Welcome
Welcome to darkrisings

You are currently viewing our boards as a guest, which gives you limited access to view most discussions and access our other features. By joining our free community, you will have access to post topics, communicate privately with other members (PM), respond to polls, upload content, and access many other special features. In addition, registered members also see less advertisements. Registration is fast, simple, and absolutely free, so please, join our community today!

Avoidance

Title says it all.

Posts: 308

Joined: Mon Jun 09, 2014 3:47 pm

Post Sun Dec 27, 2015 9:10 pm

Avoidance

Avoidance

There are several different methods of avoiding damage once you are hit by your opponent that occur in a very specific order. Each of these abilities has a specific way of checking to see if it works or not, and I hope to detail them all out for you below.

To start, there is a specific order in which the abilities are checked:
Parry, Palm Block (monks only), Dodge, Shield Block and Shadows (Gypsy guild spell).

Parry

Parry is a skill that encompasses a vast number of factors to determine if the skill succeeds or not once an attack lands. To begin, you start with a (Skill/2)% chance as a baseline for success, and add 10% to the roll if you are using either a two handed weapon OR a staff, spear or polearm (these bonuses do not stack and do not matter for a weapon held in your offhand, only a ‘wielded’ weapon counts). Another bonus is available to rangers who are dual wielding of +20%. At the end of the formula is a modifier in the event that you, the person parrying is blind. Additionally, all non-magic classes add their Hitroll/7% to their parry rate as well. Finally, all blind non-Drow (remember, they have the blindfighting racial buff), then divides the sum of this formula by 2.

This makes the formula:
Parry% = ((Skill/2)% + (2h or type modifier [10%]) + Dual Wield Modifier [20%] + Hitroll/7%) / Blind Modifier [2 if blind and not Drow]

Global Note: Dark Risings rounds down, always and forever.

Examples 1: A Draconian Ranger with 100 hit dice is dual wielding a polearm and any other weapon and has their parry skill at 100%. This would mean their parry chance works out to be (50% + 10% + 20% + 14%) = 94%. (Writers Note: This value seems absurd to me)

Example 2: A Kine Mage with 100% parry skill and a staff would be: 50% + 10% = 60%
Example 3: A Kine Barbarian with 108 hitroll wielding a 2h axe would be: 50% + 10% + 15% = 75%

Palm Block

Palm Block is a monk only parry alternative, and a much simpler skill to calculate than parry… and it’s on hold until after I talk to Kes to make sure I have it correct. The big thing is there are actual stat checks in it (so high dex matters).

I’ll let people know when this gets updated, but it matters to so few people I won’t hold up posting the rest of the write up.


Dodge

Dodge isn’t as complex as parry, but there are still a few extra factors that work into the calculation.

To start, your base rate to dodge is your skill divided by 2. There is a level modifier on the ability (you have a lower dodge rate against higher level enemies - mainly changes your rates in PvE), as well as a hitroll modifier, associated with melee classes. Melee classes will decrease their opponents chance to dodge by (Hitroll / 10)%. Additionally, in PvP, being attacked with a weapon that you’re vulnerable to will reduce your chance to dodge the attack by a flat 15%.

So, to try and put it all together in a formula, your chance to dodge is based on:

Chance = (Dodge Skill % / 2) +/- Level Differential (Usually 0 in PvP) - 15% (If vuln) - (Attackers Hitroll / 10)% [If attack is Melee]

Shield Block

Shield Block is one of the simplest checks. Your chance is (Skill / 5)% + 3%. That’s it.

Shadows

For Gypsies, shadows is the easiest check. You have a 30% chance to step into the shadows against an attack.

Dragon Punch

I’m not 100% certain where dragon punch works into the equation (the above abilities are all part of a single function for the order of checks), but the function does not seem to include dragon punch in it. I’d assume it comes after all of the above abilities.

The chance to fire is based on your base skill% / 6. There is then a level differential check, though it is very minimal as the result of the check is divided by 2. Your entire dexterity stat is added to your chance, but your opponent's entire stat value is subtracted (it acts as a check, but only indirectly in the code) and half of your strength stat is added to the chance.

If it fires, you will do your equipped weapons damoun in an attack back. If you are unarmed you will do bash damage back.

Conclusion

In addition to releasing this information, I am also releasing roughly 60,000 iterations of raw attack data I’ve collected on 1315 in the last few weeks thanks to Lafindan’s combat scraping script. I would love to hear feedback on how the system currently works (in conjunction with the new AC system) to try and figure out some reasonable changes to make combat make a little more sense. I have some basic statistics done inside the file, but the data can be looked at in far more detailed ways.

You can find the data here. If there are other iterations of the data that would be helpful for someone to look at, please let me know and I can get it taken care of.

I’m also aware that there are some oddities with the data (mainly in the miss rates), we’ll have to look into that in the near future as well.
Yeah? Well, you know, that's just like, your OPINION, man.

Posts: 248

Joined: Wed Jun 11, 2014 1:54 am

Post Tue Dec 29, 2015 8:13 pm

Re: Avoidance

I will try creating a reply on sections because it's a lot of information.

* Hit Roll scaling on parry/dodge isn't really necessary anymore. I'd rather see it scaling with something else.
It was important because stacking HR sucked when AC didn't mean a thing...and now that it does, players that want to melee already have to stack it (plus getting AC debuffs).

* Parry is high when compared to dodge and absurdly high when compared to shield block.
In reality. the main problem with it is Dual Wielding with the unexplained 20% bonus. I mean...it would make sense if rangers lost shield block in order to DW (because how the hell are you supposed to wield two weapons AND a shield? Legolas 'shield' boarded while shooting arrows but that takes things to a whole new level of non-sense).
I still believe that it deserves being slightly higher than Dodge because players lose parry when disarmed (we'll discuss disarm below).

* Dodge is alright by itself. But I believe that the HR scaling could be replaced by a DEX scaling. The reasoning is that dodge is a skill related to agility and thus should scale with its respective attribute.

* Shield Block sucks and I have no clue why its value is so low. Like weapons, shields can also be removed from the player (by dislodge) but buffing the block rate without lowering the parry/dodge rate would lead to an increase in avoidance. With that in mind, my idea is giving shield blocking an alternative use such as having a chance of blocking "direct attack spells/skills". That functionality could have a scaling based on the shield's weight and the player's constitution or intelligence.

* Dragon Punch is problematic because its real rate is lower due to how the avoidance checks work: parry first, then dodge and if you dodge, then there's the punch chance. So while it's actually something like this: [1 - parry chance] * [dodge chance] * [punch chance] instead of only [punch chance].
For example, versus someone with 60% parry, 50% dodge and 25% dragon punch chances, out of 1.000 attacks only 50 of these would be dragon punched.
However, the issue sort of solves itself if parry values are slightly lowered and blocks become slightly better.

------------------ Quick idea for making Disarm / Dislodge meaningful:

> IF the weapon isn't cursed, weapon is dropped in the ground
> IF the weapon is nodrop, weapon is 'dropped' to inventory and disarmed player gets parry debuff = disarmer weapon's weight
> IF the weapon is noremove or sticky, weapon remains but player gets parry debuff = double disarmer weapon's weight

Parry debuff can be healed through heal limb spell (it needs a function other than wither limb).

So a player, as a disarmer, can choose to wield a 20 weight weapon at the risk of losing it to STR stacking (specially if overcap is fixed) and have disarm as a way of dealing with avoidance.

On the other hand, a player being disarmed can opt for a noremove weapon if he feels like he can deal with having to heal limb every now and then to keep the parry skill up...or run a nodrop weapon for a "not as punishing" thing, even though a blind+disarm would be more crippling....or run an uncursed weapon and get a macro for pickup + wield...and risk blind + disarm too.

Apply the same concept for shields and shield block.

Posts: 308

Joined: Mon Jun 09, 2014 3:47 pm

Post Tue Jan 12, 2016 5:56 pm

Re: Avoidance

Here's a quick idea for a new parry formula:

Parry would be a base of 40% (-10% from current)+ 5% 2h/staff/spear/polearm modifier (-5% from current) + 5% Dual Wield (-15% from current) + HR/7 Melee + Weight Check (Explained Below) + Weapon Skill Check (See Below)

Weight Check = (Defender Weapon Weight / Attacker Weapon Weight) * 5%, Capped. If the formula works out where there is a modifier that would raise the parry bonus above 5%, it remains at 5%.

Weapon Skill Check = (1 is yes, 0 is no)
All checks are based around the *defenders* weapon skill.
If the attacker doesn't have the defenders weapon type, 5% bonus (assumed to be 'unfamiliar' with it)
If the attacker and defender have each others weapon type , ((Defender WS / Attacker WS) / 2) * 5%) - Remember DR rounds DOWN, so at beast a 2%


I feel the weapon skill formula could be far better, so input there is welcome. I was having trouble formulating it. It would be improved by the inclusion of meaningful attributes.


To compare this with the example above: (A Draconian Ranger with 100 hit dice is dual wielding a polearm and any other weapon and has their parry skill at 100%. I will expand it to fit the new formula to say the polearm weight is 20, the defender is using a dagger with a weight of 6, and they share the same weapon skills).

40% + 5% (polearm) + 5% (DW) + 14% (Hitroll) + 5% (Weight) + 2% (Weapon Skill) = 71%, down from a prior value of 94%.
Yeah? Well, you know, that's just like, your OPINION, man.

Posts: 248

Joined: Wed Jun 11, 2014 1:54 am

Post Wed Jan 13, 2016 12:45 pm

Re: Avoidance

I feel like the 40% part should be based on parry% so that the forget spell has an effect over it. I know...psis are rare.
It could either be: [parry% / 2.5] or [5 * parry% / 2] , depending on how the calcs are done.

At the risk of overcomplicating things, I have a suggestion for the 2h/staff/spear/polearm modifier:
+5% if the character is wielding either of those weapons WITH a shield
+10% if the character is wielding either of those weapons WITHOUT a shield

I believe the HR/7 being restricted to Melee combined with Weapon Skill check and Weight can potentially create a big gap between might and magic classes.
My suggestion is either granting the HR/7 scaling to casters as well or replacing with an attributes check: 2% * [(defender's STR + INT) - (attacker's STR + INT)]
Logic behind it is that the character needs INT to read the opponent's moves and STR to properly execute the counter (parry).

Weapon Skill is solid.

---------------------------- Dodge suggestion:

Just like parry being affected by two attributes, so could dodge.
My suggestion is replacing the HR scaling with something similar to the proposed for parry: 1% * [(defender's DEX + WIS) - (attacker's DEX + WIS)]
In this case, WIS works as some sort of gut feeling while DEX is effectively avoiding the incomming attack.

Bonus: it's an indirect buff to monks since they revolve around DEX/WIS and dragon punch is triggered by dodging. More dodges = more punches.

Posts: 308

Joined: Mon Jun 09, 2014 3:47 pm

Post Tue Mar 14, 2017 11:03 pm

Re: Avoidance

It being based on parry skill sounds good to me, now that I re-read this. As long as the end result maths out to 40, it's good. I'll revisit this soon.
Yeah? Well, you know, that's just like, your OPINION, man.

Posts: 308

Joined: Mon Jun 09, 2014 3:47 pm

Post Wed Mar 15, 2017 3:59 pm

Re: Avoidance

Ok, I'm hammering out on this formula today.

Base (Skill * (5/2)) + 5% if(2h/Staff/Spear/Polearm) + 5% Dual Wield + HR/7 (if Melee) + Weight Check ((Defender Weight / Attacker Weight) * (5/2)%) + (((Defender WS / Attacker WS)/2)*5%)

As of right now this is only a slight tweak in the base rate (just in how it works, it should still work out to 40%. HOWEVER - HR/7 for melee only comes up as a sticking point.


Right now, the /7 part of the Hitroll inclusion gives melee a substantial 11-14% buff, depending on their HR. I 100% agree that melee should be better at melee stuff than casters, but all other aspects of parry have been somewhat normalized far away from bonuses that large. HR/10 and HR/15 are two options I've explored, but it would lower parry rates to possibly around 60%, or less ( I need to run the numbers).
Last edited by Some Guy on Wed Mar 15, 2017 5:09 pm, edited 1 time in total.
Yeah? Well, you know, that's just like, your OPINION, man.

Posts: 308

Joined: Mon Jun 09, 2014 3:47 pm

Post Wed Mar 15, 2017 4:54 pm

Re: Avoidance

Or, replace HR / 7 with a Hitroll differential check.

(Attacker Hitroll - Defender Hitroll) / 6. If it goes negative, the bonus becomes 0, and not negative. This is what I'm leaning toward now.
Yeah? Well, you know, that's just like, your OPINION, man.

Posts: 308

Joined: Mon Jun 09, 2014 3:47 pm

Post Fri Mar 17, 2017 9:19 am

Re: Avoidance

Ok, one more update formula wise.

Parry Rate = ([Base (Skill% * (5/2))] + [5% if(2h)] + [5% Dual Wield] + [(Attacker Hitroll - Defender Hitroll) / 6]% + [Weight Check Eliminated] +[Weaponskill Check TBD]) / # (Blind modifier, will not be as steep as it currently is)


Staff/Spear/Polearm is removed from the list with 2h weapons because a weapon skill check would make weapons count twice.
Yeah? Well, you know, that's just like, your OPINION, man.

Posts: 4

Joined: Sun Mar 27, 2016 1:36 am

Post Fri Mar 17, 2017 3:09 pm

Re: Avoidance

Some notes on why the math for avoidance is so funky:

I had a great big explanation written out, then I accidentally deleted it, so you're getting the cliff notes version:

Total Avoidance Rate is calculated as follows:

Total = 1 - ( (1 - P) * (1 - D) * (1 - S) ), where P = parry rate, D = dodge, S = shield block

Fictional example where each of these is 50%:

1 - ( (1 - .5) * (1 - .5) * (1 - .5) ) = .875

Decrease Dodge by a flat 10% and that changes to

1 - ( (1 - .5) * (1 - .4) * (1 - .5) ) = .85

Now, that doesn't look like much of a change. However, what really matters is how many attacks are landing, since that is how you determine damage dealt.

The total land rate for attacks is simply 1 - avoidance rate.

For the first example: .125

For the second example: .15

Now, that might still look like "not much", but the relative amount is very interesting: .15 / .125 = 1.2, or a 20% increase in damage dealt.

So, yes, reducing dodge rate by a flat 10% substraction, from 50% to 40%, increased damage dealt by 20%. This is the case with any flat addition or subtraction in avoidance: it will disproportionately affect damage. Let's use an example ranger:

Kine Ranger with 100 hitroll wielding a staff and something in offhand, being attacked by a Kine barbarian with with 100 hitroll.

The Ranger's avoidance rates are:

Parry: ( (100/2) + 10 + 20 + (100/7) ) = .94 (you will recognize this as the extreme example from the first post)
Dodge: ( (100/2) - (100/10) ) = 40% = .4
Block: ( (100/5) + 3 ) = .23

This means that total avoidance is: 1 - ( (1 - .4) * (1 - .94) * (1 - .23) ) = .97228

Yes, a fairly typical ranger avoids over 97% of attacks against a fairly beefy fighter. It's even worse if the attacker is a caster.

However, let's look at how that changes if the ranger is blinded: 1 - ( (1 - .4) * (1 - .47) * (1 - .23) ) = .75514

This means that the difference in land rate is (1 - .75514) / (1 - .97228) = 8.83 = 883%

Yes, a blind ranger takes 883% more damage from attacks than an unblinded ranger, despite still avoiding more than 75% of attacks.

If you were wondering why being blind appears to be the worst thing ever in pk... well, this is a big part of it.

The important takeway here is understanding why seemingly relatively small changes to avoidances can actually make for huge swings when put together.

Return to This is how it actually fucking works

Who is online

Users browsing this forum: No registered users and 1 guest

cron
Donate Now
Donate Now


Designed by ST Software for PTF.
Hosted by © 2017 FreeForums.org | Create a free forum | Powered by phpBB
About FreeForums | Legal | Advertise Here | Investors | Contact FreeForums.org
Report Violation
suspicion-preferred