How do you try out a new build?

Curious what methods and measurements people draw on when analyzing a new mech that are a bit outside the usual glance at cooling efficiency and blatant copying of builds they are linked to/killed by.

Personally, I'm a huge fan of the Forest Colony Snow speed run. I first picked it because you used to not get pilot efficiency perks in the testing grounds, so I had to use a cold map to get a realistic performance. Also partially subconsciously, as it has the longest name and you pick it out of the menu easily.

But I still go back to it now, it gives a nice feel of what the mech can do under optimal circumstances. There are mechs at medium and long range so you can shave some off your completion time if your mech has some range to it, as well as 2 mechs hidden behind corners that brawlers can dispatch with super cool AC20/SRM racks. I take the results with a grain of salt, but a speed run through Forest Colony Snow is the first place I take a newly built mech, and its not just to set weapon groups. I find a lot of intricacies can be identified without relying on a PUG drop to be illuminating.

 

I generally run around a random map (usually Frozen City Night) to set up my weapon groups since the UI doesn't differentiate between left and right side very well. Once that's done, I'll spend a few minutes getting used to the acceleration/deceleration of the new mech, since if I can adjust to that, I'll play better once I start unlocking efficiencies. Same thing goes for turn radius and torso twist ability. Then I go find the Atlas, Cataphract, and Awesome on that map, and see how much ammo/time it takes to kill them from the front, and make decisions on my ammo load based on that information.

After some last minute adjustments, I usually just go straight into play from there, since actual combat metrics generally provide a clearer picture to me than the testing ground, and moving targets are a lot harder. The first few matches, I warn people that I'm in an unbasiced mech, and thus will not have quite the effectiveness that I would otherwise.

Half the time, in PUGs, this warning is unnecessary, since I end up doing more damage than them anyway

 

Frozen City Night to test kill times, set weapongroups, and get used to the groups. Then go farm.

 

I test on Caustic Valley where I set weapon groups and get used to things.

If I can handle the heat on that map I can really lean on the weapons in any condition.

Then I go in to PUG drops. I don't go for the whole "I live my life to play MWO 12 mans" thing any more so PUG dropping, and being successful, is good enough for me. Some times I get singled out and primary'd and other times I wreck face. Its a game so I try to do the best I can but don't get terribly upset when a 12 man comes and stomps the PUG group I'm with in CW.

 

I just drop into solo pub and see what she can do in combat.

 

I use the testing grounds to make sure the weapon groups are correct. Then I just run several (10+) solo queue matches, write down K/D/A/components destroyed for each, Then calculate average + std deviation for that.

 

It varies, I buy mechs in groups of three mostly. Only take the first into testing grounds, random map. Check weapon groups, weapon hardpoint location and verify my loadout performs per theory (quirks and modules). Especially for chassi with bullet speed variations.

Then I do ten drops, build the second chassi, do ten drops, etc.

I don't save stats, beyond a screenshot of remarkable games. Many times a video of the first runs, specially if I have a build thread.

 

Forest colony traning because it's heat neutral and it's the first map on the list so I don't have to search

 

I always do a few drops on Forest Colony. Lazy as I am, it is the first in the list and luckily for my testing also the one with the most average ambient temperature.

Contrary to my usual love for numbers and stats, when testing a new build, I do not measure time to get kills or whatever other metric one could think of. I just use the training grounds runs to get a "feel".

Commando
Cicada
Catapult
Atlas

Those targets usually are the only ones to die during testing. Only when I get a particular good or bad "vibe" with a build will I continue the tour of destruction. Just to be sure.

Now in what I look at I think is quite standard. How fast does my heat go up, how fast can I dispatch the targets, how does my movement relate to my accuracy (yes I simulate turn fighting with static targets ), how long does my ammo last,...

After that I do some PUG drops to see how much of the good vibe remains during being stomped, during good games and during "my-team-is-just-fighting-each-other-to-get-the-kills" drops.

Then some lance drops to see how it fares against organised enemies.

Then I state whether or not I like the build, check if it's on here, if not, post it.

That, or I just theory build and post it

 

Not quite right. It does to a certain degree. An example:
If you have 3 weapons in the right arm, weaponslot 1,2 and 3 are those weapons. Then you have 2 more in the left torso and the big one in the left arm... 4 ist the left arm. 5 and 6 the left torso.

IIRC the weapons aligned as RA/LA/RT/LT/CT/H. So if you know how many weapons you have in every body part you can assign them in the lab.
But it's easier to do in the testing grounds.

 

I just run it in PUGS... and tune it between games.

 

Mainly I have trouble when there are a lot of torso HPs...it alternates randomly between left and right in a lot of cases. For instance, I have a TBR where some of the weapons go like this: (1 left, 2 right
1
1
2
2
1
2

Very annoying not to do in test

 

Sometimes its good, sometimes its not. I prefer to test just to be sure. That way I don't waste ammo/FF at the start of a match.

 

That's how I do it.

I do a quick zen hash out of my weapon groups while walking out of spawn and if the sides don't mesh like I think they should (note: the first time i fire them is at an enemy) I desperately try to fix them in the thick of it.

I'm a good Mechwarrior.

 

I run the new mechs on Tourmaline Desert, HPG Manifold and/ or Crimson Straits.
1 hot map,one cold and one neutral map.
Next step PUG drops then further tweaking if necessary.

 

My conceptional test process/method (which i should use more often):

(0. Design theoretical Builds after knowledge by experience)
1. Chose one of the theoretical Builds
2. Determine the critical aspects (metric) & success criterias (thresholds) e.g. for:
- Kill Effectiveness
- Heat Efficiency
- Survivability/Sturdiness
- Usability/WGs
3. Chose Game Mode (PUG/CW) *
4. Do Testruns (e.g. 20+) incl. Collect Result Data (Game Scores(K/D/A/D)/Time to OH in Engagements/Hitbox Damage Spread/WG Handling Errors/etc.)
5. Analyse the Data and Evaluate by comparison with similar Chassis/Builds Testresults

* in PUG Mode i usually don't chose a specific Map (or better said - i don't only collect Data from a specific single map), since the game choses map for me anyways. Testscenario should map the actual game process.