Explosion falloff not related to radius

[squeek502]

See:

• fortressforever/game_shared/ff/ff_gamerules.cpp

  Line 1286 in f663579

  falloff = 0.5f; // AfterShock: need to change this if you want to have a radius over 2x the damage

• fortressforever/game_shared/ff/ff_gamerules.cpp

  Line 1418 in f663579

  flAdjustedDamage = flBaseDamage - (flDistance * falloff); // AfterShock: this means if a player is on the radius of 2x the base damage, you'll do 0 damage

This makes falloff unreliable, and makes things that have an altered damage:radius ratio (the IC and frag grens most notably) either do zero damage before getting to the edge of its radius (in the case of the IC) or do more than 0.5x damage at the edge of its radius (in the case of frag grens).

Note: There would be knock-on effects to gameplay if this was fixed, and should only be fixed if we have a plan in place to deal with those.

Here's a commit that reverted some accidentally committed test code that implements a fix and prints the differences: 24eb0aa

[squeek502]

I misremembered the effects of this. Turns out most explosions do less than they would if the falloff was "fixed" (meaning 0.5x damage at the edge of the explosion for all explosions).

Damages close to the edge of the explosion radius:

Thing exploding	Base damage	"Fixed" damage	Current damage
Frag Grenade	145.000000	68.820503	3.148529
IC Rocket	55.000000	26.769991	-9.159111
RPG Rocket	102.000000	49.961639	50.981998
Grellow Pipe	108.000000	50.931957	39.306984
MIRV	180.000000	74.475624	21.713440
MIRVlet	145.000000	57.599770	-17.745255
Unbounced fully charged rail	40.000000	20.641932	-8.395172
Backpack by an EMP	50.000000	33.925419	-27.157990

So... yeah. It has wide-reaching and substantial effects (except on RPG rockets).

[mikeparker]

Good findings, nice table.

I think personally, that I agree with valve's initial direction - that if you get hit by an explosion, you should know that you've taken at least a reasonable amount of damage. I think the game is more consistent and easier to read if you don't get hit for 1hp by a mirvlet, or 0hp from an IC rocket!

I think there are two variables we need to decide on:

• Should the radius % be adjustable per explosion, or automatic? TFC's explosions had an automatic radius based on where the damage reached 50% power.
• Should the falloff rate be adjustable per explosion, or automatic? TFC's explosions all had the same falloff rate.

So, the combination of these 2 variables gives us 4 possible outcomes:

1. TFC explosions - fixed falloff, fixed radius
2. Current FF explosions - fixed falloff, adjustable radius
3. Option3: Adjustable falloff, fixed radius (the edge will always be 50% but you can adjust the falloff so explosions are more localised or more spread out)
4. Option4: Adjustable falloff, adjustable radius (the edge will be whatever % we want, and the fall rate can be as fast or slow as we want)

Right now, we're using a hard coded value for the radius rather than a percentage based radius. If we're using a hard coded value it doesn't make sense to use a fixed falloff rate, so the first thing we should change is either varying the falloff rate by using the radius value, or changing the radius value to be percentage based.

I would either vote for option 1 or option 3. I think option 4 is too flexible to sensibly balance and I think option 2 is pointless.

[squeek502]

I'm not sure I follow what you mean by fixed/adjustable falloff. Right now, we have a fixed falloff and an adjustable radius. However, that fixed falloff does not account for the adjustable radius, so instead we have a weird hybrid of a fixed falloff that assumes a fixed/damage-based radius while we actually have an adjustable/damage-independent radius.

Given that we're choosing to fix the above inconsistency, I lean towards option 2, as it seems the most straightforward: explosion radius can be adjusted independent of the damage and the falloff will remain consistent. Option 2 is also what the "Fixed" column shows in the above table.

Option 1 is not an option if we want to keep frag gren damage the same (wouldn't be able to lower the radius of frag grens without decreasing their damage).

I don't understand why option 3 would be desirable; it's simply a slightly more flexible option 1, in a way that won't make much real difference.

And option 4 is just a more flexible option 2, which might end up being useful if the consistent 50% falloff creates too many balance problems.

[mikeparker]

To clarify, when i say fixed falloff rate, i mean that no matter what the radius is, a certain distance away from the grenade will always cause the same damage (provided you're inside the radius), which I believe is the current situation. You can "fix" the current situation by either snapping the radius to where 0.5x is (and ignoring all the radius numbers), or "stretching" the damage so that the current radius values are where the 0.5x damage is.

Maybe i should draw some graphs, haha

[squeek502]

Okay, I get what you're saying. I'd be really curious to see what the current 50% damage radiuses end up being for things like the frag gren, IC, railgun, etc. For the frag gren, it would cut the radius basically in half by the looks of it.

[mikeparker]

Yeah, it means aimed grenades are still just as vicious but unaimed / spammy / dodged grenades could be ignored rather than dealing a medium to small amount of damage.

[mikeparker]

Would be interesting to know the blue pipe values too. I guess thats one of the biggest contenders to have its radius reduced, if indeed it's radius is larger than TFC.

[mikeparker]

As a simple fix, I wouldnt mind having a lower cutoff at like 25% or something, that would be quite easy to implement and woudlnt really change the gameplay drastically. That would basically mean that if an explosion will do less than 25% of its damage, you don't get hurt.

That would result in minor nerfs to all the above except rockets and grellows.

[Largos]

"Yeah, it means aimed grenades are still just as vicious but unaimed / spammy / dodged grenades could be ignored rather than dealing a medium to small amount of damage."

Whichever option leads to this, I like. So . . . a fixed option 2 from above?

I assume buildings are affected the same by all this? Or do they undergo different explosion response properties?

[squeek502]

Was curious, so I made a table of how the radiuses would be affected if we scaled them based on where they currently do 0.5x damage or 0.25x damage.

Weapon	Base Damage	Defined Radius	Damage at defined radius	Actual Radius	% of Base Damage At Actual Radius	0.5x Damage Radius	Radius Difference (0.5 - Actual)	0.25x Damage Radius	Difference (0.25 - Actual)
IC	55	125	-7.5	110	0.00%	55	-55.0	82.5	-27.5
Frag Gren	145	270	10	270	6.90%	145	-125.0	217.5	-52.5
Railgun	60	100	10	100	16.67%	60	-40.0	90	-10.0
Mirv	180	270	45	270	25.00%	180	-90.0	270	+0.0
Mirvlet	145	217.5	36.25	217.5	25.00%	145	-72.5	217.5	+0.0
Grellow Pipe	108	130	43	130	39.81%	108	-22.0	162	+32.0
RPG	102	100	52	100	50.98%	102	+2.0	153	+53.0
Blue Pipe	108	100	58	100	53.70%	108	+8.0	162	+62.0

Spreadsheet of the table:
https://docs.google.com/spreadsheets/d/1YOUx3pYZSYi9P44bdAxJeMa-Jwmc7G-n0SWYjY3lQh0/edit?usp=sharing

[squeek502]

Did some more testing of TFC and was surprised by the results:

• fortressforever/tfc-reference-data@2a8a9a5
• fortressforever/tfc-reference-data@69a64c7

Here's a similar table showing radius falloff for TFC:

Weapon	Base Damage	Radius	Damage at radius	% of Base Damage At Edge of Radius
Frag Gren	162	352	0	0.00%
Grellow Pipe	108	120	45	41.67%
RPG	83-101 (varies)	109 max (varies)	~40-50 (varies)	~50% (very approximate)

(MIRV and MIRVlets are the same as frag grenades)

---

So it seems like TFC also decouples radius, falloff, and damage, and it varies between damage sources, and that FF's damage falloff is more TFC-like than I thought.