

Lets Debate - The Jumpjet Overhaul
#161
Posted 25 June 2021 - 01:16 PM
>Hoverjets are unfun to use but we gotta be careful to not make poptarting replace every other playstyle like it used to once upon a time (been there, done that, it was cancer)
>GasGuzzler once had the idea of drastically increasing JJ heat on heavies and assaults so that jumping around too much would cut into their damage output, forcing a choice between DPS and mobility (but JJ power would be buffed so they actually give real mobility in exchange)
>Another idea would be to make the JJ formula exponential rather than linear, i.e. 1 JJ should be absolute garbage but mounting the full allotment should have actually good mobility, which forces a real sacrifice of tonnage (and thus you have to downgrade your weapons or something to accommodate it)
#162
Posted 25 June 2021 - 01:27 PM
#163
Posted 25 June 2021 - 02:31 PM
Kiiyor, on 21 June 2021 - 07:32 PM, said:
Super nervous about this one. Pls no return to the Dragon Slayer poptart days.
I don't see why this would be an issue making an obsolete hero usable with jjets again for poptarting, as back then, it was a completely different meta pyramid from now. Nowadays... I wouldn't mind seeing mechs like the Dragon Slayer be able to poptart again and make use of it's mobility and jjets together like it SHOULD.
#164
Posted 25 June 2021 - 03:22 PM
#166
Posted 25 June 2021 - 04:20 PM
The previous Poptart was prior to the Clans released and it was 1 Gauss Rifle and 2 ERPCC.. for a total of 35 pts of damage!! And from humanoid mechs where there were nutcrackers and slightly higher. It sure didnt stop Khobai's Atlas from waddling across an open field crying while armor was being removed and he couldnt fire back cause he still had 500+ meters to go before he could do anything, always going to the center of the map

The original JJ setup was setup horribly, with most of the return relying on the 1st JJ, others were not really needed. Remove the quick drop in altitude by adjusting and lowering the default Gravity, then have certain maps.... nix that... have each map setup different gravity settings, some where the gravity is higher that default then those lighter than default.
Tarl Cabot, on 23 June 2021 - 06:07 AM, said:
Then with lower gravity baseline, have gravity differences on the maps, some maps the mechs take even longer to land, others more quickly. imho, this is what is currently missing.
#167
Posted 25 June 2021 - 04:22 PM
Using this build WHM-9D(S)
Current Jump / Values - HERE
Proposed Jump / Values / No skill nodes - HERE
Proposed Jump / Values / 6JJ duration skill nodes - HERE
As you will see the trust isn't ridiculous and investing in the skill tree actually becomes beneficial. You get ~33% more height out of it with each duration node giving ~6% more height.
#168
Posted 25 June 2021 - 04:31 PM
justcallme A S H, on 25 June 2021 - 04:22 PM, said:
Using this build WHM-9D(S)
Current Jump / Values - HERE
Proposed Jump / Values / No skill nodes - HERE
Proposed Jump / Values / 6JJ duration skill nodes - HERE
As you will see the trust isn't ridiculous and investing in the skill tree actually becomes beneficial. You get ~33% more height out of it with each duration node giving ~6% more height.
Could you do a video on a 100 tonner? Maybe with 3JJ? I'm concerned that it's not giving the boost you guys think it will because of an incorrect height formula.
#169
Posted 25 June 2021 - 05:53 PM
Nightbird, on 25 June 2021 - 04:31 PM, said:
Sure. I've also turned on coordinates so we can see the exact height gained.
Here is a MAD II-5A with 3JJs. No skill nodes:
Live_100t_3JJ_NoSkill
As can be seen, the mech goes from 225.3 Z coordinates to a peak of 246.6. That is 21.3 meters of height. Table I posted shows 24.08 for this case. 3 meter error.
With posted changes the same MADII with the same build and no skills will jump like this:
July_100t_3JJ_NoSkill
Here, the mech goes from 225.3 Z coordinates to 259.2 peak. That is 33.9 meters of max height. Table I posted shows 39.62. 6 meter error.
Also, as an additional test, July values are tested with the exact same mech and build, but with 5 JJ duration nodes (costs 6 skill nodes):
July_100t_3JJ_5_Duration_nodes
In this, mech goes from 225.3 Z coordinates to 267.6, which is 42.3 meters of max height.
The data in the tables and the height formula are approximations (specially when calculating the effect of "Instant boost" stats of a jump jet, the custom physics system when jump jet fuel runs out.).
Overall, the amount of improvement over live performance is consistent with the set goals.
Edited by Navid A1, 25 June 2021 - 05:54 PM.
#170
Posted 25 June 2021 - 06:13 PM
Hyped as heck and can't wait to try it out but also keeping fingers crossed that heavies and assaults jump jet could be made a tad bit faster down the road at cost of heat generation (there you go, more reasons to take jump jet shielding skill nodes).
"How fast" you ask? (Assuming jump height peak is kept to Cauldron proposal still)
Think of it this way:
Making Class III faster such that 60ton jumps as fast as a 55ton
Class II receives the same absolute increment as Class III (theoretically 80 ton will still be slower than 75 ton).
Then making Class I faster such that 90ton jumps as fast as 85ton. (so net double buff)
(Food for thought - why didn't PGI just make it such that 60 - 85ton all uses the same class of jump jets, anyway?)
I think this is something only PGI engineers are able to do but I'll just shoot anyway - any chances we could also see that jump jets starts recharging as long as jump jets are not being used? Currently, after using jump jets, jump jets only starts recharging when you are grounded.
Edited by Matthew Ace, 25 June 2021 - 06:16 PM.
#171
Posted 25 June 2021 - 06:49 PM
Navid A1, on 25 June 2021 - 05:53 PM, said:
Sure. I've also turned on coordinates so we can see the exact height gained.
Here is a MAD II-5A with 3JJs. No skill nodes:
Live_100t_3JJ_NoSkill
As can be seen, the mech goes from 225.3 Z coordinates to a peak of 246.6. That is 21.3 meters of height. Table I posted shows 24.08 for this case. 3 meter error.
With posted changes the same MADII with the same build and no skills will jump like this:
July_100t_3JJ_NoSkill
Here, the mech goes from 225.3 Z coordinates to 259.2 peak. That is 33.9 meters of max height. Table I posted shows 39.62. 6 meter error.
Also, as an additional test, July values are tested with the exact same mech and build, but with 5 JJ duration nodes (costs 6 skill nodes):
July_100t_3JJ_5_Duration_nodes
In this, mech goes from 225.3 Z coordinates to 267.6, which is 42.3 meters of max height.
The data in the tables and the height formula are approximations (specially when calculating the effect of "Instant boost" stats of a jump jet, the custom physics system when jump jet fuel runs out.).
Overall, the amount of improvement over live performance is consistent with the set goals.
Thanks Navid, please feel free to use the formula I provided to tweak the JJ values to hit the desired heights. I also provided my opinion on JJ changes for your group's consideration.
#172
Posted 25 June 2021 - 08:21 PM
Nightbird, on 25 June 2021 - 06:49 PM, said:
I checked your formula. I noticed that you are considering gravity when calculating the upward acceleration. When a mech is burning JJs, gravity is ignored. It's easy to test since in most cases your formula results in a negative acceleration upwards.
Also, when JJ fuel runs out, the game uses a custom velocity rule to quickly cut upward movement and puts you in free fall. It very visibly does not follow projectile movement.
I should have posted a big note saying that do not try to makes sense of JJ movement using real physics... since PGI is using a lot of custom rules.
MWO JJ mechanics do NOT follow real physics of how projectiles work. The formulas I posted are not based on physics... at least, partially not. there are a lot of extra factors. Otherwise I'm fully aware of how projectiles work and what should be the unit of each variable.
The only part that I need to fine tune is the effect of instant boost (in the game file) which I'm currently translating as initial velocity, and the custom physics rule governing the end of JJ burn time to cut the upward speed.
Edited by Navid A1, 25 June 2021 - 08:31 PM.
#173
Posted 25 June 2021 - 08:43 PM
Navid A1, on 25 June 2021 - 08:21 PM, said:
Yep, I figured this part out and posted the updated formula here:
https://mwomercs.com...ost__p__6412932
I also used the formula to calculate the current ingame values (which checks out) and the Cauldron's values, which seems to as well.
https://mwomercs.com...ost__p__6412934
https://mwomercs.com...ost__p__6412938
Edited by Nightbird, 25 June 2021 - 08:43 PM.
#174
Posted 25 June 2021 - 08:53 PM
#175
Posted 25 June 2021 - 09:24 PM
it highly irritating that you have to compensate for the unwanted forward thrust whenever youre jumpjetting I wish there was a way to turn it off.
if you want more forward thrust on your jumpjets again thats what the jumpjet skill tree is for. it just needs some buffs so putting points into it is a little more worthwhile.
Edited by Khobai, 25 June 2021 - 09:29 PM.
#176
Posted 25 June 2021 - 09:43 PM
Nightbird, on 25 June 2021 - 08:43 PM, said:
Yep, I figured this part out and posted the updated formula here:
https://mwomercs.com...ost__p__6412932
I also used the formula to calculate the current ingame values (which checks out) and the Cauldron's values, which seems to as well.
https://mwomercs.com...ost__p__6412934
https://mwomercs.com...ost__p__6412938
I missed the updated post. But even then, the tables you posted seem inconsistent with the formula (which is similar to mine minus the last term corresponding to momentum based increase in height, which as I said is not how a projectile works and is custom physics)
let's take the formula:
max height = bd*i0 + 0.5bd^2(i1) + [i0+bd(i1)]^2 / (8g)
or in cauldron's terms:
= Burn_dur * Thrust_init / mech_ton + 0.5 * Burn_dur^2 * (num_JJ * Thrust / mech_ton) + [thrust_init/mech_ton + burn_dur(num_JJ*thrust/mech_ton)^2 / (8g)
For a 100 ton mech with 3JJ Class I in live client:
bd = 3.75
i0 = 2.2
i1 = 2.25
max height = 8.25 + 15.82 + 1.44 = 25.51m
while in the table it is listed as 21.2
For a 25 ton mech with 6 Class V JJs:
bd = 3.75
i0 = 4.4
i1 = 8.88
max height = 16.5 + 62.44 + 18.13 = 97.07
while in the table it is listed as 80.6
I don't know what is the source of this inconsistency.
Also, I'm beginning to believe that thrust_init is not the initial velocity of the jump, but an initial high acceleration phase. I'll just have to pinpoint how much of the jump duration is dedicated to that initial "hop"
Edited by Navid A1, 25 June 2021 - 09:44 PM.
#177
Posted 25 June 2021 - 09:47 PM
Navid A1, on 25 June 2021 - 09:43 PM, said:
I missed the updated post. But even then, the tables you posted seem inconsistent with the formula (which is similar to mine minus the last term corresponding to momentum based increase in height, which as I said is not how a projectile works and is custom physics)
let's take the formula:
max height = bd*i0 + 0.5bd^2(i1) + [i0+bd(i1)]^2 / (8g)
or in cauldron's terms:
= Burn_dur * Thrust_init / mech_ton + 0.5 * Burn_dur^2 * (num_JJ * Thrust / mech_ton) + [thrust_init/mech_ton + burn_dur(num_JJ*thrust/mech_ton)^2 / (8g)
For a 100 ton mech with 3JJ Class I in live client:
bd = 3.75
i0 = 2.2
i1 = 2.25
max height = 8.25 + 15.82 + 1.44 = 25.51m
while in the table it is listed as 21.2
For a 25 ton mech with 6 Class V JJs:
bd = 3.75
i0 = 4.4
i1 = 8.88
max height = 16.5 + 62.44 + 18.13 = 97.07
while in the table it is listed as 80.6
I don't know what is the source of this inconsistency.
Also, I'm beginning to believe that thrust_init is not the initial velocity of the jump, but an initial high acceleration phase. I'll just have to pinpoint how much of the jump duration is dedicated to that initial "hop"
As I posted in the first table, there seems to be a constant factor of 0.83 between predicted numbers and in game numbers. Rather than try to math it out, I just accepted it as a game engine quirk. After all, our numbers are unitless. From class 1 to class 5 numbers, it all looks good once .83 is applied.
The real test was the 100 ton 3jj with the Cauldron numbers, which I predicted to be 34.9 but couldn't test. From your 34.1m result, I'm off by 0.8, which is less than 1 meter. That's why I think it's useable and further refinements may not be worth the time. You can try 0.82 and it may be even more perfect, I didn't do a check of all values.
Edited by Nightbird, 25 June 2021 - 09:54 PM.
#178
Posted 25 June 2021 - 09:50 PM
Nightbird, on 25 June 2021 - 09:47 PM, said:
Have to check that boost_inst variable to make sure it is initial velocity
That is the source of inconsistency... I'll have to reverse calculate the initial time period
Edited by Navid A1, 26 June 2021 - 03:02 AM.
#179
Posted 25 June 2021 - 10:40 PM
FupDup, on 25 June 2021 - 01:16 PM, said:
>Hoverjets are unfun to use but we gotta be careful to not make poptarting replace every other playstyle like it used to once upon a time (been there, done that, it was cancer)
>GasGuzzler once had the idea of drastically increasing JJ heat on heavies and assaults so that jumping around too much would cut into their damage output, forcing a choice between DPS and mobility (but JJ power would be buffed so they actually give real mobility in exchange)
>Another idea would be to make the JJ formula exponential rather than linear, i.e. 1 JJ should be absolute garbage but mounting the full allotment should have actually good mobility, which forces a real sacrifice of tonnage (and thus you have to downgrade your weapons or something to accommodate it)
The only thing that needs to be done ot prevent Poptarting is no more "Vertical Only Jumping" - the moment every Jump has so much forward thrust that you automatically jump out of cover every time using JJs this Poptarting thing is no thing anymore.
#180
Posted 25 June 2021 - 10:45 PM
Thorqemada, on 25 June 2021 - 10:40 PM, said:
That won't do anything.
All you do then is just ensure you start your jumping 5-10m further back to account for it. The more forward, the further you stand back. Or even just be in reverse before it...
So many ways to counter that it won't ever work.
Edited by justcallme A S H, 25 June 2021 - 10:45 PM.
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users