Celtic Heroes

OK this is an extension of the multiple topics of formulas. We are attempting to reverse engineer the formulas for our skills but to do that we need data. Everybody uses Fire Bolt so lets use it as the baseline skill. When you use a book of alt or rebirth, take the time to record the damage for every step you make. REMOVE YOUR GEAR when doing this so we don't get mixed data.

D amage (Fire Bolt)is a function based on three factors:

• L evel (of skill)
  A bility (fire)
  F ocus (stat)

So we are looking for a solution D = Function(L, A, F)

Use the following format for ease of copying.
L evel (of skill) var
A bility (fire) 0
F ocus (stat) 10

L D
1 21
2 25
3 30
4 35
5 41

This is what happens when you create a new mage, give him a lot of fettles and thus have a toon with no fire ability at all, but skill levels to put in fire bolt.
So this becomes the baseline or zero point that all formulas must pass through.

It is assumed that many of the skills use similar formulas so by learning one we can get the footprint of the others. But that one needs a lot of data.

L evel (of skill), var, var,
A bility (fire), 930, 930,
F ocus (stat), 245, 315,

L, D, D,
1 60 64
2 72 76
3 92 98
4 113 121
5 135 143
6 156 166
7 167 178
8 187 200
9 209 223
10 230 245
11 251 268
12 263 280
13 283 302
14 304 325
15 325 347
16 346 370
17 358 382
18 378 404
19 399 427
20 420 449
21 441 472
22 453 484
23 473 506
24 494 528
25 516 551

L evel (of skill) 25
A bility (fire) 931
F ocus (stat) Variable

F, Damage
10 304
20 326
30 343
40 357
50 370
60 381
70 392
80 402
90 411
100 420
110 428
120 436
130 443
140 451
150 458
160 465
170 471
180 478
190 484
200 490
210 496
220 502
230 507
240 513
250 518
260 524
270 529
280 534
290 539
300 544
310 549
320 554
330 558
340 563
350 567
360 572
370 576
380 581
390 585
400 589
410 594
420 598
430 602
440 606
450 610

hey papi ive got 8 mages i can use to get you some data. gimmie a few.

Ok, I was bored and could justify working on this as part of statistics homework, getting better at using R. Here are my findings using the data thus far-

1) There appears to be no focus dependent effect on skill level damage slope. Using the data from Papi at focus level 245 and 315 I obtained these slope intercept equations and graphs-

Focus @ 245 Damage = 19.05(firebolt skill level) + 37.40 Focus @ 315 Damage = 20.39(firebolt skill level) + 39.32
These equations are not pin-point accurate since I rounded to hundredths, but they should be good enough.

What does this information tell us? At a set level of focus, and fire ability, there is a linear increase in dmg as you add a point to the skill. Unfortunately, this linear increase does not have a fixed slope at varying amounts of focus.

By looking at a scatterplot of the data provided by Papi’s mage of focus vs dmg we see this curvilinear relationship-
At this point I don’t know what the function of the data between focus and dmg is, the linear line I fit to the graph for reference is of the equation-

Damage = 0.6251(focus) + 350.0020

Further data I think would be useful is-

-All variables the same except varied character level

-All variables the same except varied ability level (fire ability in the case of firebolt)

mancandy: Thanks for taking the time.

Damage as a function of focus is hyperbolic. if you graph the data on a semilog or log log chart you will see that it never goes linear so the equation is not a simple polynomial nor exponential.

The graph does appear asymptotic which makes sense. As we get to higher levels one would expect the relationship to be more linear as opposed to some parabolic function that would significantly reduce the effect of higher stats.

Damage as a function of focus is roughly linear. I say that because there is a pattern that follows every 5 points. It is a sawtooth curve but can be approximated with a linear relationship.


Damage vs Abilty is the big unknown. We need a lot more data points to do this one. We need to keep focus and skill level constant in order to do this though. I have an idea how to do this but it would be very time consuming for one person. If you create a new mage and feed him fettles until he is level 30 or so then you can start him out with zero ability and pick say skill level 1 and focus 10 and go level his abilty and record data points when his damage goes up. This will get us a curve for ability from 0 to 300. Using 1 and 10 would make a nice baseline or zero point that all our other curves will have to pass through.

Mancandy, welcome to the mage-scientist guild lol.

As a dev or beta tester said, it is a log-exponential graph. Use the excell function =logexp(

I havent tried it yet so let me know how it works.

Hello Papi,

I suppose I should have mentioned that I did try some log transforms and quadratic formulas to fit the dmg vs focus line but had no success and gave up as I realized it wasn't gonna be so easy . I might be too much of an R noob at the moment to tackle that one. So far as fire ability. One way I can think to get some quick data is with the use of ability boosters. Here are a few data points from Mancandy-

Character level- 107
firebolt skill level- 25
Focus- 350

ability dmg
1061 577
1261 577

...and we notice a problem that hadn't occurred to me, ability doesnt register higher dmg in the skill. So I guess I'm not sure how ability correlates to dmg. Anybody else?

currently I have had no luck with rings or diamond weapons abilty boosting showing any change to the skills. One more thign for OTM to fix.

Ya, I was wondering if that is a bug or if ability level modulates some variable besides skill dmg (i.e. hit rate percentage). One suggestion a friend of mine made yesterday, is it possible to examine source code for the game? I haven't the slightest knowledge of extracting source code from the iphone so maybe someone else has this skill? Of course, that does remove some of the investigative intrigue in figuring it out for ourselves. Depends on what you are in this for I guess