Sil: What are your least liked features of Sil?

[evilmike]

DCSS does something a bit like this with its "hints mode", which is a sort of step between the tutorial and the real game. Instead of letting you choose your race + class, it gives you a small list of preselected characters which are considered to be strong examples of various playstyles. It's up to the player how to progress in the game beyond that, but various help messages are printed during the early game to help things along.

[half]

Interesting ideas on hinting about builds. I agree that some advice would be useful, but I'm not sure how to do it. I see a couple of challenges:

1) Even realising basic combinations of Abilities that work well together is a fun task that is part of playing the game. In general it is good for games to have some fairly obvious things that go together well to give the player a feeling of progression even at the start. Taking this away has drawbacks.

2) Part of the point of the game is the open character creation as opposed to class based play. I want the game to draw in those people who will like the full game, so I wouldn't want to switch it round too much.

The tutorial already offers some advice, but I can see people would want more (and that is a jack-of-all-trades character which is best not to emulate). Maybe a shortish list of advice would be good. It could mainly focus on guiding players to abilities that are strong (as opposed to merely sounding good), and away from certain traps (such as spreading too thin, or not buying enough abilities). I could either put this in the manual, in a game screen (like the help screens), or in a series of one liner hints that show up when you start the game (like in quite a few other games).

[taptap]

It could mainly focus on guiding players to abilities that are strong (as opposed to merely sounding good), and away from certain traps (such as spreading too thin, or not buying enough abilities).

Buying too many low tier abilities and not specializing at all is probably the biggest obstacle as a beginner (but of course we have seen you can win even so). Another one for me was playing as a dedicated smith which makes the early game harder than it needs to be as does any other commitment that would pay off in endgame strength but doesn't help when you are dead as soon as you see wargs. And yes, figuring this out was part of the fun and I remember the excitement when I saw my first wargs and much later my first cats.

[taptap]

As many people say ID game is their least favourite part of the game, I guess I can put this here. It would be lovely if a staff of self knowledge would not only list the special effects allowing me to ID, but also give your char a roll to recognize them (often you have clear knowledge your weapon is strong against trolls and orcs but still don't recognize it is of Gondolin).

[HallucinationMushroom]

I posted something a little like this before. When you go to inscribe a longsword, it asks you if you want to inscribe all longswords with Gondolin, instead of the property that makes you think it is Gondolin. Like, perhaps there is a rune on it. You know it is special, at the least. Something is different about it. Perhaps the differentness of it is the same for all items of the same differentness. Man, I need to raise my int score. Anyway, I hope you get my idea. You wouldn't get experience, but at least you could sort your loot in an intelligent manner until you get perfect id.

Or, maybe I'm just being silly. Rings and amulets are always the same, so an inscription makes sense to span their particular flavors, but I can't imagine a scenario where you would want to label all longswords a particular way.

Edit: For what it's worth, I don't mind the ID game. If the perception tree weren't chocked so full of goodness, I would take loremaster. But, I'd rather have other nummies.

[tg122]

I really enjoy Sil, but I think the thing that I like least about it is that the Edain are so inaccessible to many players because they are so weak. I know they are supposed to be the challenge race, but from a lore perspective, I find them very intriguing.

Not that I think Edain should be as powerful as Sindar/Naugrim, but I would like to see them start out with maybe 1 point of dex, and 1 point of con (or some other small stat combination) instead of zeroes across the board. This would allow for some pretty versatile builds while making them a little more accessibe to many players, instead of just labelling them as a pure challenge race.

Then, maybe add Easterlings as the pure challenge race with no starting stats.

[MarvinPA]

As many people say ID game is their least favourite part of the game, I guess I can put this here. It would be lovely if a staff of self knowledge would not only list the special effects allowing me to ID, but also give your char a roll to recognize them (often you have clear knowledge your weapon is strong against trolls and orcs but still don't recognize it is of Gondolin).

I've been trying to figure out a good way to handle this - on the one hand you'll often know for sure what quite a few of your equipped items are with self-knowledge, so they should just always fully identify in that case. But just making it always identify everything you're wearing would be a stronger effect than currently since there's ambiguity in a fairly decent number of cases, and the way the effect works makes it tricky to actually have the game calculate when you've unambiguously identified something with it in a bunch of cases. I think the best solution I came up with so far was maybe just to randomly identify a few equipped items (maybe 1-3?).

[taptap]

I've been trying to figure out a good way to handle this - on the one hand you'll often know for sure what quite a few of your equipped items are with self-knowledge, so they should just always fully identify in that case. But just making it always identify everything you're wearing would be a stronger effect than currently since there's ambiguity in a fairly decent number of cases, and the way the effect works makes it tricky to actually have the game calculate when you've unambiguously identified something with it in a bunch of cases. I think the best solution I came up with so far was maybe just to randomly identify a few equipped items (maybe 1-3?).

You only get a perception roll when you see the effect. (With 0 grace your not so clever char may need time to identify even a Gondolin sword when bashing orcs.) My proposal was merely to give a perception roll (maybe with added difficulty when more items are concerned) when you simulate the permanent effects of your wielded / worn equipment with self knowledge - that would give at least a decent chance to identify (w/ xp) +0 rings / amulets and your only chance to identify adornment (other than direct understanding staff or smithing it). So people playing without loremaster had a chance to gain this bit of experience as well.

[half]

My proposal was merely to give a perception roll (maybe with added difficulty when more items are concerned) when you simulate the permanent effects of your wielded / worn equipment with self knowledge

A disadvantage of this is that I'm currently considering a previous proposal to eliminate the perception rolls for identifying things by use altogether (i.e. to make them always succeed). Adding them back here would be strange. I like the idea otherwise though.

I'm definitely inclined to have self-knowledge auto-id your weapon, as the weapon text explicitly says where the effect is coming from (as opposed to your equipment in general, or possibly even a timed effect).

I like the way that currently neither self-knowledge nor understanding dominate the other. I think self-knowledge is better earlier on for the thoughtful player, while understanding wins out in the end.

Proposals that identify things as soon as you could possibly work it out yourself appeal to people who already play perfectly and then are frustrated that the game doesn't identify an item whose identity they know. However, they actually remove some interesting gameplay for the less skilled or knowledgable people.

Keep up the proposals though. I'm interested.

Also note that the Sil way of doing identification by use is to give you a perception-linked chance (or maybe just a guaranteed success) at identifying it when it produces a clear effect, not just when you know that nothing else could have given the effect. Indeed, your boots of free action are identified even when you are wearing a ring of free action too. I'd ideally like to keep this model, though it does rather suggest auto-iding all worn items when you use a staff of self-knowledge, which is not on the cards. It would be too powerful unless I made the staves much rarer, and I don't want that as I have improved the screen in the dev version to show you exactly how all the bonuses are stacking which is useful for players.

[emulord]

Offtopic for now, but my biggest pet peeve is still the invisible forced descent clock. Can't it be a birth option that decreases score? Even a cheat option would be nice. Id like to only be constrained by the food/light clocks at first, and encouraged to advance with the marginal exp gains.

I couldnt ever get into sil because of this reason. I was forced to dive/die before I figured out AI for things and good builds. I dont mind dieing from my lack of skill, but dont force pseudo-ironman on a new player.

[Patashu]

Offtopic for now, but my biggest pet peeve is still the invisible forced descent clock. Can't it be a birth option that decreases score? Even a cheat option would be nice. Id like to only be constrained by the food/light clocks at first, and encouraged to advance with the marginal exp gains.

I couldnt ever get into sil because of this reason. I was forced to dive/die before I figured out AI for things and good builds. I dont mind dieing from my lack of skill, but dont force pseudo-ironman on a new player.

If there is no descent clock then reloading a level indefinitely will keep getting you more food, more light, etc. allowing you to sustain yourself, and more equipment, more artifacts, more consumables, etc. giving you an incentive to scum forever. We're back in Angband territory.

[Derakon]

If there is no descent clock then reloading a level indefinitely will keep getting you more food, more light, etc. allowing you to sustain yourself, and more equipment, more artifacts, more consumables, etc. giving you an incentive to scum forever. We're back in Angband territory.

Instead of forcing the player to descend by the current method, would it be feasible to simply stop generating light/food at overly-shallow depths? That is:

1) Decide on a total amount of food that the player should have over the course of the game. Call this F.
2) There are 20 levels. Therefore, over the long run, each level should have F / 20 food.
3) Track the amount of food the player has found. When they have found F / 20, newly-generated levels at dungeon level 1 cannot have more food. At 2F / 20, dungeon level 2 stops generating food. Et cetera.

And do something similar for lights.

Functionally what this means is that if you want to hang around at shallow depths, you can, but you'll be forced downwards eventually because you won't have the food or light to continue.

The biggest problem with this approach is that players will not connect the "not enough food/light" problem with the "at too shallow of a depth" problem. This is presumably why Scatha and Half decided to use the solution they did -- I don't pretend that the approach I suggested is especially novel. Sil's current approach is hamfisted but also very hard to misunderstand.

The reason why NetHack works (and IIRC why Rogue works, though I forget its exact mechanics) is because they don't allow replaying of levels. As soon as you allow that, you have to find some more artificial means of pushing the player along, or else accept that players can break the difficulty by playing at an overly-slow pace.

For what it's worth, what I eventually want to experiment with in My Own Variant™ is having an "alarm level" in the dungeon that would increase with time, which in turn determines the difficulty of the monsters that you face, but not the rewards that you get for beating them. So if you hang around at dungeon level 1, the game starts pulling in harder and harder monsters to fight, but you still get level-1 rewards. Getting the balance on this right would be tricky, though.

[Patashu]

For what it's worth, what I eventually want to experiment with in My Own Variant™ is having an "alarm level" in the dungeon that would increase with time, which in turn determines the difficulty of the monsters that you face, but not the rewards that you get for beating them. So if you hang around at dungeon level 1, the game starts pulling in harder and harder monsters to fight, but you still get level-1 rewards. Getting the balance on this right would be tricky, though.

This is mechanically almost identical to a forced descent mechanic.

[taptap]

3) Track the amount of food the player has found.

There are hunger items in the game! And there is sustenance + mind over body. This approach would make hunger impossible and sustenance an opportunity for grinding - an opportunity Angband socialized players wouldn't fail to exploit. In fact, even balanced the game attacking the player by starvation looks way more ham-fisted to me than a somewhat abstract forced descent. So even if balanced, you would lose one interesting mechanic / decision area to make a substitute for forced descent.

And for light: there are several more or less common permanent light sources in the game.

[half]

For what it's worth, what I eventually want to experiment with in My Own Variant™ is having an "alarm level" in the dungeon that would increase with time, which in turn determines the difficulty of the monsters that you face, but not the rewards that you get for beating them. So if you hang around at dungeon level 1, the game starts pulling in harder and harder monsters to fight, but you still get level-1 rewards. Getting the balance on this right would be tricky, though.

Sil actually started with this, but it didn't work so well. One issue is that it needs to be really telegraphed to the user. If they don't understand what is going on, they'll definitely won't react to the game being difficult by diving, but will stay where they are. Even then, unless you actually make it more difficult to be shallower than deeper, they will tend to stay shallow. If you do make it more difficult, then as Patashu says, why not just have forced descent?

Tight food or light clocks have similar issues. You really have to telegraph the game mechanic (which tends not to make any real-world sense, like your food version and thus breaks immersion) and even then players will die by not diving enough. They will also feel it is less heroic to be dying to lack of light and hunger. Light also has troubles with not allowing permanent light, and with the fact that good play would involve frequently turning your light off.