Ashenzari

[galehar]

Somehow, we haven't talked about the new god here

There is yet another post about Ash being overpowered.

Could you elaborate on why do you feel Ash is overpowered? Is it transfer knowledge (aka reskilling)? Or the scrying + tronado combo? Anything else?

I think reskilling does need a little nerf. I'm thinking about changing the XP penalty from 10% to 20% and maybe also slow down the transfer (half rate?).

[lucy_ferre]

- Ridiculous XP boost. I heard it was a total of 40% extra experience (5% per cursed item). My troll ended up getting so much XP per floor that I had to victory dance constantly due to maxing out.

- Cursed piety boost. x6 for complete bondage is extreme. I spammed his abilities all the time. Even x2 piety is enough to use his powers pretty often.

- Transferring skills. Aptitudes and training don't matter much when you can boost a skill that you suck at (and never use) by 15 levels from another skill you can easily raise back up. Example: my TrHu transfered 5 levels of UC (at 20-something skill) for 15 levels of spellcasting.

[dpeg]

I believe people are raging about Ash because of the xp gain boost.

I also believe that this mechanic, while strong, is opaque and too meta-gamey. People have told me that Ash without xp gain boost would be underpowered but I don't believe that and rather see it as a lack of fantasy

Since this is an Ash tread, I might as well post some Ash-y ideas here:

• Prescience: Given Ash's divination aspect, dangerous items of monsters can be known (wands, ego weapons).
• As an extension, we could use the prescience flavour for "You saw that coming and evade it"-type battle effects.
• I believe that a god who gave nothing but permanent scrying was overpowered (others disagree). From this angle, the detection powers could be improved upon. Mostly relevant are monsters, I suggest a three tier indication (at a certain piety level). Full detection, or genus (glyph) is out of the question: there are problems with tiles, and you can/should use scyring when the going gets tough. Secondly, the mapping should become finer at high piety, so that you can actually make out corridors etc.
• Currently, you curse items for better piety gain, but don't need full curse coverage (two jewellery and more than half armour items will do). However, additional curses are useful for the xp gain boost, which we are (I am) trying to get rid off. Hence, additional curses should have some effect. One approach is to divide the piety gain in two tiers: some boost for half-boundedness, more boost for full-boundedness.
  Another approach is to make cursed items behave better/special, ideally in battle. Some ideas that have been floating around: Bolts/etc. deflect (or even reflect) from your cursed shield; same for spells using cursed jewellery, and melee blows using cursed armour. The idea did not gain much praise on ##crawl-dev, but it could be turned into something that works and fits the flavour, in my opinion.

I believe there were other ideas as well, but they escape me right now. Perhaps someone can step in.

I have won a SpAs of Ashenzari with only limited cursing (one level of boundedness), and I think the god is basically functioning. The curses minigame is alright, the passive powers are all useful, and reskilling works well (whatever shortcomings it seems to have are immediate effects of the xp/skills system).

[dpeg]

- Ridiculous XP boost. I heard it was a total of 40% extra experience (5% per cursed item). My troll ended up getting so much XP per floor that I had to victory dance constantly due to maxing out.

I agree. And I want to get rid of that, see above.

- Cursed piety boost. x6 for complete bondage is extreme. I spammed his abilities all the time. Even x2 piety is enough to use his powers pretty often.

Yes, the boosts are too large. They should be toned down.

- Transferring skills. Aptitudes and training don't matter much when you can boost a skill that you suck at (and never use) by 15 levels from another skill you can easily raise back up. Example: my TrHu transfered 5 levels of UC (at 20-something skill) for 15 levels of spellcasting.

I don't agree. Reskilling does nothing you couldn't have done with victory dancing in a (minmaxer speaking) better way: earlier, and without the xp loss.
The TrHu that donates UC will not be using unarmed combat anymore, right? That's okay, but it is a meaningful transfer. In my games, I found it hard to find good skills to take xp from.
Also, aptitudes do matter. What does not matter is training: reskilling allows you to adapt after the fact. This is exactly the strategic flexibility we had in mind. With Ash, you can tune your magical skill set to the books you find etc.

[MrMisterMonkey]

Or the scrying + tronado combo?

This works perfectly fine without Scrying, though is particularly effective with known monster locations (perhaps by antennae, scrying, or knowledge of vaults), and is a problem with tornado's specification (or perhaps just a bug).

On Ashenzari itself, I like to divide reskilling usage into two primary, and more minor third and fourth (overlap, but extreme enough to warrant their own sections) categories:

• Elimination of useless skills (though the destination doesn't matter so much as it's useful), such as getting rid of necromancy for a conversion to The Shining One, shoveling Short Blades into Long Blades, or providing use to nonsense skills from sage cards and manuals (nonsense manuals are useful in ziggurats and other high-exp areas, such as to stay clear of the exp pool cap); I personally consider this sort of thing a no-brainer, provided Ashenzari worship. Choice of destination skill can be interesting, but usually it's terribly easy, and there are skills that are always good (e.g. even for a pure spellcaster, Fighting raises max hp).
• Interesting sacrifice, like unarmed combat to spellcasting, is much more appealing, to me, but I'm afraid it would prove much less common, especially with victory dancing as a cheaper (but more painful) alternative in most cases.
• Complete change in character build, like from stabber to conjurer, or undead-hearding necromancer to paladin; this is interesting when both the source and destination have appeal, but I imagine, and especially so with reskilling's cost, that players will more commonly switch when their builds become useless (e.g. how do you stab with all that hellfire in Gehenna, or use styles only effective against or in the presence of living creatures?), or more generally, go from an easy early/mid game to a more powerful late/extended game.
• Adapting to unexpected change seems to be the most popular example for intended usage, so I won't talk about it much, but I find it usually ends up being a nice reason for interesting sacrifice, and a nice break from the tiresome premeditated nature of other usage.

On reskilling as an interface around victory dancing (at the cost of some skill progress), note that with reskilling the player gets to choose when to lose skill, whereas with victory dancing, they have to live with lost potential skill when it might matter more (and it can matter a lot, e.g. in the case of playstyle switches).

[galehar]

Thanks lucy_ferre and MrMisterMonkey for the interesting feedback.

I think that tornado working through wall is a bug.

I think the XP boost also contributes to the overpowered feel of reskilling. If you're training skills 40% faster, the 10% penalty when reskilling doesn't feel expensive. Let's get rid of that asap.

Good point about reskilling not equivalent to victory dancing. If a troll uses UC for half the game then trade 5 UC levels for 15 spellcasting levels, he survived using the skill before giving them away. Also, if UC is reduced from 25 to 20, that doesn't mean that he has to stop using it. It's still plenty powerful. I'll run some numbers to see if I can find a better value for the penalty (maybe 20% isn't enough either).