As we know, there are certain skills of the DoRam which are casted twice, SU_SV_STEM_SPEAR for example. The skill descript states that chance is increased every 30 levels, so I've used this piece of code at skill.c. I've assumed a 5% base chance * Sk_lv/30.
case SU_SV_STEM_SPEAR:
rate = (sd->status.base_level)/30;
if( rnd()%100 < 5*rate )
skill_castend_damage_id(src, bl, SU_SV_STEM_SPEAR, pc_checkskill(sd, SU_SV_STEM_SPEAR), tick, 1);
break;
@@Rytech, suggested to change it to
case SU_SV_STEM_SPEAR:
rate = 5 * sd->status.base_level / 30;
if( rnd()%100 < rate )
skill_castend_damage_id(src, bl, SU_SV_STEM_SPEAR, pc_checkskill(sd, SU_SV_STEM_SPEAR), tick, 1);
break;
The problem is that it produces inaccurate results as I'm dividing the base level and then multiplying by 5. For example a level 166 DoRam, would cast a second time at 27%, whilst it should double-cast at 25% chance since every 30 level the chance is increased rather than "every level/30".
So I thought that "counting" the times the base level of the DoRam is divided by 30 should suffice, for example:
case SU_SV_STEM_SPEAR:
for (int number = 1; number <=sd->status.base_level; number ++) {
if (number%30 == 0){
count++;
}
}
int rate= 5*count.
if( rnd()%100 < rate )
skill_castend_damage_id(src, bl, SU_SV_STEM_SPEAR, pc_checkskill(sd, SU_SV_STEM_SPEAR), tick, 1);
break;
It seems a little lengthy for me, but in theory, it should return more accurate results from this one. What do you think?
