>What needs to be Adjusted?
Im only going to be addressing this point in this post, rather than focusing on set specific tweaks like what power sets might be over-tuned or under-tuned, I think that first some systemic issues need to be resolved mostly with the way DR was implemented for the majority of archetypes, as I will explain below.
Problem 1: Resistance buffs dont work.
Or to be more precise about it, they dont work when applied to Blasters, Controllers, Defenders, Corruptors, Dominators, Masterminds or Sentinels. They do work when applied to other archetypes, and that is due to the different diminishing return curve values that those archetypes get applied to them. To illustrate what I mean here is the graph showing the ratio of pre-DR and post-DR resistance on a Defender, and here is another graph showing the same ratio on a Scrapper.
Its hard to get a feel for what this does without ingame pvp experience but basically it caps “squishy” archetypes to about 43% - 48% resistance. You may be wondering, why is this a problem at all? After all they are supposed to be squishy! Well for a start, as things stand right now every new player who tries to get into pvp must undergo the learning process that yes it is a good idea to get a resistance toggle, but no it isnt a good idea to try and get any more resistance beyond a certain point, as the returns fall off to almost zero at a certain arbitrary point which is impossible to know about unless you know the DR curve, but that's just an aside. The primary reason is that there are sets revolving around buffing of resistance namely sonic and thermal and right now those +resistance buffs pretty much dont do anything.
So to fix this problem, the DR curves need to be manipulated in such a way as to make squishies as close to as they are now with their own res toggle active, but still able to get more resistance if for example a sonic were to buff them. In other words stacking their epic resistance toggle, a ton of +resistance set bonuses, and tough all together should not result in a defender beginning to approach scrapper levels of surviveability, while at the same time the +resistance from a sonic buff still needs to give a reasonable boost in effective health. Striking this balance has been tricky but after a lot of messing around with the values, these are the results I suggest:
Old A & B
New A & B
Old Free Res
New Free Res
1.20 & 1.00
1.60 & 0.75
1.20 & 1.00
1.40 & 0.75
PB & WS
0.90 & 1.00
1.20 & 0.75
Bane & Widow
0.90 & 1.00
1.20 & 0.75
1.80 & 1.00
2.80 & 0.75
0.80 & 1.00
0.85 & 0.75
0.60 & 1.00
0.85 & 0.75
Explaining why these numbers:
Currently, the resistance for a squishy before and after DR looks something like this.
In the link above, the “R” value shows pre-diminished resistance. The reason it is set to 0.835 is that with the current values of 40% free resistance and between 30% and 43.5% for epic toggles, squishies can get up to around 83.5% resistance before DR.* As you can see, with “A” and “B” values of 1.2 and 1.0, this diminishes down to 41.7% (shown by the result: 0.416969)
*Yes 83.5% is the scenario with the highest res toggles, but it doesn’t factor +res set bonuses so this balances out.
Now let us apply some theoretical Defender Sonic Resonance buffs to this squishy. It will look something like this.
The “R” value is increased from 0.85 to 1.152 because sonic resonance buffs give 31.7% with 3 resistance IO’s in them. Looking at the result, you can see that the squishies resistance has gone from 41.7% previously to 45.9% now. So how effective was the sonic buff?
Assuming a squishy has 1600 health and 41.7% resistance, they have 2744 effective health. (1600/(1-0.417)) With the sonic buff active they have 45.9% resistance so they have an effective health of 2957 (1600/(1-0.459)) Comparing the effective health before and after we can see that the sonic buff increased it by about 7.7%. (2957/2744) When I said previously that res buffs pretty much don't do anything, this ~7.7% EHP buff is what I was referring to.
Now we take the changes from the above table and reapply the math. So the first situation now looks like this.
The new value of “R” of 0.635 is because the free resistance has been dropped from 40% to 20%. The new values of “A” and “B” are 1.4 and 0.75, which brings the amount of resistance after DR to 41.5%, so more or less identical to what it was before.
Applying the Defender Sonic Resonance buffs under the new curve gives this result.
Now we compare the effective health again. The initial effective health this time is 2735. (1600/(1-0.415)) Adding the sonic buff increases the effective health to 3408 (1600/(1-0.5305)) for an increase in effective health of 24.6%. (3408/2735)
So why these numbers? The value for “A” defines how aggressively the curve seeks out the ultimate determinant of where the plateau is, “B”. Smaller values for A and B are less aggressive, as in they DR less. Looking at the table above you can see that generally A has been made higher and B lower, an example of this effect shown here:
On the left is the curve produced by the squishy values currently in the game, 1.2 and 1.0. On the right is the curve produced by the suggested values, 1.4 and 0.75. As you can see the curve on the right does steadily become more aggressive, but does not plateau completely like the current curve does. This is because by setting “B” at 0.75 the maximum amount your resistance can diminish by is now 75%, meaning that once you have reached the final slope of the curve, you get one point of resistance in pvp for every four points you accrue before DR. The values for “A” are designed to counteract the changes in “B” up until a certain point, to maintain the current resistance values that an AT can achieve on their own without buffs from others, it is this fact that is key to the successful implementation of these curve changes.
Note: While changes are made to the DR curves of non-squishy AT's such as scrappers and tankers, these changes are made for the consistency of the “B” value. For Brutes, Scrappers and Stalkers this pretty much changes nothing from a balance perspective, to see this for yourself, if you plot the old curve against the new one in these instances, they pretty much overlap. For Epic AT's I have been a slight bit more generous due to how badly all of these AT's underperform, or at least I claim they do, I hope you can get some testimony from other members of the pvp community to back me up on that. For sentinels in order to keep them in line with their current curve there was not enough free resistance to subtract, so even though they currently share an "A" value with squishies a more harsh one was needed to keep the curve the same when balancing around the new "B" value.
Problem 2: Defense buffs dont work.
The same design philosophy that went into designating the resistance DR curves seemingly was also implemented into defense DR curves, this time to an even greater extent. Here is the ratio of Pre-DR and post-DR defense on a squishy, and here it is on a scrapper. This causes an effective post-DR cap of 20% for squishies, and 60% for melee's. What's absolutely crazy about this change is that in addition to sharing the problems about learning the system for newcomers, as well as neutering sets that give defense buffs such as Cold Domination, and Force Field, is that Sentinels are also lumped in with the squishies on the 3.00 value for “A”, meaning that for example a Super-Reflexes Sentinel will have his defense diminished to about 19% normally, and then when using their T9 power Elude, that will increase their defense to... 20%.
Similarly to my methodolgy in tweaking the resistance curves, the DR curves need to be manipulated so that the self-defense buffs that squishies can grant to themselves cant be allowed to get out of hand, but the defense buffs that other players grant to them still needs to do something. Since elusivity also exists in pvp, some values need to be tweaked here to compensate for the more generous defense curves.
Old A & B
New A & B
3.00 & 1.00
1.60 & 1.00
3.00 & 1.00
1.30 & 1.00
1.50 & 1.00
1.30 & 1.00
0.99 & 1.00
0.99 & 1.00
Force Field, Deflection Shield: +5%
Force Field, Insulation Shield: +5%
Force Field, Dispersion Bubble: +5%
Cold Domination, Ice Shield: +5%
Cold Domination, Glacial Shield: +5%
Tank, Brute & Scrapper’s Super Reflexes, Elude: +20%
Scrapper’s Ninjitsu, Kuji-In Retsu: +20%
Force Field, Personal Force Field: -15%
Force Mastery, Personal Force Field: -15%
Mace Mastery, Personal Force Field: -15%
Mace Mastery, Scorpion Shield: -5%
Cold Mastery, Frozen Armor: -5%
Stalker & Sentinel’s Energy Aura, Overload: -20%
Elusivity is to be applied to the move/damage types that the defense applies to, for example on Deflection Shield the elusivity is granted to Smashing, Lethal and Melee. As an aside, a glaring inconsistency in the game right now is that the tier 9 power Overload power grants +20% elusivity whilst the other Tier-9 defense powers Elude and Kuji-In Retsu grant none whatsoever. The above proposal changes that divide so that all defense set Tier-9 powers should now grant elusivity when used by Tanks Brutes and Scrappers, but not to Stalkers and Sentinels.
Explaining why these numbers:
Unlike resistance, the value of defense is relative. Due to this, it is necessary for this explanation to draw up some typical values for a would-be attacker. The values I have decided to use for the hypothetical attacker are a bonus accuracy of +125%* and a Tohit of +15%**. In order to demonstrate what the changes above do, this “typical attacker” will be pitted against a blaster with Frozen Armor.
* +75% from enhancements and +50% from set bonuses.
** From tactics.
First the stats of both the attacker and the attacked need to be made subject to DR:
The +125% accuracy of the attacker diminishes to +100%. (Link)
The +15% to-hit of the attacker diminishes to +12% (Link)
The +16.84% defense of the Blaster diminishes to +11.8% (Link)
Our first result looks like this, meaning Frozen Armor was giving the Blaster a 10% chance to evade the attacker. Assuming the Blaster had 1847 Health, this means their effective Health was 2932 assuming 30% resistance. ((1847/0.9)/0.7)
Now we apply Deflection and Insulation Shield to our Blaster. Under the current system, applying these buffs on top of Frozen Armor diminishes from 40.57% down to 17.8% as shown here. Now we run our typical attacker situation run again with this new defense total, which gives this result. The Blaster now has a 20.4% chance of evading the same attacker, giving an effective health of 3314.7. ((1847/0.796)/0.7) This is an effective health increase of 13% (3314.7/2932).
Now we make all the same calculations under the new system. The typical attacker remains the same so the first thing to check is the new degree to which Frozen Armor’s defense diminishes, which looks like this. Knowing that, we can run the typical attacker calculation again, giving this result. As you see the Blaster’s chance of evading when running Frozen Armor alone is now around 9%.* Making their effective health 2893.
*Yes, this does mean Frozen Armor becomes weaker than currently, however please bear in mind it is less punishing under my system to stack additional defense buffs such as set bonuses, Stealth, Maneuvers, etc.
Applying Deflection and Insulation Shields again under the new system, these buffs on top of frozen armor now diminish from 40.57% down to 25.7% as shown here. Running our typical attacker calculation one final time, we see that with the forcefield shields on, the Blaster now has a 34.6% chance of evading, as shown here. This gives an effective health of 4038 ((1847/0.626)/0.7) Which is an effective health increase of 39.5% (4038/2893).
The reason why I bring sonic resonance buffs up to 24.6% effective health increase but show here against the “typical attacker” that the forcefield buffs give a much larger 39.5% effective health increase is because of the relativistic nature of defense. To demonstrate what I mean, if the “typical attacker” uses Aim the hit chance goes right back up to the clamped cap of 95% as shown here.
Problem 3: Mez
Prior to issue 13 mez durations mirrored those in pve, and followed a protection based system. This was subsequently changed in issue 13 to a new system in which protection is converted to resistance in pvp, and mez durations were shortened at the same time, after which you are given a short period of Mez Immunity. This new system has proven itself to be quite unpopular, the reason being that now there is no definitive answer to being mezzed and since spikes only take about 2 seconds anyway that is all the duration it takes for you to be incapacitated for in order for you to die. Under the old protection based mez system however, you can simply pop a Break-Free to escape the hold and potentially survive the spike. For that reason I think a lot of players would like to shift back to something like the old system.
With all that said though, there are some issues to going back to exactly the way mez worked before i13. The reasons for that are, firstly in issue 13 all attacks were given bonus damage. The relative increase of this bonus damage was not equal across all AT’s. For example a Blaster’s Flares increases from 63.19 damage in pve, to 110.10 damage in pvp, this is an increase of around +74%. In comparison however a Dominator’s Flares increases from 46.37 in pve to 105.5 in pvp, for an increase of around +128%. Secondly, with the advent of the PPM system, proc damage makes up a significant portion of the total damage that players deal to eacother, this additional flat damage makes the total damage potential of each archetype closer in relative terms.
The reason this matters is that since the relative increase in damage was in general higher for dominators and controllers, therefore reverting their mez durations back to pve values would overtune them. With that in mind I have been considering and revising a system over the last few months which I believe could strike the needed balance, that system is as follows:
It is protection based, if the magnitude exceeds your protection you are mezzed.
Mez immunity after being mezzed is removed.
Current magnitudes are kept. (Including mag 4’s).
Current pvp durations are kept.
Resistance toggles now give protection. (Rather they simply mirror what they do in pve).
The following hard Mez types now take an order of severity:
Caveat #1: After being mezzed by a mez type listed above for 4 seconds, the affected player is granted a protection buff to that mez and all other listed mez types which are considered to have a lower severity, this starts at 2 and increases by 2 for each second that the player remains mezzed up to a maximum of 10. (reached at 8 seconds mezzed). Similarly, for each second that a player is not mezzed this protection buff then decreases at the same rate. With this system players have a means to escape but it is also theoretically possible to Mez someone permanently if you can stack enough of it by yourself.*
*Due to Caveat #2.
Caveat #2: Mez stacks from multiple sources are taken separately and compared. The highest stack of mez is applied and the other stacks are ignored. This means if only one person is stacking mez on you then things go as normal. If another person then comes along and stacks mez on you as well, then either his extra stack of mez will be ignored if it is weaker, or take over as the mez being applied to you if it is stronger. This is done separately for each mez type.
The following Blaster and Sentinel primary set mezzes are limited to mag 2:
Bitter Freeze Ray
All Sentinel protection toggles have values reduced from 8.3 in pve to 5.0 in pvp, allowing them to take two mag 2 holds or one mag 4 hold without being mezzed.
The Protection values given by destiny Clarion powers are are no longer converted into resistance in PvP, instead they remain as protection but are reduced by a factor of 3. This is to balance around Clarion Core Epiphany, which gives 30 Prot for the first 30 seconds, 9 prot for the first minute, and 6 prot for the full duration in PvE, so would give 10 Prot for the first 30 seconds, 3 prot for the first minute, and 2 prot for the full duration in PvP. In conjunction with Acrobatics this will allow for protection against a single mag 4 hold, but not be sufficient to protect against stacked mag 3 or mag 4 holds.
Explanation of methodolgy:
Caveat #1 is designed to replace the immunity effect that is currently implemented in the game with a much more lenient and flexible system. The first implication of Caveat #1 is that a single mag 2 hold can’t last longer than 4 seconds and a single mag 3 or 4 hold can’t last longer than 5 seconds. This is not an arbitrarily imposed pair of rules but two implications of one system which happens to perfectly hit desirable values. Another implication is that when mezzes are stacked to create higher magnitudes, the maximum duration that can be reached also increases as a consequence. For example if you stack two mag 2 holds, you can now keep your target held for 5 seconds instead of 4. Extrapolating further, if you stack two mag 4 holds you can keep your target held for 7 seconds instead of 5.
To see for yourself how this works observe:
So given this, all that is required to hold someone indefinitely is triple stacking a mag 4 hold to reach a total of 12 magnitude. However, that is against a target without the pool power Acrobatics. So what if they do have that? We will assume that a single target hold from a control set can reach 2.3s recharge time, and 9.5 seconds duration. We also assume that it takes at least 1 second for the hold animation. Given this we will say that it is possible to hold someone once every 3.5 seconds. The 9.5 second duration is also reduced by acrobatics to 6.4 seconds. (9.5/1.4844) Given all these values this is what will happen:
The first hold is applied, it’s magnitude of 4 exceeds the 2 protection of the target. This first hold is due to last for 4 seconds because although its initial duration of 9.5s was reduced to 6.4 by acrobatics, the held player is going to be gaining 2 protection at the 4s mark.
The second hold is then applied 3.5 seconds later. The first hold still has 2.9s of duration left at this point. (6.4 - 3.5) Half a second after that the 4s mark is hit, giving the held player the 4 prot they would have needed to break free of the first hold, however by this point it is too late since a second hold has already been applied.
Five seconds pass after the first hold was applied, the held player now has a protection total of 6, still not enough protection to break out of the double stacked mag 4 holds. One second after that (Six seconds total) the player then reaches a total of 8 protection which is finally enough for them to break out. A further 0.4 seconds after that, the first hold wears off.
What happens after this? Well once the target is no longer mezzed their protection immediately begins to decrease. The result of this is that if the Dominator/Controller keeps stacking their primary hold on them they will be re-held very shortly. So the overall conclusion is that it is possible to lock down a player that is running acrobatics for 6 seconds, and once that player breaks out after those 6 seconds they will still be able to be re-held very shortly afterwards. The beauty of this system though is that the Dominator/Controller has to work constantly in order to make this happen. If they apply their primary hold only once, then the target will only be held for 4 seconds, similarly, if they only apply it twice, then once the player breaks out it will be a longer time before they can be re-held afterwards.
In order to keep the beauty of Caveat #1 from being abused, Caveat #2 makes it so that the benefit of cooperatively stacking mez only goes as far as covering the down time in each-other’s mez stacks. This essentially means that provided the balance in a 1 vs 1 scenario is nailed down properly by the variables set by Caveat #1, this balance will then scale up to as high an X vs X scenario as can be conceived.
Finally, the order of severity is implemented so that if a player is stacking a lot of different types of mezzes on a target, for example both immobilizes and holds, then being affected by the weaker mez will not provide protection to the stronger one under any circumstances.
Problem 4: Knockback requires too much counterbuilding.
Compared to the others, this is a quite simple and easy to solve problem. The issue is just that in order to safeguard oneself against a power like Force Bolt, you must somehow cram into your build about 48 magnitude of knockback protection. Solving this issue through a DR curve seems like a very easy fix, and can be done for example using the following method:
The value for “A” in the DR formula for Knockback Strength (Str) changes:
From A = 0.33 to A = 0.8
This will increase the degree by which the enhancement slotting in powers is diminished by DR, as a result the final enhanced magnitude of those powers will be closer to their base magnitudes.
Currently the highest that a knockback power can be enhanced for is +226.95%, this is when fully slotting the power using 6 level 50 basic crafted knockback IO’s, and then boosting all of them to +5. Using the current value for A in the knockback curve causes this amount of enhancement to diminish as shown here to +152.65%. Under this proposed change, the curve would be more aggressive, and instead the amount would diminish to +103.58% as shown here.
Below is a table listing what values various high magnitude knockback powers reach before and after my changes. The values given are for the 6 level 50 +5 boosted basic crafted IO’s mentioned before. Note that these magnitudes are exactly the amount of protection you must exceed in order to be protected.
Old Post-DR Magnitude
New Post-DR Magnitude
Defender Force Bolt
Defender Power Push
Blaster Power Push
Problem 5: -Jump.
-Jump powers are ubiquitously used in zone pvp, and often banned in arena pvp, grounding a player with these abilities is easy to pull off, very rewarding in restricting their ability to both chase and escape, and due to its whopping -50,000% jump height value, impossible to counter or resist. To change this I propose the following solution which involves both toning down the severity of the values for “A” and “B” in the jump-height DR formula, whilst multiplying the -Jump value by an AT scalar.
From A = 0.15 to A = 0.05
From B = 0.33 to B = 0.15
In PvE all -Jump powers do -50,000% jump height. By multiplying this value by these Archetype scalars used for various debuffs, the following new values emerge:
Post Scalar -Jump
Blasters, Brutes, Scrappers & Tankers*
-50,000% * 0.070 = -3,500%
Dominators, Masterminds & VEATs
-50,000% * 0.075 = -3,750%
Peacebringers & Warshades
-50,000% * 0.090 = -4,500%
Controllers and Corruptors
-50,000% * 0.100 = -5,000%
-50,000% * 0.125 = -6,250%
*Also probably sentinels if this is the scalar they inherited, additionally this scalar can be used for temp powers.
Explaining why these numbers:
The ubiquitously used Super Jump combined with the mandatorily used Hurdle grant from an initial combined +2946.8% jump height when unslotted, up to a combined total of +6007% when triple slotted with +5 boosted Jump IOs. In PvP, thanks to both the enhancement strength and the final total being hit by diminishing returns, this range is currently reduced to between +2112.09% and +3819.01%.* With the changes to the jump-height DR formula this difference between PvE and PvP will be less pronounced, resulting in a range of +2672.57% up to +4851.76%.
*Enhancement DR looks like this, then apply that to base values and DR the resulting total like this.
Since base jump height is 100% this means that the final range is 2773% to 4952%. When comparing with the table above, this shows that slotting will allow you to be able to make medium to small jumps even whilst hit by -Jump powers from Blasters, Brutes, Scrappers, Tankers, Dominators, Masterminds & Epic ATs. On the other hand, those Archetypes with better debuff scaling will exceed 4952% and still force completely land-bound movement. However since the values are still in a reasonable range, adding Internal Reduction into the mix allows players to jump once again.
Problem 6: The Melee and Ranged disparity.
Currently melee Archetypes are in a rather horrible place where their designated role is based around taunting enemies and throwing web grenades at them to apply -jump. The fact that this role has given these archetypes some semblance of a place in pvp does not change the truth that this should not be the only role in which they should be able to function. The thrust of what im getting at is that it is difficult to land successive melee attacks in pvp. While it is accepted that there should be some tradeoff in return for being graced with much more lenient resistance and defense DR curves which grant vastly superior surviveability, currently it is very hard for a melee to land consistent attacks against a kiting ranged character. The changes I advocate here are mild and designed to make it possible to land attacks but only if a player is focused on solely doing just that.
The methodolgy is as follows:
Values for “A” and “B” in the jump-speed and run-speed DR formula for Tankers, Brutes and Scrappers* change:
From A = 0.15 to A = 0.10
From B = 0.55 to B = 0.35
*This change regarding the DR curve values could also affect Black/White Dwarfs if it is possible to implement those values changing on the fly when formes change.
All melee attacks for Tankers, Brutes, Scrappers & Black/White Dwarfs with a range of 7 in pve get a range of 10 in pvp. (If you think this will look weird please see KO Blow with its range of 13.2).
In pvp, all melee attacks for Tankers, Brutes and Scrappers & Black/White Dwarfs apply an effect which increases their running and jumping speed cap for a small duration, similar to the way the powers Afterburner and Takeoff function, but possibly not stacking with those powers.
Explanation of methodolgy:
The first change involving changing the values for the DR formula allows for Tankers Brutes and Scrappers to reach the speed cap easier in pvp. This change alone does not make those AT’s move faster than ranged AT’s since the speed caps are still the same, it only makes it easier to reach said caps. The third change then allows for temporary increases beyond the speed cap after using a melee attack.
This means that if a scrapper is using either taunt, webnade, or even their epic ranged attacks, they will not move any faster than a ranged AT, although they will be able to hit the speed cap with less investment in their build. However if that scrapper then decides to try and land a melee attack, that will then give them an increase in speed which can help them to land subsequent melee attacks. Since the range on all their melee attacks goes from 7 to 9 this initial attack is also easier to land.
The upshot of these three changes is that it will be much easier for melee AT’s to land their melee attacks, as well as chase down enemies that ranged AT’s could not. These two factors will make melee AT’s more threatening and consistent in their damage output without having to rely so heavily on ranged attacks from their epic pool to do so. This also means that melee AT’s will no longer have to rely so heavily on their current gimmick of taunting to be relevant, but at the same time will not significantly buff a melee character designed with taunting in mind.
Note: It may be tempting to consider potential alternatives to this proposal such as having melee attacks apply a slow to enemies or something like that. While such an effect may appear at first glance to achieve the same results, the key difference is that slowing an enemy down makes it easier for all your ranged allies to catch up to that enemy again rather than just you. I do not reccomend this.
Additional Misc changes i'd like to see:
Show Post DR stats in combat attributes for all stats not just some.
Remove the “Area Factor” from pvp AOEs and Cones. The idea that an area of effect on a power would yield additional value from that ability was just a complete lack of understanding on the initial developer's part. Usually people are around 50 to 350 feet from eachother at all times, and usually closer to the 350 side at that! A radius of 15 or 20 doesnt change a power's expected number of targets hit in pvp, it always just hits one.* *Excluding very specific circumstances in which people gather for certain high value buffs briefly and then disperse.
Fix the Clamp, raising the hit chance limit from 95% to 100%.
Raise the proc chance limit from 90% to 100%.
Reduce the level of most NPC's in zone's by about 5 to make them con grey, I expect this change wont make it through the less hardcore crowd but hey, we're the ones in recluses victory all day and as far pvp is concerned the arachnos and longbow are just there occasionally messing up our fights and not much else. They also fire PvE length mezzes at you as you jump overhead, which lasts far longer than anything a player can apply.
No longer halve the values of the Core side of the Support Hybrid incarnate tree in pvp. Since other hybrids such as assault and melee are more competitive and the Radial version of support with its larger 80ft radius does not suffer this pvp penalty.
Suppress offensive toggles, for a duration following the formula: B/(1+R) + A + M
B = The Base Recharge of the offensive toggle.
R = Recharge bonuses your character has.
(eg 90% from enhancements, 80% from sets & 70% from hasten = 0.9+0.8+0.7 so 2.4)
A = The Animation time of the offensive toggle.
M = How long the mez that caused the suppression would last for.
What this formula does is estimate the time it would have taken for a user to put the toggle back on manually, then suppress the toggle for that duration. So in the example linked the user is mezzed for 2 seconds, their Radiation Infection takes 2.35 seconds to recharge and has a 1.5 second animation time, so the total time that it would take for them to reapply Radiation Infection is 5.85 seconds, if they played perfectly. So that is the duration that they would be suppressed for under this formula.
And that's my thoughts, im sorry for coming off as a bit pretentious telling you exactly down to the very numbers what I think should be done, but I did put a lot of work and fiddling about trying to get the best values for the curves and have revised my ideas several times in private, I hope you can find some use of them. Thanks, if you read this monster, for reading!