Jump to content

Archived

This topic is now archived and is closed to further replies.

B.P.M.

[Xrd] A Strange Bit of Damage Scaling Phenomena

Recommended Posts

Hi, first time (in a while) poster, long time lurker.

I'm currently programming a tool that will calculate a combo's damage based on the moves it contains.  Right now, the program parses the wiki data into an editable format and uses that for reference. I'm running into some snags, but I'm not sure if this is outdated data or an Xrd mechanic we haven't discovered yet.

For example, let's take the combo AA 5P 6K(2) 2S(2) j.H j.D j.623H with axl.

When done in-game, the combo does 166 damage on Sol at 100% health broken down like this:

5P: 28

6K: 22, 17

2S: 17, 13

j.H: 25

j.D: 23

j.623H: 25

Total: 166

However, when I do the math, according to the wiki it's done like this:

Defense Modifier: 1.00x (ignoreable)

The combo never puts sol under 50% health, so guts can also be ignored.

5P does not have initial proration as of 1.0.

None of the moves have forced proration as of 1.0.

5P: 28 (OK)

6K: 22, 18

2S: 22, 18

j.H: 38

j.D: 40

j.623H: 50

Total: 208

How come the 42 damage discrepancy?  I'm legitimately confused.

Thank you,

--BPM

Share this post


Link to post
Share on other sites

renamed to RISC damage scaling. basically as you do attacks in a combo, each hit prorates the damage of any following hits. Pretty much every fighting game in existence has this.

basically, getting hit while the RISC gauge is 0 causes damage scaling, each hit causes more scaling, how much scaling is dependant on the move. there's no visual indicator for how much scaling is in effect.

The XX series had this too, except it was named Guard Gauge at teh time, and you could see how much damage scaling was in effect. I'm pretty sure Xrd uses the same system, we just can't see the guard gauge once it's under 0.

http://www.dustloop.com/wiki/index.php?title=GGACR/Damage#Guard_Gauge

Share this post


Link to post
Share on other sites

Frame data has values for how much RISC gets reduced if the move hits. This is basically your damage scaling (ie you could manually calculate how much RISC is below 0). I do however not know the exact math behind damage calculation.

Share this post


Link to post
Share on other sites

Testing reveals that damage scaling scales with the amount of negative RISC you have.

For example, Sol's c.S 2D on Sol does 58 damage if you just add their damage values together (30 + 28).

Below are some damage values for c.S 2D on Sol depending on initial RISC.

RISC: Damage

0: 55

1: 55

2: 56

3: 57

4: 57

5+:58

Since c.S takes away 7 RISC on hit, I assume damage scaling still happens for +5 and +6 initial RISC, but it's too small to be noticeable.

More to come after I do a fuckton more testing.

Share this post


Link to post
Share on other sites

Damage scaling remains largely undocumented in Xrd, but for the most part, it works the same as it did in the XX series. The Guilty Gear XX Accent Core Break Encyclopedia provided a graph for damage scaling relative to Guard Balance, which I happen to have on-hand. While I can't provide a screenshot of the graph itself, the figures themselves were posted on various Japanese sites back during the #Reload era, and those figures eventually made their way over to Dustloop's old Web Encyclopedia, namely the Damage Scaling page, which was then copied over to our current wiki. Perhaps because it was new to Accent Core, these figures fail to include the improvements to Overdrive damage scaling over normals and specials as scaling approaches maximum, so I've added those in myself.

Guard Balance and Damage Scaling
Guard Balance Normal and Special Damage Scaling Overdrive Damage Scaling
+128~0 100%
-16 78% (79%?) 69% (70%?)
-32 59% (60%?) 50% (51%?)
-48 44% 38%
-64 31% (32%?)
-80 18% (19%?)
-96 12% 15%
-112 6% 12%
-128 3% 9%

Most of these figures are rounded down, I suspect. From testing I've done previously, I believe maximum damage scaling is 1/32, so a little higher than 3%. The percentages in parentheses are differing results that I got while testing. To find these percentages, I used a move with 100 base damage on a character with a x1.00 defense modifier. It's worth noting that I got 32 damage with Kliff's 6HS (100 base damage) at -64 Guard Balance despite an Overdrive with the same base damage dealing 31 at the same Guard Balance value, so scaling isn't exactly the same at that point, contrary to the table. I tried using moves with higher base damage (as well as using the Extra Menu to increase base damage) in the hopes of getting more accurate results, but I ended up getting differing outliers when calculating the percentages, which I suspect may be due to rounding complications.

Edit: Fortunately, JAKE is a god and has a far more detailed set of figures (complete with his own representation of the graph) on his website. He seems to have tested it the same way I did, albeit considerably more thoroughly.

Now, all that being said, there are some differences to damage scaling in Xrd. Overdrives scale to no less than 20% of their base damage now. Proration also no longer affects Overdrive damage once scaling reaches 20%, although defense modifiers, Guts and OTG state are all still taken into account.

I should also point out that damage dealt during low positive into negative R.I.S.C. Level in Xrd is calculated slightly differently from damage dealt during low positive into negative Guard Balance in the XX series, as your c.S > 2D example does "55, 56, 56, 56, 57, 57, 58" rather than "55, 55, 56, 57, 57, 58" despite all individual values being the same. R.I.S.C. Level also seems to have some slight nuances that Guard Balance does not, such as moves with RLv 0 still visibly reducing R.I.S.C. Level at positive values, which does not occur with Guard Balance. That being said, hopefully this should still be a useful reference.

Share this post


Link to post
Share on other sites

Thank you for the surprisingly detailed post.  I've actually tested a move (Axl Bomber) at all RISC levels from 0 to -61, and slapped a trend line on the results (I know, shame on me, it's probably not THAT accurate, but it's a place to start)

I'll run those figures you posted against my data and see which one is more accurate.  Currently, my data is less accurate for longer, more damaging combos, we'll see if this fixes that.

EDIT: Since the PC Version came out, I did some poking around with cheat engine, and it turns out there are two numbers for the current hit: one raw and one prorated.

 

I'm trying to find the longer, more decimal-laden version of the prorate hit (since it probably has to go from integer to decimal and then gets rounded in the game)

More to come soon.

 

Share this post


Link to post
Share on other sites

×