happened to be on the winning team.

Edited by Dee Eight, 27 October 2016 - 10:10 PM.

[Mcgral18]

BCAW, on 27 October 2016 - 09:59 PM, said:

Well, I'm relatively new here, but from what I've heard, it took over a year for the Wave 1 Clams (especially the TBRs and the SCRs) to be balanced. Seems like that is going to be the case again.

Powercreep is real. I've heard many horror stories of how ridiculous the Wave 1 Clams were, and how the TBR and SCR completely dominated their respective weight classes. Nowdays, in my personal opinion, the TBR almost feels like it is to the Clans what the Cataphract is to the IS. Out-dated, and gradually removed from the "meta tier list". With dakka and ppfld becoming the meta, the TBR has been power-creeped by the newly arrived Night Gyr, a mech that does dakka and ppfld better. Not saying the TBR is a bad mech; heck, I don't regret buying them, I love those things. It is a flexible chassis with many builds, perfect for a newcomer like me who like to switch things up once in a while without having to buy entire new chassis, and its a good energy boat and brawler. Just that those two roles are not the meta at the moment, and the its lack of high-mounted direct fire hardpoints (excluding the A left torso) hurts its "meta-compatibility" in similar ways as the Cataphract's hardpoint configurations.

You know, that is a pretty fitting comparison...

The Phract was a Meta God once upon a time, not dissimilar to the Victor and HGN
It could even (lulzily) do the 40 PP FLD 2Gauss+ERPPC

But, nerfs too their tolls, as did Power Creep

The Timby is still a fearsome robot, but nowhere near the initial invasion.
Far more blanket nerfs have assured that (and rightfully so)
But, it can stand its own against the Gyr. Both are likely to be chosen, for different roles. The speed and brawling potential of the Timby far exceed the Gyr (think SRM+SPL at 90 kph), and the main guns can also be higher mounted, but the Gyr brings the guns, no doubt.

Comparatively...the Phract is dead. Outclassed completely by the Hammy.

[El Bandito]

Dee Eight, on 27 October 2016 - 10:01 PM, said:

Furthermore, and I and others have said this repeatedly... PGI also has access to all the real numbers for the mech, and they clearly do not think its OP as they haven't done anything to it since they removed its quirks three months ago.

PGI is notorious for letting OP mechs/builds stay that way for many months. You haven't been here for the majority of the game's time period. 2013 and 2014 were especially horrible. In fact, without people spamming the forums with complaint threads, PGI wouldn't have balanced those OP mechs/builds in the first place.

Edited by El Bandito, 27 October 2016 - 10:18 PM.

[Deathlike]

Dee Eight, on 27 October 2016 - 10:01 PM, said:

As has been pointed out...I play a LOT of games each month, A LOT more than any of you...thus I see a fair number of KDK-3s.. and the OP nature a lot of you are demanding must be real based on the performance of a handful of players and a small sample size of games...isn't actually how the mech performs in the really real world amongst the majority of players using them. Furthermore, and I and others have said this repeatedly... PGI also has access to all the real numbers for the mech, and they clearly do not think its OP as they haven't done anything to it since they removed its quirks three months ago. They've merely balanced it to be the same as the KDK-1 and KDK-5 with only a CT structure quirk. Typically the only time they turn in good results is if they happened to be on the winning team.

Playing a lot of games isn't the same as playing a lot of quality games, let alone winning games.

PGI leaves metas long overdue than they should, and frankly are not the most trustworthy sources on a lot of things. Balance happens in long intervals (60 to 90 days joke is always applicable here).

Last I checked, I have winning records in Lights consistently since that garbage leaderboard had started.
ARCTIC CHEETAH ACH-PRIME 1,186 795 391 2.03 1,215 568 2.14 340,374 1,678,124 4 days 18:52:51

I would consider myself slightly above average, and have probably more insight about Lights than the average potato that complains about Lights being OP (because, that's silly - died enough in higher levels of play because Lights TTK is a pathetic joke, especially when you're bad).

I don't need people that claim Kodiaks are not OP when I have teammates that have an effing clue AND are running Kodiak-3s. It's undeniable for people that have a clue. You don't wish to have one, so please tell someone that has no clue how much Kodiak-3s are irrelevant. I'm sure they'll listen to your mediocre (or below mediocre) drivel.

[Zergling]

Dee Eight, on 27 October 2016 - 09:20 PM, said:

Ah I see...when I get good results its merely luck with random teams...when one of you tiertards does it, its got nothing at all to do with the luck of random teams.

That's because there is a correlation between damage/battle, W/L and K/D.

You have very low damage/battle figures for your W/L and K/D stats in those mechs, so those statistics are outliers caused by random chance.
Further, those figures are above your leaderboard averages, so they are not representative of your actual average scores; you cherry-picked mechs you have good stats with to claim you are a better player than you actually are.

With El Bandito and Mcgral18, their posted stats in the KDK-3 show a strong correlation between damage/battle, W/L and K/D, so the stats they produced are not outliers.
Further, Mcgral18 has over 400 battles in his KDK-3, and the probability of luck having any influence on his stats after that many battles is almost non-existent.

Dee Eight, on 27 October 2016 - 09:20 PM, said:

sorted by W/L ratio...season 3 and 4 (you have no 1 or 2 games)

11036 Zergling 133 104 1.28 197 159 1.24 238 233

10107 Zergling 195 153 1.27 340 193 1.76 348 260

That's not particularly amazing results in my opinion. Maybe it impresses others here who wanna like your posts but it doesn't impress me.

I never claimed to be a great player; I'm probably above average, but not great. Nor is my performance relevant to yours.

Further, I posted my statistics in relation to the Summoner being top tier, by showing how I was performing above the norm for my skill level in it.

As for not being impressed, I could care little about your opinion, but you seem to have little idea of what those stats actually say.
I refer again to the statistical analysis of leaderboards.

Global upper 25% in match score = 235
Global upper 25% in K/D = 1.11
Global upper 25% in W/L = 1.18

Global upper 10% in match score = 268
Global upper 10% in K/D = 1.67
Global upper 10% in W/L = 1.42

So in my first month of play after taking a break for a year, I was hitting upper 25% in average match score, and above the upper 25% in K/D and W/L.
In my second month of play after taking a break for a year, I was hitting upper 10% in average match score and K/D, and above upper 25% in W/L.

But keep on deflecting this onto my stats if you like, even though my performance is entirely irrelevant to this discussion.

Edited by Odanan, 13 November 2016 - 06:16 AM.
[redacted]