Jump to content

Why The Clans Collapsed And How To Fix It

Balance

102 replies to this topic

#101 Celtic Warrior IS

    Member

  • PipPipPip
  • Survivor
  • 70 posts
  • LocationOperation Bulldog - Tranquil - HI

Posted 02 January 2016 - 12:58 PM

View PostSader325, on 31 December 2015 - 09:05 PM, said:


Maybe it's your inability to read but SWOL takes on players of all skill level so their win rate will be lower then many units. SWOL also fields many 12 man groups many of which use trial mechs until they can afford/build those mechs up.

Looks like a Steiner village lost its idiot, your embarrassing yourself with moronic posts like this.

#102 Deathlike

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • Littlest Helper
  • Littlest Helper
  • 29,240 posts
  • Location#NOToTaterBalance #BadBalanceOverlordIsBad

Posted 02 January 2016 - 01:08 PM

View PostMcgral18, on 02 January 2016 - 12:32 PM, said:


20% of the CT mounted (which should also include the PoorDubs?)


It's just the engine DHS, not external IIRC.

Your poordubs™ are safe.


View PostAresye Kerensky, on 02 January 2016 - 05:38 AM, said:

I'm always in 100% tryhard mode. I like to win above anything else. I got 1st place for TBRs in the OCT14 leaderboard, and I've spent pretty much all of 2015 playing CW on the Clan side with CWI. The following is a brief rundown of what I have observed in terms of balance since Clans came out:

- Soon after Wave 1 hit, a group of Lords and many known top tier players held private matches pitting the Dragon Slayer against the TBR. The TBR held the edge and won the majority of games, but these were not total landslide victories, and IIRC the VTR-DS side won at least one or two of them. This was at the absolute pinnacle of Clan superiority. There were no Clan nerfs, jump jets weren't "hover jets" yet, and the TBR bunny hopping animation was completely broken. The important thing to take away from this is that the TBR in its most OP state had a slight to moderate advantage over the VTR-DS, which was considered the best mech until Clans hit.

- The last time balance was tested via solo queue, Clans won 64% of matches, but with an average Elo advantage of 100-120 (source: http://mwomercs.com/...balance-update/). We know Elo was a flawed system, however the Wikipedia entry for Elo does state that an opponent with an Elo advantage of 100-120 points SHOULD win 64% of the time. Russ stated that he was happy with those results, likely because they matched up with the predicted Elo outcome. This was after the first quirkening. The important takeaway to remember is this was before any Clan nerfs were implemented, and was also the LAST time PGI did any sort of official testing by deliberately pitting IS vs. Clan in the solo queue.

- In Tukayyid 1 the Clans won 52.2% of all matches, which I felt was actually LOW, given that the IS side was stacked with an average of 40-50 teams on defense, and the majority of them populated by solo players. The competitive premades on the IS side such as 228 therefore had much longer wait times between matches, and would not have been able to drop as often as a Clan premade. In short, Clan premades had a greater chance of running into pugs, and could drop more often = more wins. IS premades on the other hand, had a greater chance of running into a Clan premade, and would drop less often = less wins. This is why I feel the win % should have been higher for the Clans. The problem is PGI was using these metrics to judge balance, which led us to...

- The first phase of heavy handed nerfs came to Clans, including laser duration and cooldown increases for omnipods on the TBR and SCR. I believe somewhere around this time was also when side torso loss resulted in the loss of truedubs for the Clan XL, a penalty that mostly affected asymmetrical chassis and loadouts such as the HBR, but still caused symmetrical loadouts to suffer to a certain extent, as losing truedubs is more devastating than losing external heat sinks. Some of the negative quirks were scaled back from -3% per energy hardpoint, to -2% per energy hardpoint.

- A bunch of minor quirk changes happened between the original quirk pass and Tukayyid 1, in addition to a few minor quirk changes between Tukayyid 1 and the big rebalance. Overall they weren't too gamebreaking, with the exception of the TDR-9S quirk that got scaled down shortly thereafter.

At this point, I "personally" felt that balance was fairly decent. I joined MercStar for one of their IS tours and I took a bunch of the quirked IS mechs to get a feel for them, but my performance was about the same. Similar kills and damage for both IS and Clan side.

- The great rebalance before Tukayyid 2 is where things went wrong IMO. I think the quirk changes for IS were done well, in that it added more IS mechs total that performed well against the Clans, with the exception of the structure buffs. I feel the Clan XL speed/agility nerf was a little over the top at 20%, but was definitely needed. I personally feel if the weapon quirks for IS stayed, the Clan XL nerf stayed, but IS structure buffs and Clan laser nerfs were removed, we would have good parity. The extra structure + across the board Clan laser nerfs is what I feel brought things out of balance and in favor of the IS side.

- Tukayyid 2 Clans won 56.2% of matches, which should not have been possible from a balance perspective, given that IS only got stronger, and Clans only got weaker between the two events. This is a clear indication that Tukayyid 1 very likely had flawed metrics, and PGI may have acted on bad data, meaning the last REAL data gathered regarding Clan and IS balance, was the solo queue test in which Clans won the exact percentage they were SUPPOSED to win, given the average Elo advantage they had at the time.

Simply put, if that last solo queue test was correct, then the Clan mechs today are most certainly at a disadvantage. That's my opinion based on the historical events outlined above.

I justify my above opinion based on my personal observations and experiences:

- For the entirety of my time playing Clan, which was most of 2015, I had only managed to achieve a 4000+ damage game once, and part of how I accomplished that was by farming 2 AFK mechs. Within just one week of switching to IS however, I had two 4000+ damage games within 1 week, one of which surpassed my previous record damage as Clan, and neither of which involved farming any damage from AFK mechs.

- I've been stomped by superior teams on both IS and Clan sides, but I've noticed while being stomped as IS, my team's score is generally higher on average.

- I've led various pug teams to victory against large premades of 10-12 on more than 4-5 occasions during my first week playing as IS. I've also had multiple close games against large premades, one of which was a 2-3 kill difference match against a large group of KCom. During my time playing as Clan, the number of times I've led a pug Clan team to victory against a large IS premade is exactly: 0

- Most likely this is confirmation bias, but while playing with very good players in Church of Skill, SiG, or SJR, the times we would lose seemed to correlate with how many Clan mechs we took. On one night in particular in CoS, we had beaten a large group of SJR 2/3 times. The two times we won, it was a landslide victory, one of which was a 12-0. The 1 time we lost, we were running 5 TBRs. Once again, was likely confirmation bias, but there could be a correlation. Will continue researching.

Long post is long, but this is a general write up of how I've come to my opinion on the current state of balance. I generally enjoy the aestetics of Clan mechs more, so in a perfectly balanced world, I would run Clan the majority of the time. I actually enjoy the challenge of facing adversity, even though I said earlier I'm always in 100% tryhard mode. Being a tryhard and always being competitive doesn't mean I'll always take every single advantage possible. The problem right now is I've basically shelved my Clan mechs, so it's not only that I feel IS might hold a slight edge now, it's that I feel my Clan mechs will literally hamper my ability to win the game.

Once again, this is my opinion. I'm not trying to persuade people to change their opinions. I just want a freaking REAL BALANCE TEST from PGI. No "CW Metrics" that are influenced from a flawed gamemode with a dynamic population and skill disparity. No "Solo Queue" IS vs. Clans with an Elo/PSR deviation. I want PGI to do another solo queue test, and I want them to run it AS LONG AS POSSIBLE. Then, I want them to ONLY use the games where each team's PSR is as close to EVEN as humanly f***ing possible, with enough data to form a SOLID conclusion, and go from there.


TBH, PGI has been primarily operating on flawed metric analysis, in conjunction with terribly bad quirks (and the process).

Even when we had that somewhat lame hardcore unit challenge, it only affirmed that PGI's metrics tagging... *cough* stealing *cough* planets was totally broken, so in context... it's honestly how some stuff was interpreted.

There's an obvious reason that the Thunderbolt-5SS was taken primarily during both Tukayyids (more in the 1st version)... it's mainly because the alternative options sucked. There were actually a few more options (I'm told the BK and Grasshopper is better for it nowadays, though I'm skeptical for the BK), but people were going for a tried and true mech, despite the actual range nerfs that become far more self-evident (I stopped using the Wubbolt altogether for CW personally).


Flawed data analysis for flawed conclusions, when you consider "solo queue" for balance (in that Clans test) is seriously akin to balancing for "Tier Paul". No good data or good quirk changes would result for that... at best there is "balance by accident".

Edited by Deathlike, 02 January 2016 - 01:12 PM.


#103 Aresye

    Member

  • PipPipPipPipPipPipPipPipPip
  • Heavy Lifter
  • Heavy Lifter
  • 3,462 posts

Posted 02 January 2016 - 01:31 PM

View PostMischiefSC, on 02 January 2016 - 10:56 AM, said:

It's way more about who you play with than what you play right now. If you're playing IS as Kurita you're getting a very specific view. Come play Davion for a bit and pug, run with skittles on the Clan front or puggles in IS vs IS then tell me what you think.

I've played IS as Davion, FRR, and Kurita, for a total cumulative time of ~ 2 months, most of it solo drops, including many IS vs. IS fights. In terms of difficulty, I would rank my experiences as follows:

Easiest - IS (pugs) vs. Clan (pugs)
Average - IS (pugs) vs. IS (pugs)
Hard - IS (pugs) vs. Clan (unit)
Very Hard - IS (pugs) vs. IS (unit)

View PostMischiefSC, on 02 January 2016 - 10:56 AM, said:

There have been several seasons of league matches since then over the ebb and flow of rebalancing that have confirmed, repeatedly and consistently, that Clan tech was OP relative to IS tech. Ironically it still is; just that a select set of variants of select chassis with specific loadouts have been quirked on the IS side to bring them up to comparable in select situations.

It depends on the game plan. If we're talking playing a ranged game, a team utilizing Clan mechs is setting themselves up for failure. I had the relative displeasure of having to play a Clan vs. IS game in NBT where the attacking team picked Alpine Peaks for their landing phase. We took HBRs and EBJs with CERLL builds, fully complimented with targeting computers and range modules. The other team took 4 RVN-4Xs and 4 TDR-5SSs (back when it still had the 25% range quirk). Believe me when I say there was absolutely NOTHING we could do. Three 0-8 losses in a row. They had both the advantages in range, and laser duration.

I don't see much difference in saying that the IS only has a few chassis and loadouts that compare, when by "compare" you're talking about being able to stand up to the Clan's few good chassis and loadouts.

IS and Clan both have good and bad mechs. The problem I have with Clans is every good mech has virtually the same loadout and plays exactly the same. That and the fact there's only 3-4 Clan chassis worth taking in the first place.





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users