Twilight Forums

Many things in the city of twilight are still unknown. In fact, there is more of what we don't know than what we do.
I'd like to propose a somewhat coordinated spading effort to start discovering the specific data behind twilight.
For one thing, if anyone is spading anything specific they can post it here. People can then suggest how to do that properly and faster. But also, I'd like to propose specific projects for spading, and see if people are willing to participate.

Item Drop Effects
The first spading project I'd like to propose is spading item drop bonuses of the effects we are not certain of. You can find the full list of effects here. The effects of Amazing Technicolor Dreampants Pants, Clockwork kangaroo,
Crow, Echolocation, Gumshoe, Monkey and Nicely Aged are unknown.

How does one spade item drop bonuses you might ask? Well the idea I came up is to find an anchor - an item for which item drop percentages are high and well known. I believe I have found such an item: pearls of wisdom in the bay drop 80% of the time. Why this helps is detailed below. It involves a bit of math, so if you are not into this, and trust the math, feel free to skip to the list of things YOU can do to help at the end.

So how can such an anchor be use to determine the exact item drop of an effect? Assume pearls do drop 80% of the times. If you are using 25% item drop bonus, they will drop 0.8*1.25=100% of the times. However, if you have 24% item drop bonus they will only drop 0.8*1.24=99.2% of the times.

To demonstrate how this can be used I spaded the drop percentage of Metal Detector (MeD). With 8% item drop from other sources I have witnessed the pearl not dropping. So MeD must be below 17%. With 9% however, we have spent 600 turns between us and it dropped every single time. If the MeD bonus was 15%, the chance of this would be (0.99^600)=~0.2% which is low enough for our purposes.

Moving on to other effects, I hope the method I propose is clear. Say you think Gumshoe is 10%. Try first using it with additional 10% and witness it not dropping. Then raise it slowly, 1% at a time. In each case adventure until you witness it not dropping, or repeat 300ish turns with drops.

How DO we know that pearls drop 80%? Ryme already confirmed that drop rates are integers. I have spent 2609 turns with no effects or equipment in the bay and 2094 pearls dropped. By the powers of mathematics we know that with 90% certainty the real value is between 78.98 and 81.54. I believed this to be 80%. But much stronger proof was provided by Satan (see post below). With 24% item drop bonus purely from equipment worn Satan witnessed a pearl not dropping. So the exact drop rate is below 1/1.24=80.6%. Furthermore, he had 501 turns with 25% drop where pearl always dropped. The chance of this happening with 79% drop is (0.79*1.25)^501=0.18%. With very high certainty we are now sure that the drop rate is 80%.

To summarize how YOU can help (I will update the list as more data is available):

• Adventure in the bay with 5% item drop bonus from worn equipment and a Crow. Witness a case where a pearl does not drop and report it, or report count of turns it dropped.
• Do the same with Clockwork kangaroo and -7%.... that's a bit tough to get.
• Adventure in the bay with low item drop bonus from worn equipment and Gumshoe, or any of the other unknown effects, but only one effect at a time. Witness a case where a pearl does not drop and report it. Report exact drop rate bonus given by equipment. We will move on when people start reporting non-drops.

Note: Sadly, if you are a naturalist you have a passive skill "Echolocation" which adds an unknown amount of item drop bonus. Therefore you can't help in this effort until this effect is fully spaded.

Level Up Gains
Simply post what you gained at each level here

Let the data flow in, and let me know what you think! Are there other things you rather see spaded? Do you have better methods in mind?

Nicely done. You're way better at this than me.

Thank you Satan. That pretty much confirms the 80% drop. I'll update the first post with the new data.
I hope you can continue on spading MeD. I'm going to try and get preliminary data on Gumshoe tomorrow.

Satan wrote:but he tends to avoid using nice round numbers.

Correction: "avoid" isn't quite right. I just don't feel a compulsion to always go in multiples of 5.

correction: the monkey seems to give +21%, or +20% drops, i didn't use the formula correctly.

If you're not sure, just keep track of your data and save it somewhere. You can always re-process the data.

i did, i use gumballs to count turns

278 turns, 270 pearls with the monkey and no other +item

It's a good start. We know it is below 25%, we are assuming it is 20%. If you feel bold, try to capture a no-drop with 4%. Otherwise, start with 1% and go upwards.

Most all of my data is tainted ;_; Frickin' echolocation. As a reminder for the future, don't test with naturalists. Level 4 passive intrinsic skill gives +items, and most of my turn sets included a naturalist (enough to say that it wouldn't be worth trying to remember if any didn't have one). Which means all that testing needs to be redone with only non-naturalists. What kind of game creator gives that kinda bonus as an intrinsic to a single class at level 4, I mean really :p We should probably make a big bold notice somewhere that says: You can't participate in spading with a naturalist. >.> In fact, just scrap all the data I've put in. And I'm going to go check if any other classes have any intrinsics that do that, though I doubt it. If I had to guess, the naturalist bonus is +1%, but I suppose we should try and nail that down. Ughhhh.

EDIT: Wiki says no classes besides naturalists have a +items intrinsic. I don't feel like re-doing that spading work at the moment, but I'll get on it tomorrow.

RE-EDIT: According to Ryme, echolocation wasn't actually working. So, we don't have to re-spade this. It's like a miracle! But, Ryme fixed it, so the naturalist's passive skill has to be taken into consideration from now on.

Our findings:
pearl of wisdom droprate - 80%
Metal Detector - +16% items
Monkey sidekick - +20% items
Echolocation - +15% items

Crow sidekick:
2% + Crow - 6 pearls / 7 turns (less than 23%)
3% + Crow - 28 pearls / 29 turns (less than 22%)
4% + Crow - 149 pearls / 150 turns (less than 21%)
5% + Crow - 192 pearls / 192 turns

Echolocation naturalist passive:
4% + Echolocation - 32 pearls / 33 turns (less than 21%)
9% + Echolocation - 93 pearls / 94 turns (less than 16%)
10% + Echolocation - 501 pearls / 501 turns (so, Echolocation is +15% items)

I missed the message with the data for 16%. I suppose you did a lot of turns with 9% and it always dropped?

Yeah, sorry, deleted it when I remembered about echolocation. I had 501 pearls / 501 turns for MD + 9%.

lol it wasn't working?

With 1% had a nodrop with a Crow.

I updated the first post with next tasks for Crow and Monkey.

Clockwork kangaroo might be 25+%. So far 45 turns naked all dropped.
I'll work with scuba mask for negative drops.

Edit: so far:
with 0% 45/45
with -5% 100/100

You take off your old talisman and put it in the tabernacle. You feel energy drain from you, and are momentarily powerless. You pick up the Ultimate Aviator Goggles and put it on. After a time you feel better, but not quite back to normal. You lose 20000 XP.

-10%: 12/13 < 34%
-8%: 12/13 < 33%
-7%: 251/252 < 32%
-6%: 96/97 < 31%

This settles the clockwork kangaroo at 30% as the chance of 250 no drops at 22% is low enough for me (0.002)

bump, new data (Satan, if you can please bump it when you have new data, so I know to read it)

You take off your old talisman and put it in the tabernacle. You feel energy drain from you, and are momentarily powerless. You pick up the Grimm's Bulwark and put it on. After a time you feel better, but not quite back to normal. You lose 20000 XP.

Done with kangaroo, it is 30%. Moving on to age pill.

Sorry for the slow updates. Echolocation spading finished, it's +15%.

If someone's going to do Life in the Song of Keys, let me know. I have quite the collection.

I'm having problems getting enough age pills, so if you are offering a donation for the cause, I can do song of the keys next.

Lost keys:
0% : 18/19 (< 25%)
1% : 3/4 (< 24%)
2%: 16/17 (< 23%)
3%: 6/7 (< 22%)
4%: 13/14 (< 21%)
5%: 22/23 (< 20%)
6%: 8/9 (< 19%)
7%: 2/3 (< 18%)
8%: 2/3 (< 17%)
9%: 5/6 (< 16%)
10%: 7/8 (< 15%)
11%: 28/29 (< 14%)
12%: 14/15 (< 13%)
13%: 77/78 (< 12%)
14%: 20/21 (< 11%)
15%: 94/95 (< 10%)
16%: 6/7 (< 9%)
17%: 500/500 ----------> 8%

Life in the Song of Keys seems to be 8%. I got a bit bored with this so I'll do something else for a week or two and come back to it. I hope other people will take part.

Hey Ryme, are you planning on putting in exact numbers in the effects?

I wont mind exact numbers, there is still much else to be spaded, e.g. all level dependent effects.

If not, I'm going to continue the effort as soon as I get a bit more time on my hands.

Edit: Strangly enough I just had a couple of hours so I started spading Nicely Aged (age pills)

Anecdotal evidence suggests that it is quite a high bonus so I started high:

0% 20/20
-5% 30/30
-10% 53/54
-9% 10/11
-8% 19/20
-7% 25/26 (less than 32%)
-6% 429/430

This settles it at 30%

also you should spade the casino, which doesn't take turns.
how does the rock paper scissors game work against you?
we know the odds have to be against you since they're meant to be chipsinks

Uh, that's not necessarily true. It's RPS. 1/2 of a chance. etc.

I think the messages are meant to throw you off though. Which makes sense.

EDIT: teh mathz iz suckz0rz.

just make like 10 thousand bets of 10 and record

and heck no, i WISH i could afford spading casino

the SUPERnaturalist wrote:just make like 10 thousand bets of 10 and record

That's 100k.

I'm sorry, but spading the casino simply doesn't interest me. Feel free to do it of course, I'll stick to game mechanics.

MagiNinjA wrote:Uh, that's not necessarily true. It's RPS. 1/3 of a chance. etc.

Actually, it's a 1:2 chance of winning.

if it was 1:2 it wouldn't be a chipsink.

the SUPERnaturalist wrote:if it was 1:2 it wouldn't be a chipsink.

That's why it isn't. It's a chip-break-even-on-average-er. Or something. :/

the SUPERnaturalist wrote:if it was 1:2 it wouldn't be a chipsink.

I'm just saying... since you don't lose on a draw (and it doesn't really affect the outcome at all), the odds are 1:2, not 1:3.

Im trying to spade the drop rate of Brass Knuckles. Mostly because no one has done it yet, and it's pretty pointless, making me more interested to finish

Strlikecrazy wrote:Im trying to spade the drop rate of Brass Knuckles. Mostly because no one has done it yet, and it's pretty pointless, making me more interested to finish

If you are going to do a brute force spade you are looking at a few thousand turns to get a reasonable estimate. I can do the exact math, but I'm too lazy at the moment.

If you want to help in a reasonable number of turns you can try spading the gumshoe effect.

Also: updated some more data above regarding age pill.

it has to be a chipsink though, otherwise what's the point of it?

Not everything has a point. Not everything NEEDS a point.

the SUPERnaturalist wrote:it has to be a chipsink though, otherwise what's the point of it?

Fighting enemies isn't a chipsink. In fact, it gives you chips! What's the point of having it?

Olaf wrote:

the SUPERnaturalist wrote:it has to be a chipsink though, otherwise what's the point of it?

Fighting enemies isn't a chipsink. In fact, it gives you chips! What's the point of having it?

He was not referring to fighting but to the casino.
And I don't think it HAS to be a chip sink.

Anyway, Nicely Aged is now spaded - 30% item drop. I was very near to give it 31%, but after 430 turns it didn't drop for one time.

The MMG of KoL, minus the house cut, isn't really a meatsink. It's a meatsink for one player, but it's a huge boon to the other. Then again, that's between two players.

Nice job, Muhandes. Fancy little pill indeed.

Muhandes wrote:He was not referring to fighting but to the casino.

Yes, I know. I was giving an example of something else that wasn't a chipsink. Trying to point out how his argument didn't make any sense.