Stat1cVoiD, on 23 January 2016 - 08:26 AM, said:
Gman stated everything clearly from a/his competetive point of view and - regarding this standpoint - everything he said is true. And 95 % of all posts in this thread don't even try to contradict his arguments, but just blindly flame him for playing competetively and judging a map around those factors.
The map does indeed favor high hardpoints immensely and punishes low mounts harshly. Even if you find a corner to peek within the hills, the shot might still land in the next hill. And yeah, you can go to some of the few buildings to compensate for this problem, but that still means that a chassis with high hardpoints is way more flexible in picking its position than one with lower ones. Other maps are way more flat and punish suboptimal convergence considerably less.
Brawl loadouts can work if your opponent also runs mostly mid or shortrange builds and there can happen some really nice fights up and down the hills. If however one teams consists mostly of ERLL boats and decides to kite and snipe you, there is just not enough cover to close the 900m gap without opening up to their LoS over and over again and getting widdled down without any chance to fight back.
LRM Cheese is real. Period. Cover is just not good enough and I am again not talking about 1 or 2 LRMtards who wait blindly behind a hill for a light to kill em, but about 5 to 6 boats which kite you to your doom.
If played halfway decent, QKDs or BLRs go just completely ham on this map.
And this has little to do with "thinking" or "tactics", but with the fact that this map offers extremely open 360° firing lanes on a lot of spots for mechs with high range and hardpoints. You don't have to be a brilliant mind or a master tactician to realize you just have to move 400m to the left to look into a certain trench. As long as your range is high enough its extremely easy to lock your enemy down in a lot of spots, because of how linear a lot of thetrenches are laid out.
So imho most of Gmans criticism is justified. If you however say you like the map because you like sniping, SkillRMs, Snow or running for several minutes without seeing a single enemy, thats your personal preference and fine... But it doesn't change the fact that the map is imbalanced from a competetive standpoint: JJs are mostly useless, ERLL and LRM reign supreme if played halfway decent and Brawl or Midrange Builds can become pretty obsolete. If you like playing on huge maps or not however is just personal preference.
Ever heard the "arena style shooter" argument? Some people waited years for a map like Polar, some gave up hope for sth like that long time ago. Some left the game, but let's not talk about those last ones as they're not here.
Anyway, after all this time, a miracle happened. PGI managed to make a map that actually is like those people wanted all along. A map they've gave up hope for. These guys are wetting their pants from happiness.
aaaannnd
Some other people don't like it because reasons. These are ignored, nothing can please anyone
buuuuut
A well-known competitive(s) brings up the idea of "fixing" that map by making it more like ALL THOSE OTHER MAPS WE HAD FOR YEARS . Competitives, as I heard, have PGIs ear sometimes. So certain people throw a fit - they've just got their dream come true map and some people PGI just might listen to want to destroy it.
Don't be suprised, this game was founded and survived mainly on people's sentiment. So strange many people playing it are emotional about loosing their magic little dream-come-true map? Just because some of the top-level-super-duper-meta-competitive per mill of the players find it bad?
--------
EDIT: Comps have map voting too, and map select in private matches. We should just settle on a compromise that people acknowledge not all maps can be liked by everyone. Besides, maybe simply agree with PGI or whoever that only certain "super competitively balanced" maps should be used in tournaments and such, while leaving those chaotic pug and <12 group drops with all the maps there are. And let them decide with voting.
Edited by Prof RJ Gumby, 23 January 2016 - 09:32 AM.