Tree of Savior Forum

Critical rate formula (iCBT2 updated)

Disclaimer: I did not datamined client nor bribed staff member. Everything I did was spending hours whacking things and recording every hit made. Therefore, all listed below is nothing more than estimation - take it with a grane of salt. Feel free to contribute and point out possible mistakes.

Disclaimer 2: Math overload in this post may cause irreversible changes in your brain. There is TL;DR section below :stuck_out_tongue:


The reason why game client doesn’t show the exact % for stats like evasion or critical is because final rates depend on stats of your opponent like accuracy and critical resist and may vary greatly. But since the first Beta players started to assume that the final formula looks somewhat simple, like

(CritRate - CritRes)*N ; where N is some kind of stat modifier which magically converts stat numbers into %'s

There were also people (including me) stated that N being ~0.4 after some testing. Were we right? Well, yes and no. There were also rumors that crits count differently for different class branches, were they right? Again, yes and no.

Below is the table with crit statistics on different monsters. It also takes into account different character levels and different class branches (swordsman and archer). Special thanks goes to Hanamings for being so cooperative.


CharLv . . MobLv . .Crit Rate Stat. .Crit Res Stat . .Hits . . .Crits . . .Crit Chance . . Estimated Modifier

. . . 52 . . . . . . 3. . . . . . . . . . 88 . . . . . . . . .6 . . . . . . . . 522 . . . . 306. . . . . . . 58,6%. . . . . . . . . . 0,7
. . . 52 . . . . . 42. . . . . . . . . . 88 . . . . . . . .31 . . . . . . . . 240 . . . . 103. . . . . . . 42,9%. . . . . . . . . . 0,75
. . . 64 . . . . . . 3. . . . . . . . . 102 . . . . . . . . .6 . . . . . . . . 495 . . . . 296. . . . . . . 59,8%. . . . . . . . . . 0,62
. . . 64 . . . . . 42. . . . . . . . . 102 . . . . . . . .31 . . . . . . . . 397 . . . . 184. . . . . . . 46,3%. . . . . . . . . . 0,65
. . . 83 . . . . . . 3. . . . . . . . . 152 . . . . . . . . .6 . . . . . . . . 308 . . . . 220. . . . . . . 71,4%. . . . . . . . . . 0,49
. . . 83 . . . . . 42. . . . . . . . . 152 . . . . . . . .31 . . . . . . . . 302 . . . . 186. . . . . . . 61,5%. . . . . . . . . . 0,5

From this point it started to feel weird, so I decided to take a break and test on mobs with different levels after I will be close to cap.

. . 192 . . . . . . 3. . . . . . . . . . 90 . . . . . . . . .6 . . . . . . . . 251 . . . . . 50. . . . . . . 20,0%. . . . . . . . . . 0,24
. . 192 . . . . . . 3. . . . . . . . . 150 . . . . . . . . .6 . . . . . . . . 236 . . . . . 98. . . . . . . 41,5%. . . . . . . . . . 0,28
. . 192 . . . . . . 3. . . . . . . . . 206 . . . . . . . . .6 . . . . . . . . 177 . . . . . 93. . . . . . . 52,5%. . . . . . . . . . 0,26
. . 192 . . . . . . 3. . . . . . . . . 288 . . . . . . . . .6 . . . . . . . . 296 . . . . 196. . . . . . . 66,2%. . . . . . . . . . 0,23
. . 193 . . . . 153. . . . . . . . . 153 . . . . . . . .58 . . . . . . . . 130 . . . . . 30. . . . . . . 23,0%. . . . . . . . . . 0,24
. . 193 . . . . 153. . . . . . . . . 290 . . . . . . . .58 . . . . . . . . 201 . . . . 100. . . . . . . 49,7%. . . . . . . . . . 0,21
. . 193 . . . . 191. . . . . . . . . 290 . . . . . . . .72 . . . . . . . . 186 . . . . 100. . . . . . . 53,8%. . . . . . . . . . 0,25


Well…

What does it mean?

The ‘magical’ modifier N is not a constant. So far the only variable it depends on seems to be character level - the more you level up, the harder it becomes to land a crit. The dependence also may be more nonlinear but for levels below 200

N = ~42/CharLv

is close enough for build planning. And altogether

CritChance = (CritRate - CritRes)*42/CharLv

For a
CharLv = 50 . . N = ~0,8
CharLv = 100 . N = ~0,4
CharLv = 150 . N = ~0,3
CharLv = 200 . N = ~0,2

So after all number 0,4 appeared because test character was somewhat close to lv100. And ‘different class modifiers’ simply means that archers/swordsmen/clerics were all different levels during test. All close, all miss.


TL;DR

Q: So how do I know my rates?
A: ¯\_(ツ)__/¯

  1. Take Critical Rate stat from your character info window.
    Like this
  2. Substract from it value of Critical Resistance stat of your opponent. Monster database here
  3. Multiply resulting number by 42
  4. Divide result by your current level. Yes, I’m too lazy to make a spreadsheet now for one formula, bear with me :stuck_out_tongue_closed_eyes:
  5. Resulting number is your chance in percents. -xx% means no crit, 1xx% means you overdid with DEX.

CritChance = (CritRate - CritRes)*42/CharLv

5 Likes

And just to answer it before anyone ask:

The amount of crit rate to get 100% crit chance at lv 200, on a 0 crit res enemy, is 476
Or about 700-ish Str if you turn all those Dex into Str.

It’s not actually possible to get 100% crit chance without a crit chance modifying buff though, yeah?

@Noah_ Very nice, thanks!

On the assumption of using 0 buff/temporary benefits and no awaken bonus (since that’s apparently changed in OBT)
And no specific weapon, but I’ll assume 3 slots total (2 main hand +1 offhand or 3 main hand)

I’ll assume 300 Dex is reachable through raw stat, since just the stat from leveling alone gets you 260~ and there’s a lot of stat bonus from quests/Zemyna statues.

Archer get 1 crit rate every 5 level, so at lv 200 that’s 40 crit rate = 340 now.

Let’s say lv 4 green gems just for the sake of being more reasonable to get 3 of (don’t know how’s the nerf to monster gem exp) = 42-46 crit rate = 382-386.

Petamion/Max = 8-14 rate = 390-400
2 Cicel = 74 = 474-484
(the far inferior option of 2 rapid bangle would be 16 total, so that’s -56)

And then there’s +Dex on armor pieces.

If you consider some specific weapon choice (Heart Seeker) then Archer can potentially maintains 100% crit rate even after monster crit res is taken into account.

But then your base damage is kind bleh and you have no HP or SP pool to speak of.

I’m speaking of the assumption that there is a crit cap, so that even if you had 1000 crit rate vs a 0 crit def monster, that you would be capped at some crit chance %.

I did some tests with finestra 15, and even with a pure dex build ( + green gems ) with a very high crit rate I wasn’t able to get what felt like near 100% crit chance. It seemed like around 70-75%. However if I had any crit chance modifying skill ( barbarian’s old aggressor, ranger’s critical shot, rogue’s sneak hit ) I was easily able to get 100% for sure crit chance.

What I believe is that there is some cap when calculating crit rate, but that cap is bypassed when using a crit chance modifier.

Are you able to confirm this one way or the other?

Or make a SPR Krivis/Priest your personal slave, that also can solve the problem.

No, because I did not use any buffs on purpose. But pure DEX build with Finestra and green gems is exactly 70% or so. Don’t forget that as you level up flat bonus to crit start to drop while percent bonuses don’t have this problem.

Sorry I was replying to rofldat there :smile:

Hrm, 70% vs a 0 crit def enemy? At the lvl you’re expected to be at for hoplite3? (~126) Seems a little low… ( I had a bit over 320 crit rate ).

I guess to be 100% for sure, I’d have to have a SPR Krivis to debuff a monster / buff me, and test this out, yeah.

However in any case, this is useful. I was thinking hoplite 3 was a bit of a waste for finestra 15, since 10 was good enough ( and you still want DEX to be tanky, so you aren’t really saving points IMO ). Since you need more and more crit rate, this makes finestra 15 more valuable to my build.