The primary problem with this public test, as can be seen from multiple other threads, is that in addition to adding the energy draw system, there are several other tweaks which make the effect of energy draw harder to ascertain.
These other changes include:
Heatsink changes
Weapon cooldown increases
Nerfs to AC5s and LPLs
Removing Gauss charge but nerfing its range
To anyone, like myself, with some background in running controlled experiments, this is extremely bad design. Any assessment of how energy draw will affect the meta is now conflated with all these other changes.
Is weapon X coming back into vogue because of energy draw or because you tweaked the mechanics on competitor weapon Y? Who freaking knows, because you have one measurement and two things that changed.
***
Now if PGI had a good handle on the meta and how MWO plays, having a few other tweaks in the same Public Test might not be so bad. But it continues to baffle me how PGI can't already predict the main impact of this mechanic as currently implemented -- which I can figure out just by reading the notes, let alone playing the on the test server.
3x PPCs on a fast mech with JJs is the obvious best mech. It will beat everything else, hands down, with these mechanics.
Do not pass go, do not collect $200 -- go directly to 2013's meta.


This Public Test Changes Too Many Mechanics At Once!
Started by Vincent Quatermain, Aug 18 2016 10:32 PM
3 replies to this topic
#1
Posted 18 August 2016 - 10:32 PM
#2
Posted 18 August 2016 - 10:55 PM
Vincent Quatermain, on 18 August 2016 - 10:32 PM, said:
3x PPCs on a fast mech with JJs is the obvious best mech. It will beat everything else, hands down, with these mechanics.
Do not pass go, do not collect $200 -- go directly to 2013's meta.
Do not pass go, do not collect $200 -- go directly to 2013's meta.
And if they hadn't nerfed AC5s hard, that would also be comparable. But yeah, could have called this. Oh wait, I did.
#3
Posted 18 August 2016 - 11:06 PM
I agree. They're running a test. In science, when you run a test, you try to minimize and control all the variables to find out if the one you're most concerned about plays a role in the results. So naturally they did the opposite.
#4
Posted 19 August 2016 - 12:14 AM
So, I played some PT matches and I can confirm that PPCs are super strong. The new no-delay Gauss is also crazy broken. Put them together, and sweet jebus that will be insane.
PGI set out to nerf pinpoint alpha builds and on their first pass managed to grant a massive buff to the most obnoxious flavor of pinpoint alpha builds in the game.
PGI set out to nerf pinpoint alpha builds and on their first pass managed to grant a massive buff to the most obnoxious flavor of pinpoint alpha builds in the game.
2 user(s) are reading this topic
0 members, 2 guests, 0 anonymous users