Hollow Point Ammo feat. Fixing Dex?

AlexeaNikkole noticed here that her ArmaLite was being fully mitigated by armour when using HP rounds.

It has been stated that HP rounds multiplies opponent armour by 1.5x, meaning you would expect to do 2/3rds the regular damage. Link to this claim: https://www.torn.com/forums.php#!p=threads&f=19&t=16015530&b=0&a=0&start=0&to=17954864

The claim was made because of a bug report in which damage went to zero, and it was explained that this 1.5x factor could lead in rare situations to armour going over 100 (i.e. 100% reduction - damage). This is no longer possible with current armour values, so they only way to get to 0 damage would be some rounding down in rare cases, but the damage done with regular ammo would have to be super small too.

I tested this myself:

HP Ammo - https://www.torn.com/loader.php?sid=attackLog&ID=c70a076e76dc1569293550f48e568ba2

Regular Ammo - https://www.torn.com/loader.php?sid=attackLog&ID=1acefad033c1cc962ea62b69297ad1a0

As you can see, there is no cause for my damage to be rounded down to 0.

Was the 1.5x value increase armour not correct? Has it been changed? Is there an additional effect not mentioned?

Attacks on CrazedOne (e.g. https://www.torn.com/loader.php?sid=attackLog&ID=e307e89dc4857b0a4a2a45166576ec83) lead me to believe it's multiplying the STR/DEF value used to calculate mitigation, rather than armour perhaps?


=====================================================================================

Thread Summary


- Started by identifying a problem with the way "Special" ammo mechanics work (in particular Hollow Point).

- Chedburn reconfigured to correct the issue, and mentioned the next thing that needed sorting out was the way weapon "Accuracy" interacts with battle stats (finally!)

Currently: Display Accuracy -> Converts to Speed Multiplier, obtain new SPD/DEX ratio -> Calculate hit chance
This system allows for tiny Speeds to hit huge Dexterities, making it very unbalanced relative to Defense.

- A few ideas were discussed, the favourable contender at the moment is a system where the "boost" to hit chance, given by a weapon's accuracy, will be most effective when hit chance is 50% and will taper off to 0 at the extremes (0,100%). Graph incoming or Table located top of Page 2.

[image: i.gyazo.com]

- Weapon accuracies will range between MAX and -MAX, with MAX likely being 15. This will display on a weapon as 50.00+accuracy, so a 15.00 would be 65.00, and a -15.00 would be 35.00.

- Accuracy boosters (mastery, edu, XP, mods, jobs, ammo, faction) will be much simpler boosts of 1-5, for an absolute of maximum of +18, and will apply directly on to the weapon accuracy (meaning the absolute best weapon would be 65.00 + 18 = 83.00).

- Being passed to committee.

Couldn’t agree more! Would love to help ;-)

Thanks.

#ProxForDev

#HireTylerToWorkOnAPI

Prox for committee, screw being a dev :P

I won't change Armor/Defense in case you want to continue to use me as a lab rat Prox. Load image: smiley

If my understanding is correct, the idea is to make accuracy buffs similar to armor and damagae buffs in terms of how they're applied?

Specifically, armor and damage currently scale the final damage value only after the str v def calculations, meaning that a small value can only be raised or lowered by a small, absolute amount. A value of 0 remains 0.

The proposal looks like the new accuracy system would be either:

• A) If someone's spd v dex hit rate was only 2%, +100% accuracy would only given them 4% final accuracy.
• B) If someone's spd v dex hit rate was only 2%, accuracy would be a flat modifier, e.g. +10% accuracy = 12% final accuracy; 0% hit rate would be achieved by the initial hit rate going negative.

I think either would be an improvement. At a glance, the multiplier idea feels more "accurate." The flat rate version seems like it would suffer from the same problems of scalability that the current system suffers from.

This just might make dex sex again.

I’m on mobile right now so haven’t yet been able to respond how I want to but..

- I generally agree, I think multiplication gets out of control and the logical solution becomes to scale it with the hit chance you’re already at, which is what you’re trying (correctly imo) to move away from

- flat 35% is insane. Off the top of my head that would respresent 50x-ing SPD (effectively) in extreme cases. E.g. for SPD/DEX = 2, hit chance is slightly more than 65%. This becomes 100% which requires 100x SPD/DEX without the bonus.

I’ll comment my thoughts/idea in about an hour.

I see the issue, good point, but it seems like the proposed flat increase breaks in the opposite direction.

spd vs dex: 1% + 35% accuracy = 36% hit chance

1/100th speed shouldn't be able to hit so easily. I presume that includes weapon mastery and experience, so that 1:1 spd vs dex is 85% at most?

Some alternative options off the top of my head:

A) Convert all accuracy bonuses to speed bonuses and adjust the scaling, e.g.: +20% accuracy should not be +50% speed, but something lower.

B) If using flat increase from accuracy, adjust the speed vs dex formula to make the hit rate drop off significantly faster, e.g.: commonly available weapons + mastery + exp + mods should still result in 1% or lower hit rate against 100x dex.

C) If using multipliers, scale the accuracy bonus values down, e.g.: +20% becomes +10%, so the sum of your mods + mastery + exp + weapon can only amount to 70% or so, equivalent to +35% flat rate at 1:1 spd vs dex.

not finished.

Key: SD = User Speed / Opponent Dexterity

OK here we go..this might be long, not sure. Sorry.

Problems

1)

Speed multiplying really doesn't work. This isn't a new conclusion, feel free to skip the rant if you like, but just in case anyone isn't yet convinced...we take the most accurate weapon in the game, Alexstrasza's AK with maxed mods, 79.64 accuracy....

(a)

79.64 display accuracy is a flat ~4.6x boost to speed across the board. The most basic point here I can make is that, even starting at quite a silly SD needed for 0% hit chance (1/100), this is further reduced to (1/460). The two defensive systems are not even nearly analogous to one another, and that is extremely bad for game balance.

Conclusion (a) - Dexterity is ruined by accuracy at extreme SD region. Can hit opponents with <1/400 SD!

(b)

The graph below illustrates how it affects the final hit chance. Blue is regular hit chance (y-axis) as a function of SD (x-axis). Purple is the same thing, but for someone using the GAK mentioned.

[image: i.gyazo.com]

Green shows the % point difference this makes across the board, peaking at 35% or so around SD~0.46. y-axis is the %-point difference, x-axis is SD again.

[image: i.gyazo.com]

[image: i.gyazo.com]


Not only is +35% reached (this is purely weapon accuracy it should be pointed out - with XP, edu, Mastery etc it becomes even further detached from sanity), but this graph peaks in what you would expect to be the most common SD region for a typical fight (i.e you are fighting people with similar stats to your own!).

Conclusion (b) - Dexterity is ruined by accuracy in typical fighting range SD. Dexterity sucks all round. Booooo.

I know this has been done to death by now, but I just needed to vent and hopefully we can never return to this shit show :-)


Solutions?

2)

Flat +35% is too big. At extremes this will provide 50x multiplier to speed which is not acceptable, 2 posts above have mentioned this.

Flat bonuses in general are tricky, trying to find the balance between being too overpowered in particular regions and completely negligible in others isn't something that is easily done.

3)

One idea I had was to define a maximum Hit Chance boost that you're comfortable with. In this example I'll use +15%, but it could be adjusted if too high or low. (15% would provide a speed multiplier of ~7.2x at the extremes, which isn't horrible considering the rest of the scheme).

Weapon display accuracy would then provide a fraction of this boost -

- Fraction = (display_acc - 50.00) / 50.00

--> 85.00 becomes 10.50% Hit Chance boost, 75.00 becomes 7.5%, 65.00 becomes 4.5%
--> 15.00 becomes -10.50% Hit Chance boost, 25.00 becomes -7.5%, 35.00 becomes -4.5% etc..

Based on my knowledge of the SD vs Hit Chance formula, it would need to be adjusted for the 1/100 regime. If you're set on having 1/100 as the break-point for 0% regular hit chance, then my recommendation would be moving to a scheme such that at SD = 1/200, Hit Chance = -(maximum Hit Chance boost), e.g. -15%.

Mastery, XP, etc would all multiply upon the weapon display accuracy. At 10/10 mastery for example:

--> 85.00 (becomes 110.50) becomes 16.575% Hit Chance boost.
--> 35.00 (becomes -45.50) becomes -1.35% Hit Chance boost

This could even absolutely max out at the maximum Hit Chance boost. (e.g. 85.00 would only provide 15% boost).

Any way, it's just one idea for now that provides what I think is a solid-but-not-OP boost to Hit Chance, while alleviating the problem of dexterity being a bit shit when facing an average weapon.


4)

Man, just fuggin throw out this 1/100 -> 100 shit and lets make it consistent with STR/DEF!

Make it 1/14x for 0% hit chance, even make it 32x for 100% hit chance, and we can all have a party.

You could even keep (but reduce) the speed multiplying factor, and just impose a restriction that the final hit chance without the accuracy boost has to be > 0 (again consistent with STR/DEF, as once 1/14 is reached no weapon damage can lift the full mitigation).

- What you described with the arms race is what currently occurs with STR/DEF because you can fully mitigate someone. Some people are even going for it with Dexterity, but obviously finding it slightly harder because of the aforementioned issues.

- One thing that should be done is enforce a hard line on what ratio has a 0% hit chance. In the calculation you should first ask, is SPD/DEX >0.01 (personally in favour of increasing this, but lets say it's 1/100 still). Only after this condition you should calculate a basic hit chance, and then modify it with any accuracy boosts. This completely eliminates any "10,000x less speed receiving default hit chance" etc, and it becomes moderately consistent with the STR/DEF scheme.

- You should absolutely not implement a guaranteed hit and miss chance.

- The curve it's self works quite well I think actually, I like the reciprocal function aspect. If you want accuracy effects to not be flat but rather for example, be half as effective at 25/75, something along the lines of:

weapon_accuracy_boost * |[1 - |(spd_vs_dex_hit_chance% - 50%)| / 50]| 

would work. At 1:1 ratio = 50% hit chance, a weapon would provide it's full bonus. At 75% and 25% hit chance it would do half as good etc.

Looks pretty great from a brief inspection!