Making Poor Mech Variants Work
#1
Posted 11 May 2013 - 01:50 PM
Anyway, lately the min-maxing in MWO has been annoying me to no end, and since I honestly can't compete with the PPC+Gauss pros out there, I've decided to go completely the other way and play my least successful mech, the Raven 4X.
I have to say it's been a blast. Every kill is a hard fought battle, every match survived is a job well done. The few matches where you actually get 2 or 3 kills, barely limping to safety with your smoking MGs, your heart rate is through the roof. I'm really enjoying it, even though it's still frustrating to get one-shotted by 4 ER PPC stalkers now and then.
I've even managed to raise my success rate. At 50 matches, my W/L rate was 0.78, my KDR was 0.32. At 100 matches, my W/L rate is now 1.04 and my KDR is a perfect 1.00.
(My key to success has been this build, and only playing Assault. Ironically, Assault is safer for a Raven 4X, as there are no Raven 3L swarms who dare play with the big boys)
Anyone else out there trying to hit the magic 1.00 ratio with the ugly ducklings of MW:O?
#2
Posted 11 May 2013 - 02:03 PM
#3
Posted 11 May 2013 - 03:44 PM
Alistair Winter, on 11 May 2013 - 01:50 PM, said:
After around 100 matches, you should hit the magic 1.00 W/L ratio with any mech; that's the magic of Elo. Of course, get back into a good one in the same weight class afterwards and you'll be a god amongst beggars. You learn the mech, and make the most of it, but as you keep losing when the MM expects you to win, your Elo rating drops and you end up facing ever worse opponents until you start winning again.
This is why those people who say "Oh, X weapon/mech chassis is fine, I do just great in it!" are totally missing the point. You'll do fine in any mech eventually.
Right until you try to play in, say, an 8-man where Elo matching doesn't exist, or at odd hours of the night when the player pool is smaller. Then you're right screwed.
#4
Posted 11 May 2013 - 03:49 PM
Alistair Winter, on 11 May 2013 - 01:50 PM, said:
I've given up on my ugly ducklings; the non-ECM Spiders. Their WLR and KDRs are 1.43/0.29 and 0.68/0.25 (I'll let you make an educated guess as to which one's the 5V and which one's the 5K...). As a comparison, my 5D is 1.33/1.28.
But I have been chasing the magic 1.0 with a self-imposed ugly duckling - it's not a bad 'mech, but it's a rather weird build: My COM-1D Ghetto Panther PNT-10K. Getting kills in a Commando with just an ERPPC and ASRM-4 is rather challenging in the current game-play.
#5
Posted 11 May 2013 - 03:59 PM
NRP, on 11 May 2013 - 02:03 PM, said:
Well, if you have streaks, it's perhaps the second best Cicada out there. So not exactly an ugly duckling as such. If you're using SRMs, I tip my hat to you, sir. If you're using LRMs, you're probably a weirdo who shaves his cat.
Wintersdark, on 11 May 2013 - 03:44 PM, said:
I don't really understand the details of ELO. Do I have an ELO score for each mech variant, or each weight class, or how does it work?
I already had 150 matches with my Raven 3L and Raven 2X before I started dusted off my old Raven 4X. And those mechs had much better stats. So unless I have an ELO score for each mech, I'm not sure that's a factor. Indeed, my overall KDR and W/L ratio has been climbing steadily the last few months, which I hope is increasing my ELO score overall. Otherwise, the ELO system is more flawed than I thought.
stjobe, on 11 May 2013 - 03:49 PM, said:
But I have been chasing the magic 1.0 with a self-imposed ugly duckling - it's not a bad 'mech, but it's a rather weird build: My COM-1D Ghetto Panther PNT-10K. Getting kills in a Commando with just an ERPPC and ASRM-4 is rather challenging in the current game-play.
I'm sorry to hear you've given up on your spiders. A friend of mine actually managed to do very well with all of them, even the 5V. Though I must admit that the 5V was the first mech I ever sold in this game, out of sheer frustration. If it comes back as a champion trial mech, I'll give it a spin again.
However, I'm ever more sorry to see light pilots equipping PPCs to deal with the latest PPC craze. Even in a Commando 1D, which I'm sure is a very challenging mech. Kudos for using SRM4+Artemis instead of 2 streaks though. That's character-building stuff, in my book
Is the effect of Artemis on an SRM4 really worth the tonnage?
#6
Posted 11 May 2013 - 04:19 PM
Alistair Winter, on 11 May 2013 - 03:59 PM, said:
I already had 150 matches with my Raven 3L and Raven 2X before I started dusted off my old Raven 4X. And those mechs had much better stats. So unless I have an ELO score for each mech, I'm not sure that's a factor. Indeed, my overall KDR and W/L ratio has been climbing steadily the last few months, which I hope is increasing my ELO score overall. Otherwise, the ELO system is more flawed than I thought.
You have a totally independant score for each weight class. What happens is the matchmaker attempts to put teams together of similar Elo scores. If you always played 3L's, you'll have a comparatively high Elo score. This means when you first attempt to play a poorer chassis (assuming you play them equally well) you'll be outmatched initially.
As you lose more than you win, this will drop your Light class Elo score gradually. Not fast, it's a slower process, but total number of games overall doesn't matter. So long as you lose more matches than you win (this is modified by how likely the MM thought you were to win or lose a match - if it thought you'd lose initially, and you do lose, your Elo doesn't drop significantly) your Elo will continue dropping.
Because you're in a weaker mech, then, you lose a bunch of matches initially, your Elo rating drops until you reach equilibrium and you acheive a roughly 1.0 ratio again.
Incidentally, this works the other way too: If you consistently win more than you lose (when the MM expects you to lose, that is) your Elo rating continues to climb, facing you against ever more skilled opponents until once more you reach equilibrium.
Note that this is hard to see in practice, because your overall W/L/R won't reflect this nearly as fast as it happens, as it includes "old" data by necessity.
Quote
However, I'm ever more sorry to see light pilots equipping PPCs to deal with the latest PPC craze. Even in a Commando 1D, which I'm sure is a very challenging mech. Kudos for using SRM4+Artemis instead of 2 streaks though. That's character-building stuff, in my book
Is the effect of Artemis on an SRM4 really worth the tonnage?
I find the effect of Artemis on SRM4's is very much worth the tonnage. The closer spread is always good, but far, far more valuable is that is that your SRM4's are perfectly effective clear out to 270m instead of spreading past usefulness.
Edited by Wintersdark, 11 May 2013 - 04:22 PM.
#7
Posted 11 May 2013 - 05:36 PM
stjobe, on 11 May 2013 - 03:49 PM, said:
Did the 6MG Spider visit you recently? Did it tell you the true meaning of Dakka?
#8
Posted 11 May 2013 - 05:41 PM
#9
Posted 11 May 2013 - 11:23 PM
Alistair Winter, on 11 May 2013 - 03:59 PM, said:
Is the effect of Artemis on an SRM4 really worth the tonnage?
I think you misunderstand, I don't mount an ERPPC on the 1D because of the "PPC craze", I do it because it's a build that tries to get as close as possible to a Panther PNT-10K. It's supposed to mount nothing but an ERPPC and an Artemis SRM-4. Of course, it's also supposed to be ten tons heavier and mount a bit more armour, but this is as close as I get. Also, I originally built it long before the PPC buff.
I've used PPCs on Commandos since closed beta; they never were more than a novelty until the PPC got buffed a number of patches back (especially the projectile speed increase) and HSR. Now they're suddenly the best weapon in the game.
And yes, Artemis helps with the SRM-4. Whether it's worth the tonnage or not wasn't a consideration in this build.
Edited by stjobe, 11 May 2013 - 11:29 PM.
#10
Posted 12 May 2013 - 12:09 AM
Oh, and jumpjets make everything more fun!
#11
Posted 12 May 2013 - 12:31 AM
97 matches, 54 wins, 35 losses, w/l ratio 1.54, 77 kills, 53 deaths, k/d ratio 1.45, damage done 20089, 71k exp, 8 hours 59min played
max engine 245xl, 124kph with speed tweak.
2ML
1streak 1 ton ammo
1 ac2 3 tons ammo
1dhs
2 jumpjets
Basically I run it as constantly ac2 firing sniper, as well as brawling with the 2ml streak and the ac2.
Of course streak heavy lights equal death without backup but this thing is wonderful for jjing behind larger targets and shredding their armor.
It may not seem like much but since the ballistic rewind came in I have been doing well with it, something like 2-300 damage a match with 1-3 kills.
I ran only this in the last torney for the first few days and my best ranking was 53rd.
I have ran all the builds from small weapons to LL or erppcs and every single ballistic weapon there is, but I have to admit that I have had the most fun in this little sniper/scout.
Edit- about 90% pugger these days.
Just did a random match 357 damage, 6 assists 1 kill and survived.
I dont know why people say the 4x is useless.
Edited by MonkeyCheese, 12 May 2013 - 12:45 AM.
#12
Posted 12 May 2013 - 01:29 AM
MonkeyCheese, on 12 May 2013 - 12:31 AM, said:
If one weapon does 10 damage, and another does 9 - all else being equal - people will say the weapon that does 9 damage is useless.
I think that Internet Folks don't understand what the word means, to be honest. Not as good? Certainly. But that doesn't even make it bad, let alone useless.
It's much the same with ugly duckling builds. They're not as good. Well, some may be outright bad. But even if you have no weapons at all, you're not useless - you can cap, you can act as ablative armor for better mechs, you can delay a cap.
#13
Posted 12 May 2013 - 03:08 AM
Wintersdark, on 11 May 2013 - 04:19 PM, said:
Well, I don't use streaks on my Raven 3L, so it's not particularly better than my 2X. It's just faster.
Wintersdark, on 11 May 2013 - 04:19 PM, said:
Well, I certainly haven't noticed anything. And I've still been playing the Raven 2X and 3L while playing the 4X. But back in the day, I just had the wrong build and the wrong tactic with the Raven 4X. For example, I was determined to make 2AC2s and jump jets works. That was hard.
I'm not sure how much ELO factored into this, to be honest. I'm still seeing the same names I'm always playing with, I'm still doing well when playing in pre-mades (though it remains a mystery how ELO works when you're in a big group) and getting kills with my Raven 4X didn't get progressively more difficult as my stats improved - it got easier. In other words, ELO may have an effect, but it seems to have been a small one.
You do make an interesting point though, which leads me to wonder how matchmaking works. Let's say an ELO score range is between 1000 and 3000, or whatever the numbers are. So the worst players are 1000, the best players are 3000. Now, since they fixed the scoring, there may very well be a perfect bell curve to the scores. But if match making is adding anyone within, say, 500 points of the average of players already within a game, then that could easily be anyone from the majority of players. It just prevents the very best from meeting the very worst.
Wintersdark, on 11 May 2013 - 04:19 PM, said:
I'll have to try it on my Raven 2X or 3L, thanks.
stjobe, on 11 May 2013 - 11:23 PM, said:
I've used PPCs on Commandos since closed beta; they never were more than a novelty until the PPC got buffed a number of patches back (especially the projectile speed increase) and HSR. Now they're suddenly the best weapon in the game.
And yes, Artemis helps with the SRM-4. Whether it's worth the tonnage or not wasn't a consideration in this build.
In that case, I tip my hat to you. That last sentence is something I wish I heard more often in this game.
aniviron, on 12 May 2013 - 12:09 AM, said:
Oh, and jumpjets make everything more fun!
I gave up on the Raven 4X for the longest of time too. Although for me, the entertainment value is definitely not directly proportional to firepower. As I wrote in another thread, I've started using less guns to make room for STD engines instead of XL engines, and it's superb for my playing style.
Speaking of ugly ducklings though, I'm trying out the Ilya Muromets lately, and it's much harder to use than I would have thought. It's 3 UAC5s may have made it king of the hill a few months ago, but the 3 UAC5 Jager is just so much better. Hell, I even did better with my 4 UAC5 Jager. So I've gone with 3 x AC10. And in this case, the amount of firepower is actually directly proportional to entertainment value. Untill someone blows a hole through my XL 250 engine.
#14
Posted 12 May 2013 - 04:19 AM
#15
Posted 12 May 2013 - 04:48 AM
NRP, on 11 May 2013 - 02:03 PM, said:
replace beers with shots and mixed drinks and i was in about the exact same boat, except i did plan to level the cicadas eventually just not for awhile....
in the right situation almost any mech can work, and if you stick with your team and use teamwork alot more varients can also work.
#16
Posted 12 May 2013 - 04:51 AM
mekabuser, on 12 May 2013 - 04:19 AM, said:
I've been looking pretty hard, but there doesn't seem to be anything awesome hiding anywhere in the SDR-5V or the 5K.
I tried running the 5V as a pure scout, with a single ML and TAG, BAP and ESR2. It wasn't awesome, and it didn't make me any CB/XP. I tried running it with an ERLL, but that's just a poorer 5D.
I tried running all kinds of configurations of the 5K, but until MGs get a serious buff there's absolutely nothing awesome about it, and certainly not the kind of awesome that brings in the CB/XP.
I'd love for you to show me and the rest of the world where the awesome hides in the 5V and 5K.
#17
Posted 12 May 2013 - 04:53 AM
Edit: Just for funsies I decided to take out my 4MG LPL SDR-5K build out for the first time since I elited it. (Before stat wipe so I had clean stats.) After 8 games I now have 5 wins to 3 loses with 4 kills and 3 deaths. Usually putting out more damage than a few of the assaults on my team. This is horrible and wrong.
Edited by Farpenoodle, 12 May 2013 - 06:14 AM.
#18
Posted 12 May 2013 - 06:15 AM
Ph30nix, on 12 May 2013 - 04:48 AM, said:
in the right situation almost any mech can work, and if you stick with your team and use teamwork alot more varients can also work.
Actually, sticking to the team is often a death trap, I've learned.
Some of my success comes from being cynical enough to see when it's time to leave my team behind. Like when the score is 1-4, and 2 or 3 your team mates actually charge forward, as if expecting to kill the 7 remaining enemy mechs on their own.
That's when I do 2 things. 1) See if other light mechs want to escape with me. 2) See if there's any way to win by cap. If not, get the enemy to spread out and use guerrilla tactics to take out their weakest members.
A few times, your teammates will scream at you for being too much of a coward to die with them, but then you have to remember that you're a Raven 4X pilot. You can't play by their rules.
If you're talking about something like a Dragon or Awesome 8V, unpopular heavy and assault mechs that need to stay close to their team to be effective, then I agree with you.
stjobe, on 12 May 2013 - 04:51 AM, said:
I tried running the 5V as a pure scout, with a single ML and TAG, BAP and ESR2. It wasn't awesome, and it didn't make me any CB/XP. I tried running it with an ERLL, but that's just a poorer 5D.
I tried running all kinds of configurations of the 5K, but until MGs get a serious buff there's absolutely nothing awesome about it, and certainly not the kind of awesome that brings in the CB/XP.
I'd love for you to show me and the rest of the world where the awesome hides in the 5V and 5K.
I play with a guy called MechKilla. He went through kind of the same thing I've done with the Raven 4X. As he progressed, many times, we would be in a 3-man group and he would end up getting the top score or most kills on our team. With two medium pulse lasers, or sometimes 1 large laser. I played a few games with him in his Spider the other day, and he averaged 50 damage, I think. I don't think he got any kills. Some of it may be the increased accuracy of lasers and ballistics in the last few patches. But I think the most important part is that when you play 50 games straight with any mech, you'll get a real good feeling of what it can and can't do, and you'll reach a zen state, like Neo in the Matrix. But clearly, that needs to be maintained. If you get used to playing heavy mechs, you can't just switch back to a Spider 5V.
#19
Posted 12 May 2013 - 09:40 AM
Alistair Winter, on 12 May 2013 - 06:15 AM, said:
I play lights about 75% of the time (Commandos and Spiders), and mediums (Centurions, rather, fast ones) the rest. I don't own any heavies or assaults, except for my founder 'mechs. I think my light skills are pretty maintained.
Still waiting for someone to point out where the "awesome" is hidden in the 5V and 5K.
#20
Posted 12 May 2013 - 09:50 AM
stjobe, on 12 May 2013 - 09:40 AM, said:
Still waiting for someone to point out where the "awesome" is hidden in the 5V and 5K.
No, we both know it can't compete with other light mechs. The point of this thread wasn't to say that every mech in the game is equally good. Merely that playing the underdog is a fun challenge when everyone else is min-maxing. Note that I've just fought my way up to a 1.00 KDR. It's not exactly something to write home about.
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users