Difference between revisions of "User:Hordes"
m |
|||
| Line 12: | Line 12: | ||
---- | ---- | ||
| − | Djinn have +11 aptitudes in magic, but are forced to train all skills equally. This equates to training a single school at a -3.3 aptitude - with a caveat. By the time a caster reaches | + | Djinn have +11 aptitudes in magic, but are forced to train all skills equally. This equates to training a single school at a -3.3 aptitude - with a caveat. By the time a caster reaches mid-game, their effective aptitude is often higher than -3 in practice.<ref>Djinn split XP between 12 skills at a +11 aptitude. As XP cost = <code>2^(-apt/4)</code>, each skill costs <code>2^(-11/4) * 12</code> = 1.78x the XP of a single skill. This equates to a <code>-4 * log2(1.78)</code> = -3.32 aptitude. But there's a quirk - since Djinn train the "Spellcasting" skill for free, they get more out of every skill level. |
| − | *Say you wanted to cast a single-school Conjurations spell. If you only cared about spell power/failure, a [[Kobold]] would be most efficient a point like {5 Spellcasting, 15 Conjurations}. This would cost | + | *Say you wanted to cast a single-school Conjurations spell. If you only cared about spell power/failure, a [[Kobold]] would be most efficient a point like {5 Spellcasting, 15 Conjurations}. This would cost 7275 skill points. |
| − | *A Djinni would reach the same power as the Kobold at {13 Spellcasting, 13 Conjurations}. Each skill has | + | *A Djinni would reach the same power as the Kobold at {13 Spellcasting, 13 Conjurations}. Each skill has 713 skill points, or 8556 skill points in total. This is only +17.6% more XP than the Kobold, or a '''-0.9''' aptitude. |
*Note that if the Kobold were to train more Spellcasting, for either MP or spell level purposes, they'd become less efficient. A Kobold at {10 Spellcasting, 13.5 Conjurations} would reach the same power, but at a cost of 79.13 skill points. | *Note that if the Kobold were to train more Spellcasting, for either MP or spell level purposes, they'd become less efficient. A Kobold at {10 Spellcasting, 13.5 Conjurations} would reach the same power, but at a cost of 79.13 skill points. | ||
: | : | ||
| Line 98: | Line 98: | ||
==Strategy== | ==Strategy== | ||
| + | Djinn split XP between 12 skills at a +11 aptitude. As XP cost = <code>2^(-apt/4)</code>, each skill costs <code>2^(-11/4) * 12</code> = 1.78x the XP of a single skill. This equates to a single school at <code>-4 * log2(1.78)</code> = -3.32 aptitude. However, this ignores the effect of Spellcasting skill. Once you take into account Spellcasting, Djinn are generally better than "-3.32 aptitude" suggests. There are two ways to think about it: | ||
| + | |||
| + | :1. Djinn need ''a lower skill level'' to get the same amount of spell power. For a regular species: if you wanted to maximize power, you would invest little into Spellcasting, and a lot into <Spell Schools>. For Djinn: you train medium-high <Spell Schools> and get medium-high Spellcasting for free. For example, going from 9 -> 10 skill is cheaper than 11 -> 12 skill. | ||
| + | |||
| + | :2. In practice, power isn't everything. A regular species will want to train Spellcasting for spell levels and/or max MP. But since Djinn don't need spell levels or max MP, they can dedicate more XP into spell power. Thus, Djinn are more XP efficient. | ||
| + | |||
| + | To take a worked example: | ||
| + | *If you ''only'' cared about spell power, a Kobold would be close-to-max efficiency at {4 Spellcasting, 15 Conjurations}. This would cost 7275 skill points. | ||
| + | *A Djinn would reach the same amount of power at {13 Spellcasting, 13 Conjurations}. Each of the 12 skills cost 713 skill points, or 8556 skill points in total. This is only +17.6% more XP than the Kobold, or a '''-0.9''' aptitude. | ||
| + | |||
| + | This happens because going from 12 -> 13 skill is cheaper than going from 14 -> 15 skill. | ||
| + | |||
| + | If the Kobold were to train more Spellcasting, for either MP or spell level purposes, they'd become less efficient. A Kobold at {10 Spellcasting, 13.5 Conjurations} would reach the same power, but at a cost of 7913 skill points. | ||
| + | |||
| + | ----- | ||
Revision as of 03:32, 13 December 2023
Relevant credentials: Hordes (greaterplayer) | Moriya (greaterplayer #2) | HorDDes (tournament)
(The rest of the page is just writing stuff that's WIP)
Djinn have +11 aptitudes in magic, but are forced to train all skills equally. This equates to training a single school at a -3.3 aptitude - with a caveat. By the time a caster reaches mid-game, their effective aptitude is often higher than -3 in practice.[1] And, since every school is trained at once, dual-school spells are very easy to train. Overall, a dedicated mage will not have trouble casting spells.
Tengu are a versatile, but frail species. Their aptitudes are good across the board, with impeccably good skill in Conjurations and Air Magic. And with 2 auxiliary attacks, melee is not out of the question. However, Tengu are among the most fragile species in the game: they have -20% HP and can't wear helmets/boots. Unlike Spriggans and Felids, Tengu do not get increased movement speed to compensate.
Note that Tengu are not small, so they don't get a bonus to the Dodging skill. It is true that Acrobatic and Evasive Flight grant EV, however, these bonuses are unaffected by armour encumbrance. Therefore, Tengu can wear heavy armour like any regular species.
List of gods, sorted by "when you get the first strong perk". (IMO)
Gods that benefit you on first worship:
Gozag has gold distraction (maybe worth listing here?), Fedhas pacifies sleepcaps
Gods that benefit you at 1*:
- Ashenzari (ID items)
- Gozag (Potion Pet)
- Hep (Ancestor)
- Nemelex (1st set of cards)
- Okawaru (Heroism)
- Trog (Berserk)
- Sif Muna (MP recovery)
- TSO (Halo, Divine Shield)
IMO these gods don't provide good benefit at 1*, but provide some benefit: Ely Purification, Zin Recite, Makh heal on kill, WJC Lunge/Whirlwind, Dith stealth, Kiku Wretches (hard to use w/o good spell).
Gods that benefit you at 2*:
Gods that benefit you at 3* or more:
- Kiku (Variable; based on skill training)
- Vehumet (Variable; based on skill training)
- Makhleb (3*; Lesser Servant)
- Ru (3*; Draw Out Power)
- WJC (3*; Serpent's Lash)
- Zin (3*; Imprison)
- Dithmenos (4*; Shadow Mimic)
Gods that hurt you on first worship:
When you start with a melee-oriented background, you'll get a starting weapon of choice.
Simple Starting Weapon
Used by most weapon-wielding backgrounds.
-
Short sword - Short Blades aren't great at regular combat, though they are good at stabbing. Uses dexterity instead of strength. -
Mace - Maces & Flails are the strongest "strength" weapon, at least in terms of stats. -
Hand axe - Axes have the ability to cleave. In addition to hitting a "main" target for 100% damage, they'll hit every monster adjacent to you for 70% damage. However, axes deal less damage than other weapons. -
Spear - Polearms have reaching - they can attack from an extra tile away. Despite their worse stats, spears are the best option to fight single targets, due to their extra range. -
Falchion - Long Blades are similar to Maces; higher base damage, but no gimmick. They use dexterity instead of strength. -
Unarmed Combat - Unarmed Combat is usually the weakest of the available "weapons". Certain species, like Trolls and Ghouls, receive special bonuses to Unarmed attacks. Unarmed can become strong at high skill, but that's a long while away.
Good Starting Weapon
Fighters and Gladiators start with a better ("good") starting weapon than other backgrounds.
-
Rapier - Short Blades are good at stabbing, but weaker at plain combat. Rapiers are the only "good" weapon that stats below 1.0 attack delay, meaning monsters won't get a chance to hit you twice. -
Flail - Maces & Flails have higher base damage, but no special gimmicks. -
War axe - Axes can cleave, but deal less single-target damage. Great option for strong species, like Hill Orcs and Minotaurs. -
Trident - Polearms can attack from a tile away. Like the spear before it, tridents are usually the best option against single targets. -
Long sword - Long swords are slightly more powerful than flails, though long swords use dexterity instead of strength. -
Quarterstaff - Quarterstaves deal more damage/turn than any other option, but they are two-handed, meaning you can't wear a shield with one. Because of this restriction, only Gladiators and Formicid Fighters can start with one. In addition, Staves are rarer than other weapon types. - Unarmed Combat - Unarmed Combat is usually the weakest of the available "weapons". Certain species, like Trolls and Ghouls, receive special bonuses to Unarmed attacks. Unarmed can become strong at high skill, but that's a long while away.
Strategy
Djinn split XP between 12 skills at a +11 aptitude. As XP cost = 2^(-apt/4), each skill costs 2^(-11/4) * 12 = 1.78x the XP of a single skill. This equates to a single school at -4 * log2(1.78) = -3.32 aptitude. However, this ignores the effect of Spellcasting skill. Once you take into account Spellcasting, Djinn are generally better than "-3.32 aptitude" suggests. There are two ways to think about it:
- 1. Djinn need a lower skill level to get the same amount of spell power. For a regular species: if you wanted to maximize power, you would invest little into Spellcasting, and a lot into <Spell Schools>. For Djinn: you train medium-high <Spell Schools> and get medium-high Spellcasting for free. For example, going from 9 -> 10 skill is cheaper than 11 -> 12 skill.
- 2. In practice, power isn't everything. A regular species will want to train Spellcasting for spell levels and/or max MP. But since Djinn don't need spell levels or max MP, they can dedicate more XP into spell power. Thus, Djinn are more XP efficient.
To take a worked example:
- If you only cared about spell power, a Kobold would be close-to-max efficiency at {4 Spellcasting, 15 Conjurations}. This would cost 7275 skill points.
- A Djinn would reach the same amount of power at {13 Spellcasting, 13 Conjurations}. Each of the 12 skills cost 713 skill points, or 8556 skill points in total. This is only +17.6% more XP than the Kobold, or a -0.9 aptitude.
This happens because going from 12 -> 13 skill is cheaper than going from 14 -> 15 skill.
If the Kobold were to train more Spellcasting, for either MP or spell level purposes, they'd become less efficient. A Kobold at {10 Spellcasting, 13.5 Conjurations} would reach the same power, but at a cost of 7913 skill points.
- ↑ Djinn split XP between 12 skills at a +11 aptitude. As XP cost =
2^(-apt/4), each skill costs2^(-11/4) * 12= 1.78x the XP of a single skill. This equates to a-4 * log2(1.78)= -3.32 aptitude. But there's a quirk - since Djinn train the "Spellcasting" skill for free, they get more out of every skill level.- Say you wanted to cast a single-school Conjurations spell. If you only cared about spell power/failure, a Kobold would be most efficient a point like {5 Spellcasting, 15 Conjurations}. This would cost 7275 skill points.
- A Djinni would reach the same power as the Kobold at {13 Spellcasting, 13 Conjurations}. Each skill has 713 skill points, or 8556 skill points in total. This is only +17.6% more XP than the Kobold, or a -0.9 aptitude.
- Note that if the Kobold were to train more Spellcasting, for either MP or spell level purposes, they'd become less efficient. A Kobold at {10 Spellcasting, 13.5 Conjurations} would reach the same power, but at a cost of 79.13 skill points.
- If you wanted to maximize power on a regular species, you would train low Spellcasting, high <Spell Schools>. XP costs rise as skill levels rise. So a regular species, wanting to train to a higher skill level, would face relatively higher XP cost. This matters less in the early game as going to 10 skill is less expensive than 27 skill.
- In practice, regular species will want to train Spellcasting for spell levels and MP. Djinn have infinite spell levels and high max MP, so they can dedicate more XP into increasing power.
