0001568: Nemelex Xobeh seems to gift decks faster than you use them up

Notes
(0004850) KiloByte (manager) 2010-05-17 01:18	In several places, you assume waiting 600+ turns between drawing every card, to get _slightly_ more decks than you consume. Except, you forgot that Nemelex' piety decays at a rate of 1 per 700 standard turns. This makes piety using decks just a flavourful addition and an incentive to use the decks rather than a perpetuum mobile.

(0004851) pixnaps (reporter) 2010-05-17 02:59	I think Luca just meant to wait 600 turns after being gifted a new deck, after which time you can draw new cards in quick succession -- wouldn't that offset the 1 piety loss? (That said, I like the flavour of piety/rewards from using cards; but if Luca's calculations are right then it might at least need some tweaking...)

(0004852) Luca (reporter) 2010-05-17 03:44	Yes, the strategy is exactly the one described by pixnaps. Simulations appear to show that the number of decks in your possession tends to infinity, and piety generally stays above 190 (having a very small effect on gifts). Using an amulet of faith speeds this up even more (but you can't then wear stasis or resist mutation, which may be useful at least for some decks).

(0004853) syllogism (updater) 2010-05-17 08:36 edited on: 2010-05-17 09:26	I've utilized this strategy, for most of the game, and I did eventually run out of decks. That could have just been variance, however. Something should probably be done about it regardless as this allows you to get exactly the types of decks you want.

(0004856) Luca (reporter) 2010-05-17 14:20 edited on: 2010-05-17 14:21	Was that a local game? I can't find any very recent morgue by you. You likely need at least 5-10 decks when you reach full piety to have a good chance of not running out, and also must avoid wasting cards (draw on an unoccupied square with a summoned monster in sight that you attacked to make it hostile) and should avoid using any Nemelex ability except Draw One. When playing normally, sacrificing items can probably compensate for not doing the above.

(0004858) syllogism (updater) 2010-05-17 15:09 edited on: 2010-05-17 15:32	Not a local game and I didn't intend to suggest the strategy does not work; indeed, the experiment convinced me it does and I mentioned it on crawl-dev. Can't find the specific log as I've, to some extent, abused the mechanic in all my 'recent' nemelex games.

(0004861) Cryptic (developer) 2010-05-17 15:53	Lots of gods have this issue. You can scum oka gifts by waiting Note: 0000600 turns each time he gifts something, and withing ~50 piety events, he'll likely gift you again. It would seem to me that gift timeout should only decrease under certain circumstances...such as piety events.

(0004863) syllogism (updater) 2010-05-17 15:56 edited on: 2010-05-17 15:58	It's not the same situation at all. Okawaru piety decays faster and the only way to keep piety up is to play him normally.

(0004864) pixnaps (reporter) 2010-05-17 15:59 edited on: 2010-05-17 16:00	I recall a proposal somewhere to use an exploration-based timer for these sorts of things. Would that solve the problem?

(0004865) Cryptic (developer) 2010-05-17 16:04	Gameable gift timeout is problematic for numerous gods, decay rate doesn't change the fact that both are easily gameable following a fairly straightforward method of waiting x turns after gifts. Keeping oka piety up is actually really essy, even while making use of might and haste. Pix, do you recall the specifics of that proposal, or do you have a link?

(0004866) KiloByte (manager) 2010-05-17 16:07	For all gift-givers, decaying the timeout only for actual piety gains rather than time would be a good idea. But for Nemelex, I think piety for cards should stay -- otherwise, there is no reason to ever use decks in an unsafe manner. We already have two mostly redundant counters -- just tuning them to remove any benefits for waiting should do the trick.

(0004867) pixnaps (reporter) 2010-05-17 16:14	https://crawl.develz.org/wiki/doku.php?id=dcss:brainstorm:misc:time#exploration_timer [^]

(0004868) Cryptic (developer) 2010-05-17 16:34	Yeah, exploration timer would just lead to scumming gifts before exploring branches, instead of arbitrarily by waiting. Gift-giving timeouts working off piety gains less gamey, I think.

(0004870) dpeg (administrator) 2010-05-17 17:02	Back when Haran and I designed the new Nemelex, we explicitly wanted to shift the piety gain from sacrifice to card use. (Actually, we phrased it like this: you shouldn't be able to get piety higher than some threshold without actually using cards. In other words, only saccing and stacking decks should not work.) This does not imply that sacrificing alone should keep you getting decks infinitely... It is good to see Nemelex feedback. We were (and are) aware that the god is not perfect. This includes mechanics (like the gifts) as well as abilities, deck composition and card effects. For the case at hand, what about (besides changing numbers, which is probably warranted anyway) making piety gain from drawing cards depend on deck rarity and piety itself? In other words, drawing from a common deck at low piety would increase piety like now; drawing from a rare deck would yield more, and drawing from a legendary deck even more. At very high piety, drawing from a common deck would yield very little piety... re exploration timer: They're good for many things, but not so much for gifts, in my opinion. Both waiting out some duration and optimising exploration can be optimised. The real problems are (a) repeated gifts and (b) not using the appropriate scale (piety) well enough. On (a), I hope that Nemelex can circumvent this problem by demanding use of the decks.

(0004872) KiloByte (manager) 2010-05-17 18:05	dpeg's last comment is spot on -- both about exploration timer not being good here (even if it's a good idea for other things), and about requiring both sources of piety. My solution, not reducing timeouts with time but solely with sacrifices and cards use, can probably work but dpeg's suggestion is definitely more flavourful.

(0004880) Luca (reporter) 2010-05-17 21:06	Demanding use of decks has the problem that it encourages using them in an useless fashion just to game the gift system, after taking precautions to avoid bad effects. And it seems quite hard to automatically determine whether the card when used in an useful/risky fashion or not. There is the "tension" system, but I doubt it can work properly for advanced characters that can just Controlled Blink away from a supposedly dangerous situation. Also anything decreasing with time obviously encourages uselessly wasting turns. The only robust way of doing this in my opinion is to tie it to some resource that the game is already careful to not give out for free (which are to some extent all equivalent to useful progress in the game or at least interesting gameplay): - XP - Skills trained via XP - Explored terrain - Gold found (not spent) - Items found Of course all of these can be easily obtained by wandering indefinitely in Pandemonium, but that requires having essentially already won and is no worse than anything else in the game. In addition, it should be something that the god likes, and counting only what happens while worshiping that god. For instance, Sif Muna rewards training spell skills, which is limited by XP. An option could be to have Nemelex reward training Evocations, which would basically reward drawing cards with XP in the pool (very frequent if you actually use them in combat), achieving both the flavor requirement of deck usage and the robustness requirement of requiring XP. The only problem is that you would get no more decks at Evocations 27, but I think this can be fixed by simply letting the Evocations skills go arbitrarily high, and making values higher than 27 count as 27 for all purposes (perhaps even display to the user). Or equivalently, gobbling up XP as if Evocations was trained even when it is at 27. The only user-visible difference would be that with Nemelex, you would still be able to enable/disable Evocations when mastered. As for sacrifices, is there some flavor reason that a god of cards should take item and even corpse sacrifices? I think they could be totally removed, and you could still get decks if you run out by doing anything else using Evocations with xp in the pool. It is also rather tedious to, say, sacrifice a whole Ziggurat floor while making sure to only sacrifice the right type of items to get the decks you want the most. For instance, if you just cleared an orc floor and want decks of defence but not destruction (the latter are useless because Fire Storm is better, while the former have Tomb), you will have to visit each of the around 50 squares an orc died on, pick up the weapons, sacrifice the armor, and then drop the weapons (or alternatively apport all the weapons to one square, and all armors to another, also not very fun). It even encroaches on Jiyva's item sacrifice theme, which is instead at least not tedious since it happens automatically and you can just sacrifice anything indiscriminately. Maybe Nemelex could be even expanded to be a general "miscellaneous items"/Evocations god instead limiting it to consumable decks. Note that this would encourage uselessly evoking a staff of channeling or similar to train Evocations and get deck gifts, but this would be limited by the XP pool and a general solution for skills would solve this too. As for which decks to give, how about having a fixed chance of giving a random deck, and a fixed chance of making a weighted random choice, based on the decks you drew the last 50 or so cards from? (or perhaps with an infinite exponentially weighted mean, or something like that) Another consideration is that perhaps all the non-tactical cards such as Sage, Experience, Shuffle, Focus, Trowel, Helix and Genie should be dropped since they seem to be primarily useful for scumming purposes, or for partially or totally breaking the game in various ways. Without these, even infinite decks would not be exploitable so badly as now, where you can (theoretically) maximize your XP and all your skills with Sage+Experience, get arbitrarily rich and get an infinite supply of randarts with Genie, and get a great mutation set with Helix. The overpowered Tomb, Elixir and Portal might also benefit from some nerfing.

(0004886) syllogism (updater) 2010-05-18 08:32	Nemelex doesn't gift decks with Elixir anyway

(0004901) dpeg (administrator) 2010-05-18 21:10	I am very interested in feedback by players who have played Nemelex quite far. Please reply :) Luca: did you win a Nemelexite? (Serious question!) When we re-designed Nemelex, we were willing to sacrifice some aspects of the old trickster god, but not all of them... Continuity is a good thing. So the portable altar is no more (which hurts quite a bit in the flavour department), and the god is much more about using cards in tactical situations than before. The availability of strategical cards is also inherited (you could say that old Nemelex was essentially only about this). Now, there's no principle reason to keep strategical effects if they harm the game, but my take on this is a bit different: Nemelex is a god which completely dominates your game (in contrast to most gods, who only complement a game). Also, Nemelex is a very broad god: you can make good use out of the god, almost regardless of species, background or build/style. Therefore, having very strong as well as long term (strategical) cards seems like an appropriate reward. I am now addressing your points one by one: (1) Decks will be spammed, not used properly. It's true that abuse prevention is hard to do, but we have a bit of that: you don't want to use combat cards alone etc. Much of this could be improved upon, of course. A more interesting reply is my former proposal (which you didn't reply to): piety gain from drawing cards should depend on deck rarity (positively) and current piety (negatively). In other words, spamming cheap decks wouldn't help you improve your piety much, beyond some point (much like sacrificing). And to combine these two ideas: while it is hard to nail down card spamming, it is easy to reward drawing from decks (especially rare and legendary ones) under high tension. If decks need more double-edged cards, then so be it. (2) Using a timer is bad. I agree. Having a fixed duration after a gift only encourages complaints as by the OP. Yes, there is piety decay during the duration, but wouldn't it be much easier to just reduce piety by that amount (over the 600 or so turns) at once, when getting a gift? (3) Shifting to xp/skills rather than piety. This is not good, in my opinion. We'd have to change a basic mechanic (if allowing Evocations training beyond 27 for Nemelex piety purposes), and piety gain can be linked to finite resources in much more interesting ways. For example, I still support a model like this: * only sacrificing can only get your piety up to somewhere (say 100) * likewise, drawing cards alone will only get your piety up to that limit * the higher your piety, the harder it is to get more: draw from rarer decks, high tension helping sacrifice better items * each of these methods should have diminishing returns The last point means that you'll want to sacrifice (good!) stuff, thereby disabling the deck perpetuum mobile: you will have to go somewhere, so that you find good items to sacrifice. The diminishing returns mechanic should set in right away (from piety 0). (4) Why sacrifices? First, they were the defining property of old Nemelex. Second, items are a finite resouce. Third, I really like the idea of steering the gifts by choosing what you sacrifice. I don't see a change of this mechanic soon. It is possible to improve the interface, of course. For example, I believe that players should be able to select item categories in the ^! screen (which explains which items go towards which decks). Prayers would then only sacrifice items of the selected types. It is also conceivable that for Nemelexites, autoexplore moves towards items of the selected types. (5) Remove cards. The decks and cards are by no means in a finished state. However, I have learned (from Erik) that allowing some aspects of endless play can be good for the game. I believe that Nemelex does this in a rather positive way (via decks of Wonders). If you have ideas for new card effects, or how to modify existing cards, please tell us. (This goes to everybody who has an interest in Nemelex.) It is also an option to come up with new decks. (Two good souls once designed and partially implemented a deck of Night, revolving around stealth and stabbing. Unfortunately, they moved on before the patch was finished.)

(0004905) pixnaps (reporter) 2010-05-18 21:40 edited on: 2010-05-18 21:45	I like dpeg's suggestions. (I've won with a Nemelexite SpAs, and a second got killed by Daevas in Zig-11. It's my favourite character type to play.) Is extra piety currently given for drawing from unidentified decks? That might be a good tension-booster, especially when combined with dpeg's proposal to weight piety gains by deck rarity. (Though I'd like to hear more about how "high tension" situations are measured -- does this mean something like low HP, lots of high-level monsters in LOS, etc.?) The 'Deck of Night' sounds really fun. Maybe sacrificing blades and light armour (cloaks, robes) could steer gifts this way, leaving heavy armour for Escape and maces/axes/etc. for decks of Destruction? (The latter I never use except to spam piety anyway. I'm also not such a fan of the current non-gifted decks of defence and war; but maybe that's just me.)

(0004914) Luca (reporter) 2010-05-19 02:23	I did a lategame conversion to Nemelex (during which I noticed this issue), but haven't played it from the start. Limiting gain for spamming common decks seems helpful. Note however that it will still be good to spam them if there is any gain and you don't need them (why not?) and it is still useful to spam ornate/legendary decks, especially of types you don't need in real combat. The current "anti-spam" system for decks of destruction can be trivially worked around by blowing the horn of Geryon to get an hostile beast near you before drawing cards (I suppose Summon Small Mammals, possibly followed by attacking it to make it hostile also works at XL 1). This also makes Banishment harmless. Ignoring summons and using tension can help in early game, but I think it's really hard to estimate tension correctly for middle/late-game characters. For instance, an extremely powerful melee monster is no threat to a spellcaster with Swift and Flight that can simply prevent it getting close, an Ice Fiend is no serious threat to someone with good defenses, rC+++ and in lich form, and so on. You would at least need a complex system accounting for player resistances, skills, obtainable speed, blink/teleport capabilites, and so on to make this non-abusable, and I'm not sure it is actually possible. As far as I know, the current tension system is more geared towards making it less likely for Xom to screw you up even more in difficult situations, and this is much harder to exploit and it's not very useful to do so anyway. A similar mechanic would be Nemelex replacing bad or less useful cards with more useful ones in case tension in high (not sure if it is a good idea). The piety decay is currently negligible (with the exploit strategy, piety doesn't seem to drop below 190), especially because drawing cards returns piety to full almost immediately. Note that the problem is not that getting high piety is too easy, but that getting piety increases is too easy, since currently piety increases drive gifts. In particular, the fact that the gifts themselves can be used to get piety increases and thus more gifts causes the gift loop issue. With such a diminishing-rewards mechanism, one could conceivably let piety drop so that increases and thus gifts become more frequent. The fact that high piety is needed for rarer decks and increased gift chance makes this less desirable though. Still, it creates a mechanic where you can choose, by manipulating piety, between more gifts of low powered decks (and it will happen normally at first) and rarer gifts of high powered decks (which will happen normally later): not sure if this is desirable. Note that piety can be trivially decreased by wearing and removing an amulet of faith. Items are indeed a finite resource (except for Pandemonium and god gifts), and that part mostly works. The idea of selecting sacrifice types would be a great improvement. There is still the problem that if you convert to Nemelex late, it is optimal to sacrifice everything in the dungeon, which is boring to do. A "Sacrifice Dungeon" option upon converting to Nemelex acting on all squares not marked with !p would work, but seems somewhat weird. Note however that the Trowel card might allow to access portal vaults with sufficient loot to expect to get more than one Trowel card, leading to another infinite loop. Item sacrifices also seem to directly contradict the rationale for not letting players sell items to shops in the Crawl help, since they are essentially the same thing (except you get paid in piety and cards instead of gold). As for steering gifts, how about steering it depending on which decks you actually use? That seems more intuitive and useful, does not need instructions, and also works correctly automatically. Of course, some other resource needs to be used to avoid infinite decks (e.g. XP as I proposed, or item sacrifices regardless of type). It's not clear to me why decks of wonder allow endless play. It seems to me they facilitate either getting more experience and skill early with Sage/Experience (which is a unique effect and risks to be game breaking) and good mutations with Helix, or doing the same late-game with the intent of rapidly producing a character with all skills maximized (see for instance the recent Naga all-rune wins with all skills maximized, hyperbolic in particular). Focus and Shuffle allow to manipulate stats in strange ways (e.g. ultra-intelligent trolls), which I'm not sure if it is good or not, but still is a "scumming" kind of Nemelex usage (i.e. infrequent, unrelated to tactical strategy and encouraging to take Nemelex just for this). Drawing Shuffle accidentally might also totally ruin your game. Also, Jiyva has similar effects to Focus, Shuffle and Helix, but I think with much more interesting tradeoffs for them. The Potion card is only useful tactically or perhaps to cure mutations if desperate, and Wild Magic is bad. Decks of dungeons have Trowel which indeed allows to visit a lot of portal vaults. Pandemonium also provides this and already has an 8% chance of spawning a Ziggurat on each level. Adding other interesting portal vaults there may be a better system than relying on Trowel, since it cannot break the early/middle game, and is accessible to all types of characters. Only Bazaars are likely to be interesting though, since Ziggurats already provide the most challenge and loot (and it seems a bad idea to, say, let a player see all wizlabs in a single game, to preserve variety). At any rate, the current situation seems possible to minimally fix: just remove the decks countdown (or remove its decrease with time), and increase the gift timeout (which does not decrease with time) to a value good enough to prevent this by a good margin even with an amulet of faith on, and having on average a similar effect on non-exploitive play as the current system. Decks will still be spammed, but at least you will get much less decks than the one you are consuming, so you will do this only for decks you don't need. Also make loot generated in portal vaults created with Trowel be worthless for sacrificing purposes (unless there is a large gold cost to enter the portal) and make sure that bazaars don't allow you to multiply gold by buying scrolls of acquirement for less than the amount of gold you can acquire.

(0004915) nicooo (reporter) 2010-05-19 02:57	How about giving players an incentive to not spam decks of destruction? Like in https://crawl.develz.org/wiki/doku.php?id=dcss:feedback:god:nemelex [^]

(0004932) Kyrris (reporter) 2010-05-20 02:02	I'm almost certain that Experience cards don't funnel into Sage. Experience potions certainly don't.

(0004940) Luca (reporter) 2010-05-20 13:06	Right, with &c in wizard mode it doesn't: good. You can still victory dance the filled XP pool from Experience though (and you gain levels).

(0005023) Ryak (reporter) 2010-05-24 19:49	I have played 1 winning game with Nemelex and he is quite possibly my favorite current god. Just make sure you don't throw out the baby with the bathwater. I already don't like the gift timeout because it penalizes the unspoiled player who sacrifices or uses cards during the timeout. Is it possible to have "gift likelihood" build up over time with the caveat that no matter what he will always wait a certain amount of time between gifts? If you did this then you could safely decrease the "gift likelihood" buildup so that the unspoiled player gets gifts at about the same rate they were before, but the person who was exploiting the mechanic actually has to do more since there's now no benefit to waiting around before using cards/sacrificing. In this situation, there would be no point in spamming wait to wait out the timeout. And as a bonus, it encourages (or no longer discourages) playing the intended way.

(0006389) JoachimSchipper (reporter) 2010-07-18 14:41	(Partly in response to dpeg's 0004901.) Couldn't this be fixed by switching to two faith pools, one dependent on deck usage (and possibly other "gambling" stuff like drinking from sparkling fountains) and one dependent on item sacrifice? Total faith should simply be the sum of both pools, and both could decay. This should at least make it hard to get an infinite number of the good decks...

(0014701) Mental Mouse (reporter) 2011-09-18 15:58 edited on: 2011-09-18 17:48	Depending on Nemelex cards for your dungeoneering needs is plenty of gambling for any character! If you want to cut down on deck payoff, then just lengthen the deck timeout a tad, or reduce the odds of higher decks according to piety, no need to completely rejigger the system! JoachimSkipper: We already do have two pools, counting the "deck timer" which gets reduced by deck use. Ryak's idea is good, too-- I was looking at a more random version, but repeated small varying chances would probably be too random! Sacrificing during gift timeout does get you piety, anyway.

(0016558) nht (reporter) 2012-01-06 23:36	This 0.6 issue is still a problem in 0.10-trunk. Playing VpMo of Nemelex (with an amulet of faith) and sacrificing purely towards Wonders and Dungeons decks (Wonders for Experience cards, Dungeons for portal vaults from Trowel cards). Converted to Nemelex, sacrificed some junk towards Wonders. Once I had obtained a few decks, stood on a cleared level manually drawing cards, then waiting 400 turns, then drawing again. By this method I reached endgame xl/skills before Lair, see dumplog/ttyrec: http://crawl.develz.org/morgues/trunk/nht/morgue-nht-20120106-174602.txt [^] http://crawl.develz.org/ttyrecs/nht/2012-01-06.16:04:12.ttyrec [^] (possibly add a .bz2 at the end of the ttyrec url) This is an extreme way of playing Nemelex - you can see that I was near-stat-dead by the end due to repeated Focus cards. Most Nemelex games still abuse infinite decks in a similar manner, though not so early in the game. A lot of good work has been done balancing the individual cards and decks since Luca's original submission. As I see it two issues remain, the problem cards (Experience, Trowel) and being able to obtain infinite decks with very little danger. The principle dangers being abyssed by Wild Magic or drawing a nasty Wrath card, both being rare. First the problem cards Experience - Too common. Sage card gives a similar xp boost in a more tactically interesting way. I suggest making Experience rarer, perhaps as rare as Shuffle. Power 2 Trowel - Occasional free portal vaults is fine, this becomes an issue if you're with Nemelex for a long time. I'm not sure how many portals of each type you get now, if 0004411 is current you get at least 35 bazaars. Capping at one of each portal vault wouldn't be unreasonable (and use power 0/1 Trowel after you run out). Re: infinite decks - I like the suggestion to have Nemelex reward using decks under tension, even so far as giving zero piety for deck use if there are no enemies around. This addresses the particular abuse of sitting somewhere safe drawing cards forever.

(0016566) The Mantis (reporter) 2012-01-07 14:56 edited on: 2012-01-07 15:34	The situation's different from 0.6, but the above comments remain, and the reason why this stopped getting a bunch of comments is because it's difficult to solve for little benefit. Rewarding the use of decks only under tension is scummable, and the last thing that Wonders needs is a reduction in the likelihood of acquiring Experience, which is its redeeming value. The problem above is not with Nemelex Xobeh, but rather with Vp's food clock and the lack of punishment for simply spending time in a level, and moving on. To counter nht's statement, I want to point out that I ran an XL27 VpCK of Xom and was able to farm a whole late level's worth of Experience at maximum SL, which I judge is roughly the same risk/reward as this. You can do whatever the hell you like when your food clock is 'drain a corpse, cast Regeneration'. And in addition I don't really think that wearing an amulet of faith as a Monk with no food clock is indicative of typical play. Nemelex Xobeh drains your gold or food and summons hostile 1s, or banishes you, in exchange for really very little. The risk is high, and the reward is quite low eventually. The same goes for bazaars, since after all Nemelex Xobeh doesn't reward you with gold (quite the opposite). The biggest benefit can be gained from a high Evocations race that somehow gets an amulet of faith early, preferrably a Monk. Once you get past a certain point, you can grab quite a few skill levels. Unfortunately, you lose the benefits of worshipping Nemelex Xobeh normally during that period. It's kinda' like Ashenzari's EXP bonus, but in a lump sum. Inferior later, superior early. The best you could do is divide it amongst UC and Evocations so that you have good offense and get higher bonuses more quickly, but every turn you scum you lose score, and since the returns diminish so fast, it's really a silly idea and a silly issue. The benefit vanishes so fast after normal play. Barring dumping it all into UC with good Evocations or wasting a ton of turns. Focus gives counterbalance and then some to Wonders, crippling you over time unless you have high, average stats, like 14\|14\|14. FTFY, Triple Draw means piety loss, which brings you closer to the nadir where you have a choice between sacrificing more wands, scrolls, food or books. Unless you're a VpMo of Nemelex Xobeh with an Amulet of Faith, in which case 'junk' includes ALL THE FOOD IN THE GAME. And yeah, if you rest out a gift timeout, that does do it. You know who also has gift timeouts? Trog. Trog gives weapons, too! Also you can reduce the rest of his timeout by bashing things very hard! You know what bashes things really hard? Weapons! In case you're not entirely sure, try sacrificing only Wonders without an amulet of faith. Any character. Go on. The good ideas aren't tension, they're increasing piety from legendary decks and penalizing keeping a billion decks in inventory. Lastly, way back in 0.6, Luca started this off. Scroll down - you'll see he converted pretty late, and that's the trick. If a god relies on you sacrificing items, and you wait until late, you'll sure get a hell of a boost. And I hate to notice it, but did this guy say Elixir and Portal cards? Those come from the Deck of Escape and the virtual deck Enchantments, so I don't really know how he's getting his facts about all this, bar codediving. And yeah, I want the Deck of Night so hard. x)

(0016567) jeanjacques (reporter) 2012-01-07 18:12	One way to tackle the most severe problem with decks of wonders is to not allow a situation where you only have gift weight for one specific deck. In particular every deck should have some base gift weight. Maybe 10% each or 15% for the three "cheap" types (escape, summonings, destruction) and 5% for the other two. That reduces the benefit of the heavily spoiled behaviour of turning deck types off and starting out with only wonders. The "food clock" is simply too trivial to deal with to prevent this type of abuse. Even sacrificing all permafood but a couple of rations it is very hard to run out of food if you plan ahead and continue exploration between the wonder deck usage.

(0016569) nht (reporter) 2012-01-07 19:17 edited on: 2012-01-07 19:18	"In case you're not entirely sure, try sacrificing only Wonders without an amulet of faith. Any character. Go on" TrMo (max food clock, partly offset by racial gourmand) of Nemelex, no amulet of faith this time. I start worshipping at turn 9474. Explore and clear levels only when I have no chunks left, always wait 400 turns before drawing after a gift. Sacrificing items only towards Wonders. The decks come slowly at first (always Wonders decks), many times I am without a deck. By turn 30000 I always have at least one deck, by turn 40000 deck generation appears to be infinite (6 wonders decks available). Thanks to all the Experience cards I am XL27 by the time I've cleared Lair, Orc and down to D:19. Thanks to Focus cards I have a natural intelligence of 4 (I've been adding all my stat bonuses on level up to the lowest stat). At this point I've hit XL27, can triple draw out the remaining Wonders decks and switch to Jiyva to fix the low intelligence. Dumplog, ttyrec http://pastebin.com/raw.php?i=wP2puaJs [^] http://crawl.develz.org/ttyrecs/nht/2012-01-07.14:45:30.ttyrec [^] (possibly add a .bz2 at the end of the ttyrec url) I'm not suggesting this is optimal play, or that it is fun to play this way (unless you like playing games in cheat mode). It clearly demonstrates the need to make Experience cards either weaker or rarer. So this would make Wonders decks less attractive, good! At the moment, to quote the learndb, these should be triple drawn somewhere safe, presumably for Experience cards. This, the optimal strategy, is not interesting gameplay. It would be better make Sage cards be more common, these give a more reasonable xp boost and you have to killdudes to get it (and you don't get to pick the skill, which is more flavourful for Nemelex). Something along the lines of jeanjacques suggestion would be good. Didn't realise you could exclusively weight a single deck type by only sacrificing towards it (until I tried with this troll).

(0016570) The Mantis (reporter) 2012-01-07 19:20 edited on: 2012-01-07 19:25	The key issue is most likely starting with 60 piety and starting late. I favor the other ways of reparation, as mentioned above. ** piety with one deck is silly when one piety is sacrificing one item. It's probably the Monk bit more than the Nemelex Xobeh bit - that's why there's not been interest 'til 0.10... I support indirectly editing the way that piety is gained, more than anything. More piety for high rarity decks at high piety, almost no piety for low rarity.

(0016583) nht (reporter) 2012-01-08 16:37	Perhaps there is a balance problem with the monk starting piety, I don't find it excessive. Though it does allow some interesting tricks, like having insta-heroism with Okawaru. Of course a monk is strictly worse than a transmuter in the long term, so there should be some incentive to pick Mo over Tm. I don't like the solution of piety gain as a function of deck quality. If you're sacrificing a dungeon full of junk of all kinds, deck gifts are mostly ornate and legendary anyway (the plain ones are the rare ones). And why would Nemelex gift you lower quality decks if she doesn't want you to use them? The edge case we need to fix - the Lichform/Vampire/Mummy player sitting somewhere without spawns (eg. a portal vault), drawing cards forever, or at least in between getting banished by some card effect. In the infinite turn limit you get infinite Experience cards (max skills) and infinite Genie cards (free acquirement). Even nethack limits you to 120 genie wishes! As dpeg says, it would be a shame to lose the strategic cards completely, so we need to make the rate of deck gifting negative for the player sitting around drawing cards. Linking deck gifting to exploration or tension would fix this. Of course both are scummable, but such scumming is more dangerous than hitting 5 four times, then v until you get a gift. Creating a piety/rarity minigame fixes the problem in a roundabout way - but it's creating unnecessary complexity to fix an edge case.

(0016584) The Mantis (reporter) 2012-01-08 16:51 edited on: 2019-06-23 21:45	It seems as though ultimately we disagree. You seem to miss the point that getting sixty piety without weighting decks to anything but Wonders is extremely hard, perhaps impossible without having the whole game's supply of wands, scrolls and food. The more powerful a deck is, the more dangerous it is. Nemelex Xobeh gifts decks, but rewards trusting in their power. The more powerful the deck, the better. Even at maximum piety, you're likely to get a fair few plain decks. Those decks just won't offer you much piety any more - that piety goes to the legendary decks, meaning that eventually you'll be forced to sacrifice items (you ignored the point about having multiple unused decks being a negative conduct for him). If you want to get loads of great items and experience as Mummy or character with Lichform in the end-game, why not try the Abyss? Because you seem to have missed the whole 'fifty percent chance of rotting or stat drain' that Genie offers, which seems to lead to my next question: have you tried doing this? Because it seems highly unlikely from my perspective, and I have played hundreds of Nemelex Xobeh games. You seem to be fixed on a single solution, and I accept you believe it is a good solution. If that's the case, this is not the place for it; why don't you draw up a proposal on the developer's community? What you're describing now is inevitably going to result in Wonders decks getting removed from the standard rotation then removed entirely, and probably going to result in the entire deck acquirement system being nudged progressively towards being glorified Invocations with no profit from using cards, can't you see that?

(0017058) KiloByte (manager) 2012-02-16 14:41	In 0.10, waiting doesn't decrease the timeout anymore. You need to use decks or sacrifice, and the former is not enough by itself. Let's see how it fares on the tournament before making any further changes.

(0017113) XuaXua (reporter) 2012-02-20 22:05 edited on: 2012-02-20 22:07	I just played a 5* piety felid who had only a deck of escape on him with 3 cards; sacrificed everything on lair 1-3 and d8 and no new decks. Trunk 0.11-a0-175 I did use a Mark 5 on the deck of escape prior to not receiving any more decks. It had 3 cards in it (swap, swap, warpright) and as my only defense (no spellbooks yet!) I didn't draw them for a while. Still, I expected to see at least one more deck.

(0018024) KiloByte (manager) 2012-05-09 14:58	While you get quite a bit less decks, it's not totally out of whack anymore. Marking as resolved, further rebalancing might be needed but at least there's no abuse possible, and the god is playable.

Issue History
Date Modified	Username	Field	Change
2010-05-17 00:56	Luca	New Issue
2010-05-17 01:18	KiloByte	Note Added: 0004850
2010-05-17 02:59	pixnaps	Note Added: 0004851
2010-05-17 03:44	Luca	Note Added: 0004852
2010-05-17 08:36	syllogism	Note Added: 0004853
2010-05-17 09:25	syllogism	Note Edited: 0004853
2010-05-17 09:26	syllogism	Note Edited: 0004853
2010-05-17 14:20	Luca	Note Added: 0004856
2010-05-17 14:20	Luca	Note Edited: 0004856
2010-05-17 14:21	Luca	Note Edited: 0004856
2010-05-17 15:09	syllogism	Note Added: 0004858
2010-05-17 15:32	syllogism	Note Edited: 0004858
2010-05-17 15:53	Cryptic	Note Added: 0004861
2010-05-17 15:56	syllogism	Note Added: 0004863
2010-05-17 15:58	syllogism	Note Edited: 0004863
2010-05-17 15:59	pixnaps	Note Added: 0004864
2010-05-17 16:00	pixnaps	Note Edited: 0004864
2010-05-17 16:04	Cryptic	Note Added: 0004865
2010-05-17 16:07	KiloByte	Note Added: 0004866
2010-05-17 16:14	pixnaps	Note Added: 0004867
2010-05-17 16:34	Cryptic	Note Added: 0004868
2010-05-17 17:02	dpeg	Note Added: 0004870
2010-05-17 18:05	KiloByte	Note Added: 0004872
2010-05-17 21:06	Luca	Note Added: 0004880
2010-05-18 08:32	syllogism	Note Added: 0004886
2010-05-18 21:10	dpeg	Note Added: 0004901
2010-05-18 21:40	pixnaps	Note Added: 0004905
2010-05-18 21:45	pixnaps	Note Edited: 0004905
2010-05-19 02:23	Luca	Note Added: 0004914
2010-05-19 02:57	nicooo	Note Added: 0004915
2010-05-20 02:02	Kyrris	Note Added: 0004932
2010-05-20 13:06	Luca	Note Added: 0004940
2010-05-24 19:49	Ryak	Note Added: 0005023
2010-07-18 14:41	JoachimSchipper	Note Added: 0006389
2011-09-18 15:58	Mental Mouse	Note Added: 0014701
2011-09-18 17:48	Mental Mouse	Note Edited: 0014701
2012-01-06 23:36	nht	Note Added: 0016558
2012-01-07 14:56	The Mantis	Note Added: 0016566
2012-01-07 15:01	The Mantis	Note Edited: 0016566
2012-01-07 15:04	The Mantis	Note Edited: 0016566
2012-01-07 15:08	The Mantis	Note Edited: 0016566
2012-01-07 15:09	The Mantis	Note Edited: 0016566
2012-01-07 15:11	The Mantis	Note Edited: 0016566
2012-01-07 15:13	The Mantis	Note Edited: 0016566
2012-01-07 15:21	The Mantis	Note Edited: 0016566
2012-01-07 15:26	The Mantis	Note Edited: 0016566
2012-01-07 15:27	The Mantis	Note Edited: 0016566
2012-01-07 15:29	The Mantis	Note Edited: 0016566
2012-01-07 15:30	The Mantis	Note Edited: 0016566
2012-01-07 15:31	The Mantis	Note Edited: 0016566
2012-01-07 15:34	The Mantis	Note Edited: 0016566
2012-01-07 18:12	jeanjacques	Note Added: 0016567
2012-01-07 19:17	nht	Note Added: 0016569
2012-01-07 19:18	nht	Note Edited: 0016569
2012-01-07 19:20	The Mantis	Note Added: 0016570
2012-01-07 19:22	The Mantis	Note Edited: 0016570
2012-01-07 19:24	The Mantis	Note Edited: 0016570
2012-01-07 19:25	The Mantis	Note Edited: 0016570
2012-01-08 16:37	nht	Note Added: 0016583
2012-01-08 16:51	The Mantis	Note Added: 0016584
2012-01-08 16:53	The Mantis	Note Edited: 0016584
2012-02-16 14:41	KiloByte	Note Added: 0017058
2012-02-20 22:05	XuaXua	Note Added: 0017113
2012-02-20 22:05	XuaXua	Note Edited: 0017113
2012-02-20 22:07	XuaXua	Note Edited: 0017113
2012-05-09 14:58	KiloByte	Note Added: 0018024
2012-05-09 14:58	KiloByte	Status	new => resolved
2012-05-09 14:58	KiloByte	Fixed in Branch	=> 0.11 development branch
2012-05-09 14:58	KiloByte	Resolution	open => done
2012-05-09 14:58	KiloByte	Assigned To	=> KiloByte
2019-06-23 21:45	The Mantis	Note Edited: 0016584

Relationships