1. Set the heat capacity of ALL mechs regardless of heatsink quantity or type to 30
2. Increase dissapation of all DHS to 2.0
3. Remove alpha heat scale penalties
Reasoning:
While David made a rational arguement in the last Ask the Devs for the current heat capacity calculation I beleive that PGI have made a fundamental error. To say that a mech with 30 heatsinks firing a 30pt heat alpha in TT never increases in heat is not true. It does - it is just that at the end of the turn (10s) it is back to where it started so the net gain is 0. The capacity of ANY mech in TT to exceed their cooling is the same at 30 heat each. Regardles if it is a Jenner with 10SHS or an Awesome with 22 DHS. This is also tied to the fact that weapons fire a lot faster in MWO than TT - i.e. PPC every 4s rather than 10s. By setting the heat capacity of all mechs to 30 it automatically limits heat dependent high alpha builds (ballistics high alpha builds are weight limited). Note that with the current master level pilot skills in play this cap raises to 36.
Increasing the dissapation of DHS external to the engine to 0.2hps will help hot, heatsink dependent mechs overcome the reduced cap. It also makes bringing more heatsinks important for sustained combat. It also removes the penalty of mechs who are unable to fit a 250 engine. Finally it is simpler to understand and makes a lot more sense.
With the drastic reduction in heat capacity the possiblity of 4/6ppc alpha strikes ceases to be a threat. So there is no real need for the alpha heat scale penalty system. This removes a massive amount of unnecessary complexity from the game, again making things simple and intuitive oncce more.
Methodology:
In order to develop this proposition I sat down and constructed a spreadsheet to calculate the heat position of any given mech firing alpha-strikes over time for both the currently implemented method (including heat scale penalties and the impending PPC heat increase) and my proposed method (with changes mentioned above). I was also able to easily calculate the total alpha damage over time as an additional metric. I then plotted these as graphs to make it easy to visualise.
I selected a number of mechs to look at, full list is below but I can easily add more to this. Note that due to somewhat cumbersome method of calculation I can only work with alphas that have the same cycle time on all weapons. Full list of mechs I considered are:
6 LL Stalker (20 DHS) - silly hot alpha build
6 PPC Stalker (17 DHS) - ridiculous hot alpha build
3 PPC Awesome (22 DHS) - sensible hot alpha build
ERPPC + 2 PPC + Gauss Highlander (14 DHS) - recently nerfed popular alpha build
9 ML Swayback (18 DHS) - recently nerfed alpha build
6 ML Jenner (14 DHS) - core out in Atlas in 4 seconds!
3 UAC Ilya (10 DHS) - cool sustained DPS ballistics
2 Gauss 2 ML Jagermech (10 DHS) - cool high alpha sustained DPS ballistics
2 AC20 Jagermech (10 DHS) - cool high alpha sustained DPS ballistics
AC20 + 2 LL + 3SRM Atlas (17 DHS) - high heat high alpha brawler
2 ERPPC + Gauss Highlander (17 DHS) - popular alpha build
4 LL Stalker (24 DHS) - sensible hot alpha build
Note I did my calculations working with the pilot skills at master level, so heat dissaption of 1.2 and heat capacity of 1.2 (making it 36 for all mechs), also 0.95 reduction in cycle time for Fast Fire. But I also built in the option to easily adjust these back to new pilot status of 1.0 - which makes a massive difference btw (more on that later).
Results:
Looking at the results it is very hard to compare the different mechs directly with each other as the different weapons systems obviously work very differently. It is more interesting to look at the differences between the two systems within a specific mech, notably a mech considered to be quite good (or bad) in the current system. Also when looking at the graphs it is probably wise to remember that in most typical sniping duels you are only firing a single alpha, while in mid-range engagements you may only have 5s of sustained fire, and even most brawls only last 20-30s before somebody goes down. In other words read with a grain of salt as spreadsheets only give you a fraction of the story.
The 6 LL Stalker actually still works with the existing heat system (just), but under my proposed system in overheats and shuts down after a single alpha (42 heat vs 36 capacity).
The 6 PPC completely overheats and explodes in both systems as you would expect.
The 3 PPC Awesome is interesting so here is the heat graph:

Under the existing system the 3 PPC Awesome (or indeed any heavy/assault with 3 PPCs) is mildly nerfed by the heat scale alpha penalty - but it still chucks out 3 alphas in less than 10s. The revised system actually slows this alpha down slightly, but then gives much superior sustained DPS. BUT it becomes VERY easy to shut yourself down as you can see that you need to wait for your heat to drop to about 25% before you can safely alpha. Chainfire would be recommended.
Here is the Damag vs Time plot for interests sake:

Ok another heavy hitter that has been nerfed by the heatscale system is the Highlander with ERPPC + 2 PPC + Gauss. This is interesting to compare with the 3 PPC boat as the Gauss sucks up a lot of tonnage that could otherwise be used for cooling.

Quite similar to the previous 3 PPC mech the proposed changes would slow down the second alpha strike, with increased dissapation allowing for an eventual increase in total damage. Note that this graph is slightly misrepresentative as the Gauss rifle could keep firing independently of the PPCs as it is pretty much free from the heat limitations. BUT that is the entire point of the exercise - breaking up those initial big alpha strikes. Again VERY easy to overcook in this as you need to alpha from almost 0 heat.
Next up I wanted to show you the current meta favourite - the 2 ERPC + Gauss in either a Highlander/Cataphract:

Well, doesn't that paint a picture! As you can clearly see, without any heat scale penalty this badboy can alpha 5 full times before it has to even worry about overheating under the current system! The proposed system would cut that to two, and slow down the subsequent alphas while also forcing the pilot to actually watch his heatmeter! Interestingly the damage over time is actually pretty comparable here out to a minute thanks to the massive heat capacity it benefits from under the current system.
Also note that total DoT is similar to the 3 PPC build - the extra weight of the gauss limiting HS dissapation is telling. Under the current system this loadout is hands down better than 3 PPCs in all ways - bigger alpha, better sustained damage - which limits choice when building competitive mechs.

Ok next up check out some heat neutral (or close to it) ballistic builds:

As this only has 10 DHS in the engine the dissapation is actually identical. So the parameters of a dakka build like this remain unaffected by the proposed system. It is also worth pointing out that this build does a ridiculous 870 damage from 60s of continous firing. So we already have high DPS, heat neutral builds in the game - yet they aren't breaking it.
The double Gauss is basically the same - but trades overall DPS for alpha potential.
What is problematic however is the double AC20 jager-boom:

The changes to alpha heat scale have reigned this monster in, but barely. You can still fire 3 full alphas for 120 pinpoint damage before you have any chance of overheating. The proposed system doesn't really help either - in fact it would make the JM close to heat neutral. But that raises a key point. AC20s are ballistics - they are not meant to be heat limited. This demonstrates that the heat system is only one aspect of the problem PGI face. There needs to be some type of adjustment to pinpoint damage to solve this underlying issue. Trying to use the alpha heatscale to control AC20s is really a square peg in a round hole - and doesn't actually work that well as you can see here.
The 6ML Jenner and 9ML Swayback both came out reasonably close to where they already are but again with slightly less frontend and better sustained DPS. Chainfire would be necessary on the 9ML swayback as its alpha heat (36) would hit capacity (36).
The brawler Atlas loses an alpha at T=4s due to lack of heat (but it is a monster alpha of 74 damage!) so overall damage is less. A slightly more sensible build with 2 SRM6 and extra DHS would probably run overall much nicer.
The 4LL Stalker is viable under both systems - sustained damage is up overall (shows a similar trend to the 3PPC build). This is fair enough as 4LL seems pretty sensible and runing that close to heat neutral probably isn't going to break the game.
Conclusions:
This proposed heatscale would certainly limit large, high heat alphas - and would simplify the very complex ruleset that PGI are contructing to manage heat. I also think it would make adding extra doubles in mechlab a tough decision at the expense of endosteel (for more weight) or a larger engine. Forcing players to chainfire or risk overheating would be a big change to the current 'alpha alpha alpha' gameplay - and whether the gameplay would be more enjoyable or not is certainly debatable. But alternatives do exist and I am a little dissapointed that PGI are being so incredibly conservative with their test server changes.
Unfortunately there is no clear heat-based solution to the problem of the double-jager-boom mech, other than the current, slightly awkward, heat scale penalty. It seems to me that there needs to be supplemental changes to address the issue of pinpoint alpha damage at a fundamental level, which would leave the heat system free to function as it should.
Finally, having embarked on this task it has become more clear to me the enormity and complexity of the task facing PGI. So I will probably be a bit more easy on them. BUT I still think they need to try a lot more different ideas, and learn to admit when they have made a mistake (it's not too late for ECM to change!).
btw if anybody knows of a good free file hosting servie I can chuck up a link to my spreadsheet and you can play around with your own values for capacity/dissapation etc. or add new mechs to consider.
Edited by Jungle Rhino, 31 July 2013 - 06:59 AM.