The test map is a great place for examining the ability to stack abilities as it provides precise metrics. For this experiment we can simply use the damage up ability and see how it affects the damage. Additionally i chose to time the duration of the bubbler, to see how it compares.
Data
main denotes abilities in the main slot of the gear (at the top); sub respectively denotes a sub ability of which there can be up to three per gear.
Splattershot
Raw: 36.0 dmg
1main: 39.2 dmg
1main 1sub: 40.1 dmg
1sub: 37.0 dmg
1sub 1sub: 38.0 dmg
2main: 41.9 dmg
2main 1sub: 42.6 dmg
Aerospray
Raw: 24.5 dmg
1main: 24.9 dmg
2main: 24.9 dmg
Rapid Blaster
Raw: 80.0 dmg
1main: 87.2 dmg
1main 1sub: 89.2 dmg
1sub: 82.3 dmg
1sub 1sub: 84.5 dmg
2main: 93.2 dmg
2main 1sub: 94.7 dmg
E-Litre 3K (charged)
Raw: 180 dmg
1main: 196.3 dmg
1main 1sub: 200.7 dmg
1sub: 185.2 dmg
1sub 1sub: 190.1 dmg
2main: 209.8 dmg
2main 1sub: 213.2 dmg
3main: 220.3 dmg
1main (Head): lacking gear
1main (Clothes): 196.3 dmg
1main (Shoes): 196.3 dmg
1sub (slot 1): 185.2 dmg
1sub (slot 2): lacking gear
1sub (slot 3): 185.2 dmg
Bubbler
Due to the audio measured for the raw loadout being close to 5 s, it will be assumed that this is actually the default length. More precision could possibly be achieved using visual measurement via high speed video recording.
Raw: 5.00 s
1main: 5.555 s
1main 1sub: 5.749 s
1main 2sub: 5.874 s
1main 2sub 1sub: 6.004 s
1sub: 5.151 s
1sub 1sub: 5.324 s
2sub 1sub: 5.509 s
2sub 1sub 1sub: 5.656 s
2sub: 5.323 s
2main: 6.055 s
2main 1sub: 6.182 s
Swim Speed Up
Roughly measured time swimming from one wall of firing range to the other. (No Ninja Squid ability has been used which is said to reduce speed.)
Raw: ~5.0s
1main: ~4.6s
2main: ~4.4s
3main 1sub: ~4.25s
(Measured this because it was suggested that Swim Speed Up does not stack.)
Results
Now for some observations and calculations based on that data.
Weapons
The stacking appears to work the same for the different guns, with one exceptional behaviour: Capping.
Damage Up stacking caps for certain weapons at some point. This is believed to be due to the amount of shots being needed to splat someone with that weapon being made constant. So if a weapon takes three shots to splat someone without any abilities then it will also splat at three shots with 3main abilities. Chargers are not affected by this. (Likely because in the charged state they all kill at one shot.)
This effect can be seen with the Aerospray already when applying 2mains. The damage caps at 24.9 => 24.9 * 4 = 99.6, thus five shots are required. If 0.1 more damage where done that number would go down to four.
The larger damage counts yield more precise results so calculations are done based on the E-Litre 3K:
Raw to 1main: +9.06%
1main to 2main: +6.88%
2main to 3main: +5.00%
Raw to 2main: +16.56%
Raw to 3main: +22.39%
1main to 1main 1sub: +2.24%
2main to 2main 1sub: +1.62%
Raw to 1sub: +2.89%
1sub to 1sub 1sub: +2.65%
Raw to 1sub 1sub: +5,61%
(These are still fairly inaccurate, to control for rounding errors one could calculate the ratios for all different damage profiles and average them.)
Bubbler
Raw to 1main: +11.1%
1main to 2main: +9.00%
Raw to 1sub: +3.02%
1sub to 1sub 1sub: +3.72% | What is going on here...
1sub 1sub to 2subs: ~0% (When chalking up difference as measurement error)
When plotting the times for the number of subs (0 to 4), treating them all as equal, we get this:
Weapons vs. Bubbler
Falloff (Raw -> 1sub) to (1sub -> 1sub 1sub):
Is this even possible? My hypothesis is, that the initial duration simply is not exactly 5 but something smaller. Looking at the bubbler plot the first point also seems out of line and looks like it should be lower. Also subsequent measurements were all affected by the initial assumption, which may further change the data.
This is why the damage numbers are better: They only contain rounding errors, not measurement errors.
Discussion
The count relative results from the bubbler trials are ignored for now, as the trials appear to have been faulty.
The results seem to suggest that stacking is in fact quite real, and that the falloff becomes fairly steep once the already applied bonus is high. For example, when already being equipped with one damage+ on main, an additional damage+ main will only be 75.9% effective. When adding a sub to another sub the additional sub is still 91.7% effective.
The penalty appears to go down again, as a third damage+ on main is still 55.2% effective. If it were still increasing one would expect a value below 51.8%.
If the slot position of a sub ability matters, then it is quite unintuitive, as at least between slot 1 and 3 there is no difference. The sub slots being the same was to be expected; while the exact values of falloff and absolute gains may not be public, it would be odd to have such a mechanism. Likewise, the type of the gear does not appear to matter either.
The results also show that weapon damage caps at the amount that still yields the same number of shots on one target necessary to splat. This is probably a mechanism very specific to Damage Up. Other abilities might have their own peculiarities, for example i doubt that Ninja Squid stacking has any effect (though it would be interesting if it lowered the swimming noise or made the trails disappear).
The two bubbler data points for 1sub 1sub and 2sub seems to suggest, that having abilities spread over different gear does not affect their bonus.
I will investigate this further once i acquire more gear with damage+. With more data points it might be possible to calculate a falloff curve.
Other things that need to be investigated:
Notes
Data
main denotes abilities in the main slot of the gear (at the top); sub respectively denotes a sub ability of which there can be up to three per gear.
Splattershot
Raw: 36.0 dmg
1main: 39.2 dmg
1main 1sub: 40.1 dmg
1sub: 37.0 dmg
1sub 1sub: 38.0 dmg
2main: 41.9 dmg
2main 1sub: 42.6 dmg
Aerospray
Raw: 24.5 dmg
1main: 24.9 dmg
2main: 24.9 dmg
Rapid Blaster
Raw: 80.0 dmg
1main: 87.2 dmg
1main 1sub: 89.2 dmg
1sub: 82.3 dmg
1sub 1sub: 84.5 dmg
2main: 93.2 dmg
2main 1sub: 94.7 dmg
E-Litre 3K (charged)
Raw: 180 dmg
1main: 196.3 dmg
1main 1sub: 200.7 dmg
1sub: 185.2 dmg
1sub 1sub: 190.1 dmg
2main: 209.8 dmg
2main 1sub: 213.2 dmg
3main: 220.3 dmg
1main (Head): lacking gear
1main (Clothes): 196.3 dmg
1main (Shoes): 196.3 dmg
1sub (slot 1): 185.2 dmg
1sub (slot 2): lacking gear
1sub (slot 3): 185.2 dmg
Bubbler
Due to the audio measured for the raw loadout being close to 5 s, it will be assumed that this is actually the default length. More precision could possibly be achieved using visual measurement via high speed video recording.
Raw: 5.00 s
1main: 5.555 s
1main 1sub: 5.749 s
1main 2sub: 5.874 s
1main 2sub 1sub: 6.004 s
1sub: 5.151 s
1sub 1sub: 5.324 s
2sub 1sub: 5.509 s
2sub 1sub 1sub: 5.656 s
2sub: 5.323 s
2main: 6.055 s
2main 1sub: 6.182 s
Swim Speed Up
Roughly measured time swimming from one wall of firing range to the other. (No Ninja Squid ability has been used which is said to reduce speed.)
Raw: ~5.0s
1main: ~4.6s
2main: ~4.4s
3main 1sub: ~4.25s
(Measured this because it was suggested that Swim Speed Up does not stack.)
Results
Now for some observations and calculations based on that data.
Weapons
The stacking appears to work the same for the different guns, with one exceptional behaviour: Capping.
Damage Up stacking caps for certain weapons at some point. This is believed to be due to the amount of shots being needed to splat someone with that weapon being made constant. So if a weapon takes three shots to splat someone without any abilities then it will also splat at three shots with 3main abilities. Chargers are not affected by this. (Likely because in the charged state they all kill at one shot.)
This effect can be seen with the Aerospray already when applying 2mains. The damage caps at 24.9 => 24.9 * 4 = 99.6, thus five shots are required. If 0.1 more damage where done that number would go down to four.
The larger damage counts yield more precise results so calculations are done based on the E-Litre 3K:
Raw to 1main: +9.06%
1main to 2main: +6.88%
2main to 3main: +5.00%
Raw to 2main: +16.56%
Raw to 3main: +22.39%
1main to 1main 1sub: +2.24%
2main to 2main 1sub: +1.62%
Raw to 1sub: +2.89%
1sub to 1sub 1sub: +2.65%
Raw to 1sub 1sub: +5,61%
(These are still fairly inaccurate, to control for rounding errors one could calculate the ratios for all different damage profiles and average them.)
Bubbler
Raw to 1main: +11.1%
1main to 2main: +9.00%
Raw to 1sub: +3.02%
1sub to 1sub 1sub: +3.72% | What is going on here...
1sub 1sub to 2subs: ~0% (When chalking up difference as measurement error)
When plotting the times for the number of subs (0 to 4), treating them all as equal, we get this:
Weapons vs. Bubbler
Falloff (Raw -> 1sub) to (1sub -> 1sub 1sub):
Weapon: 91.62%
Bubbler: 123.2%
Bubbler: 123.2%
Is this even possible? My hypothesis is, that the initial duration simply is not exactly 5 but something smaller. Looking at the bubbler plot the first point also seems out of line and looks like it should be lower. Also subsequent measurements were all affected by the initial assumption, which may further change the data.
This is why the damage numbers are better: They only contain rounding errors, not measurement errors.
Discussion
The count relative results from the bubbler trials are ignored for now, as the trials appear to have been faulty.
The results seem to suggest that stacking is in fact quite real, and that the falloff becomes fairly steep once the already applied bonus is high. For example, when already being equipped with one damage+ on main, an additional damage+ main will only be 75.9% effective. When adding a sub to another sub the additional sub is still 91.7% effective.
The penalty appears to go down again, as a third damage+ on main is still 55.2% effective. If it were still increasing one would expect a value below 51.8%.
If the slot position of a sub ability matters, then it is quite unintuitive, as at least between slot 1 and 3 there is no difference. The sub slots being the same was to be expected; while the exact values of falloff and absolute gains may not be public, it would be odd to have such a mechanism. Likewise, the type of the gear does not appear to matter either.
The results also show that weapon damage caps at the amount that still yields the same number of shots on one target necessary to splat. This is probably a mechanism very specific to Damage Up. Other abilities might have their own peculiarities, for example i doubt that Ninja Squid stacking has any effect (though it would be interesting if it lowered the swimming noise or made the trails disappear).
The two bubbler data points for 1sub 1sub and 2sub seems to suggest, that having abilities spread over different gear does not affect their bonus.
I will investigate this further once i acquire more gear with damage+. With more data points it might be possible to calculate a falloff curve.
Other things that need to be investigated:
- Is the falloff the same for other abilities (might be hard to measure for other abilities though)
Notes
- Damage capping does not render the ability useless! The other direction is not capped, so if an enemy equips defense abilities the needed amount of shots can go up. Also, when in a firefight with your team you can combine your added damage with theirs, e.g. a .96 Gal does 62 damage, a Splattershot 36, so if two people equip these weapons and hit an enemy once each they do 98 damage, now if the .96 Gal user is equipping 1main damage+ it does 67.6 damage; the total now is 103.6, which is one splatted enemy.