6 replies to this topic

[Kin3ticX]

I was goofing with skill trees and decided to share how someone goes about figuring out whats going on with heatgen and coolrun.

Step 1: Go to smurfys, build your mech, goto weapons lab, get 2 numbers.



Step 2: Now you are going to compare two skill tree builds where you are wondering which is better. Obviously if you have the points to get max coolrun and max heat gen, you dont need this. If you can't get max of both, this might come in handy.

I built two trees on a HBR, take note on how many heat gens and coolruns you have in each test tree

tree 1: 7 heat gens, 5 coolruns

tree 2: 11 heat gens, 3 coolruns

Step 3: Complicated (not really) Mathz



here is a link to the simple spreadsheet (sorry its ugly i made it fast)
https://docs.google....dit?usp=sharing

Step 4: Results

Cooling wise its a wash, but I do gain ~4% range and some cooldown going for the heatgen along the way. Going for coolrun is cooler only by a tiny tiny amount.

So clearly tree 2 is better.

Edited by Kin3ticX, 13 November 2017 - 01:51 PM.

[Growlly]

The result favoring Tree 2 makes sense to me because it always seems easier to reduce heat problems by taking off a hot weapon than adding enough heatsinks to compensate for it.

[Kin3ticX]

Growlly, on 13 November 2017 - 02:13 PM, said:

The result favoring Tree 2 makes sense to me because it always seems easier to reduce heat problems by taking off a hot weapon than adding enough heatsinks to compensate for it.

The HPS is ~double that of the CPS. The heat gens have more bang but they take more points to reach depending on how many points it takes to reach clusters of them. It only takes 17pts to get max coolrun, however, even though its 2% per skill, CPS is also about half that of HPS (in this mech)

So it gets tricky because we are talking

.75% per skill on 8.77 HPS

&

2% per skill on 4.25 CPS

Edited by Kin3ticX, 13 November 2017 - 03:11 PM.

[Tarogato]

Here's the smurfy for the mech: http://mwo.smurfy-ne...909f9848d9f0a0f

Smurfy heat/s: 8.77
Smurfy dissipation: 4.25


Smurfy has the dissipation right, but before heat/sec it assumes that you are firing all weapons as fast as they will fire. Consider that with a loadout like 2x cHLL + 4x cERML you won't fire your cERML until the cHLL is ready to fire again. So instead of adding the heat/sec of both weapons together, you need to add the heat of both weapons, and then divide by the largest cooldown value. For this build, that would give you a corrected alpha heat/sec of 7.83. This better reflects how the loadout is actually used in game. BUT... this is nitpicking. It doesn't actually affect the results at all whatsoever.



Here's the three trees you provided on reddit:

Your first skill tree (5 coolrun, 7 heatgen, max tweak, ECM, minimal survive)
Second skill tree (3 coolrun, 11 heatgen, max tweak, ECM, minimal survive)
Third skill tree (5 coolrun, 11 heatgen, no mobile, ECM, most survive)









First let's talk about the Survive tree you did in the third tree.

Spoiler

(ROI means return on investment, just an arbitrary value where higher is better)
(flat structure is what the bonus actually says. "Structure value" is 90% of that, because most weapons on average deal about ~~10% extra damage to structure. I think. I made that calculation ages ago, maybe it was 9%. Maybe it was 15%. But it's somewhere generally around 10%. _{I think.)}


25 points (your version)
14.4% armour (0.576 ROI)
18.6% flat structure
16.7% structure value (0.670 ROI)
---> 1.2456 total ROI

27 points (extra points)
16.0% armour (0.593 ROI)
18.6% flat structure
16.7% structure value (0.620 ROI)
---> 1.213 total ROI

28 points (route that I often use)
16.0% armour (0.571 ROI)
21.7% flat structure
19.5% structure value (0.698 ROI)
---> 1.269 total ROI

31 points (everything except the one skeletal that you avoided)
16.0% armour (0.516 ROI)
27.9% flat structure
25.1% structure value (0.810 ROI)
---> 1.326 total ROI


(Even when) Correcting for structure only being worth ~90% of its face value (because crits), you find that the more points you invest in survival the better your return on investment. (Except for that 27-point tree, avoid that one, it's bad). All of these trees avoid one skeletal density, but I always pick that one up. I just avoided it because you avoided it, and wanted to stay consistent. =3

ECM tree

Spoiler

I just wouldn't take ECM. It costs 1 ton and 13 points. That's the same cost as one heatsink and 4 cool runs. Is ECM worth that much? I don't think it is, so I don't equip it. Even if you consider that you're already taking cool runs, look at what you can take instead of ECM.

Yes ECM
No ECM ---> 13 points instead go into: extra coolshot, +coolshot effectiveness, +12.4% flat structure, +1.6% armour, 1.5% better heatgen, 0.75% worse cooldown, 3.5% better heat dissipation from the extra heatsink instead of ECM

That's a lot of useful stuff. While ECM is ... uh... what's the point of it? It only makes your mech invisible against noobs that can't aim anyways, and it makes your mech un-lurmable by other noobs. It's basically a noob-proofing crutch that doesn't really do anything against good players. If you constantly play against noobs and you need the crutch, then sure... I guess. But I prefer having real, tangible bonuses instead.

Oh, and before anybody says, "but you're also giving up 30% sensor range and 15% faster target info gathering!" ....... why yes, that's true. But those perks aren't worth anything anyways. Nothing of value lost. Also consider that the two radar derp nodes you're forced to go through to pick up the ECM... are nullified by the ECM itself.

Firepower vs Ops

Spoiler

Pretending that everything above doesn't exist, and going back to your OP. You gave two trees.

42 points = 5 coolrun / 7 heatgen = 0.5625 efficiency
44 points = 3 coolrun / 11 heatgen = 0.5599 efficiency

Now, you must have missed my prime suggestion,



44 points = 4 coolrun / 13 heatgen = 0.6120 efficiency




You give up one duration node, but I think the better heat efficiency is worth it.
Personally, I pretty much never take any duration nodes at all, but that's just comes down to personal preference. I don't notice an improvement when taking them, but other people claim to, and I'm not going to argue it. =D


Now here's the part that will blow your mind. Kind of the secret that a lot of people just don't understand. Weighing the value of coolrun vs heatgen is always the same regardless of what weapons and heatsinks you have. I'll hear people often claim "oh, coolrun is better when have more heatsinks! because you're adding a bonus onto a higher base dissiptation!"

But that's not how this works. You get the same relative bonuses regardless if you're running 10 heatsink or 32.



As you can see, I can put in any random numbers for the "baseline heat stats" and still get the same improvement for all three different skill tree configurations. Your build doesn't matter, your quirks don't matter... all that matters is balancing the %improvement that the node selections grant you.

Edited by Tarogato, 13 November 2017 - 04:33 PM.

[Kin3ticX]

Tarogato, on 13 November 2017 - 04:33 PM, said:

Here's the smurfy for the mech: http://mwo.smurfy-ne...909f9848d9f0a0f

Smurfy heat/s: 8.77
Smurfy dissipation: 4.25


Smurfy has the dissipation right, but before heat/sec it assumes that you are firing all weapons as fast as they will fire. Consider that with a loadout like 2x cHLL + 4x cERML you won't fire your cERML until the cHLL is ready to fire again. So instead of adding the heat/sec of both weapons together, you need to add the heat of both weapons, and then divide by the largest cooldown value. For this build, that would give you a corrected alpha heat/sec of 7.83. This better reflects how the loadout is actually used in game. BUT... this is nitpicking. It doesn't actually affect the results at all whatsoever.



Here's the three trees you provided on reddit:

Your first skill tree (5 coolrun, 7 heatgen, max tweak, ECM, minimal survive)
Second skill tree (3 coolrun, 11 heatgen, max tweak, ECM, minimal survive)
Third skill tree (5 coolrun, 11 heatgen, no mobile, ECM, most survive)









First let's talk about the Survive tree you did in the third tree.

Spoiler

(ROI means return on investment, just an arbitrary value where higher is better)
(flat structure is what the bonus actually says. "Structure value" is 90% of that, because most weapons on average deal about ~~10% extra damage to structure. I think. I made that calculation ages ago, maybe it was 9%. Maybe it was 15%. But it's somewhere generally around 10%. _{I think.)}


25 points (your version)
14.4% armour (0.576 ROI)
18.6% flat structure
16.7% structure value (0.670 ROI)
---> 1.2456 total ROI

27 points (extra points)
16.0% armour (0.593 ROI)
18.6% flat structure
16.7% structure value (0.620 ROI)
---> 1.213 total ROI

28 points (route that I often use)
16.0% armour (0.571 ROI)
21.7% flat structure
19.5% structure value (0.698 ROI)
---> 1.269 total ROI

31 points (everything except the one skeletal that you avoided)
16.0% armour (0.516 ROI)
27.9% flat structure
25.1% structure value (0.810 ROI)
---> 1.326 total ROI


(Even when) Correcting for structure only being worth ~90% of its face value (because crits), you find that the more points you invest in survival the better your return on investment. (Except for that 27-point tree, avoid that one, it's bad). All of these trees avoid one skeletal density, but I always pick that one up. I just avoided it because you avoided it, and wanted to stay consistent. =3

ECM tree

Spoiler

I just wouldn't take ECM. It costs 1 ton and 13 points. That's the same cost as one heatsink and 4 cool runs. Is ECM worth that much? I don't think it is, so I don't equip it. Even if you consider that you're already taking cool runs, look at what you can take instead of ECM.

Yes ECM
No ECM ---> 13 points instead go into: extra coolshot, +coolshot effectiveness, +12.4% flat structure, +1.6% armour, 1.5% better heatgen, 0.75% worse cooldown, 3.5% better heat dissipation from the extra heatsink instead of ECM

That's a lot of useful stuff. While ECM is ... uh... what's the point of it? It only makes your mech invisible against noobs that can't aim anyways, and it makes your mech un-lurmable by other noobs. It's basically a noob-proofing crutch that doesn't really do anything against good players. If you constantly play against noobs and you need the crutch, then sure... I guess. But I prefer having real, tangible bonuses instead.

Oh, and before anybody says, "but you're also giving up 30% sensor range and 15% faster target info gathering!" ....... why yes, that's true. But those perks aren't worth anything anyways. Nothing of value lost. Also consider that the two radar derp nodes you're forced to go through to pick up the ECM... are nullified by the ECM itself.

Firepower vs Ops

Spoiler

Pretending that everything above doesn't exist, and going back to your OP. You gave two trees.

42 points = 5 coolrun / 7 heatgen = 0.5625 efficiency
44 points = 3 coolrun / 11 heatgen = 0.5599 efficiency

Now, you must have missed my prime suggestion,



44 points = 4 coolrun / 13 heatgen = 0.6120 efficiency




You give up one duration node, but I think the better heat efficiency is worth it.
Personally, I pretty much never take any duration nodes at all, but that's just comes down to personal preference. I don't notice an improvement when taking them, but other people claim to, and I'm not going to argue it. =D


Now here's the part that will blow your mind. Kind of the secret that a lot of people just don't understand. Weighing the value of coolrun vs heatgen is always the same regardless of what weapons and heatsinks you have. I'll hear people often claim "oh, coolrun is better when have more heatsinks! because you're adding a bonus onto a higher base dissiptation!"

But that's not how this works. You get the same relative bonuses regardless if you're running 10 heatsink or 32.



As you can see, I can put in any random numbers for the "baseline heat stats" and still get the same improvement for all three different skill tree configurations. Your build doesn't matter, your quirks don't matter... all that matters is balancing the %improvement that the node selections grant you.

Wow, that was a lot of work. Nice

I'm going to have to try going down the left side of the OPs tree like that

(and doing the 3 of 4 burntime)

Edited by Kin3ticX, 13 November 2017 - 05:02 PM.

[Kin3ticX]

thats my first game with Tarogatos build, felt like i was gonna win for sure, then everyone melted.......

[J0anna]

Probably one of the clearest ways to weigh the value of heat gen vs cool run. Nice work.