User talk:Flargen/archive2

From TheKolWiki
Jump to: navigation, search

6900 bricko bricks recipe (unconfirmed)

No seriously, how do you know that there is one for 6900 bricks? What is the reasoning behind it? Because it just looks like someone pulled a random number out of their ass --Lemon-claw 21:25, 4 February 2010 (UTC)

  • There are recipes that take 11, 1100, 23, 2300, 37, 3700, and 69. Thus, a recipe involving 6900 isn't particularly far-fetched. --Quietust (t|c) 21:34, 4 February 2010 (UTC)
  • Well (if Flargen doesn't mind me stepping on his toes), as the person who actually put it there...it's because there was a construct for 11, 23, 37, 69, as well as 1100, 2300, and 3700. 6900 would be a logical fit. There is one undiscovered item between the cathedral and pyramid currently, and we know that there's at least one more undiscovered monster. However, this was done before the gilded BRICKO was discovered. It's more likely now that whatever the slot is for can be formed by multi-using something like 6 gilded blocks as opposed to my guess at 6900. I'm basing 6 off of the minimum amount to make an egg-like shape.--Toffile 21:35, 4 February 2010 (UTC)
    • More specifically, the BRICKO chick hasn't been found (except through chameleons), and there's a gigantic bricko chicken monster we have the image for--which presumably drops the hatchling. --Flargen 02:32, 5 February 2010 (UTC)
Brickogchicken.gif

Useful Script

Hey why did you undo my revision to the Shivering Timbers page? --Lemon-claw 03:09, 11 December 2009 (UTC)

  • It's not useful. It's not a script. It's a javascript code sniplet used for url manipulations, in order to exploit holes in the game that let you access things you normally wouldn't. And it has long since lost its relevance, and was a short-lived bug that was quickly fixed. It has no place on the wiki. The only time it does is when it is content intentionally made accessible only through such means. Meaning that it is hot stuffing. --Flargen 03:12, 11 December 2009 (UTC)

Template:Item

Next time you modify this, can you get Gift Items to autocategorise themselves as well? If you could get the DoD potions autocategorising as well that would be nice, but I'm not sure it's worth the effort. And by "DoD potions" I mean "items that have a type of combat / usable; like the DoD and procrastination potions".--Ryo_Sangnoir 16:10, 4 May 2009 (UTC)

  • Yeah, should be pretty easy to do the gift items. As long as the slash doesn't cause an issue for the wiki the bang potion types can be handled, as well. I might take a bit to do that, as next time I look at it I might try to think if there's any good way to make it more readable/compact, as well as a way to throw in an auto-cat parameter so it can be used on standards pages and such without categorizing the page. --Flargen 20:25, 4 May 2009 (UTC)

A Menacing Phantom

You said "no, not really, people do that all the time". Jick uses hyphenated "-o" for plays on breakfast cereals and snacks that end in "-O" and for plays on words that end in an "oh" syllable, like "ooze-o" for "orzo". What is your hypothesis, in the context of an ongoing Jar-Jar Binks gag, why Jick would hyphenated the item with a "-o" or do you think that "people do that all the time" for no reason at all? —BozoTheScary 14:26, 23 May 2009 (UTC)

  • If I remember an old radio show right, it's because of the Metallica cover of "Whisky in the Jar" ("Whisky in the jar, ohhhh"). It was a running gag for a little while. --TechSmurf 14:40, 23 May 2009 (UTC)
That does sound like a better explanation. The hyphenation seems unnecessary in that case, but what the hell. Thanks. —BozoTheScary 13:25, 24 May 2009 (UTC)

Did I miss something?

Getting a captcha every time I edit something claiming an external link. Was it because of the spambot earlier last week?--Toffile 14:03, 1 June 2009 (UTC)

  • Uh...maybe, not sure. As an admin I get to bypass captchas. I haven't heard others complain. So maybe there's something weird going on, a la those ModSecurity issues that prevented us from using words like "coalesce" on pages we actually wanted to be able to edit. If no one else seems to be getting this issue, you could try asking Jinya. Or Quietust if you don't quite want to go straight to the biggest wig. Quietust seems to be pretty knowledgeable about the various technical specifications of the wiki. --Flargen 14:10, 1 June 2009 (UTC)
    • I wouldn't know about this particular issue, but I've edited on a few other wikis powered by MediaWiki which would present a CAPTCHA each time you added an external link. If you're getting it on every page, then there's probably something wrong, but you'd have to ask Jinya/Mag/Frostbyghte/whoever. --Quietust (t|c) 14:19, 1 June 2009 (UTC)
      • Hm. It wasn't quite every page now that I double check, but it was popping up in weird places. (For example changing NeedsContent to NeedsSpading or creating a metadata page). I'll have to do some tests I suppose.--Toffile 15:27, 1 June 2009 (UTC)
        • Maybe it's your user name it's choking on. It sort of looks like "to file", so maybe that's...uh...not sure. But, yeah, something weird alright. --Flargen 15:29, 1 June 2009 (UTC)
          • Quickest test ever. Something's defeinitely up. See User:Toffile/Test. Something seems to be up with captchas and NeedsSpading.--Toffile 15:38, 1 June 2009 (UTC)
            • No idea what's up. There are no external links in either of those templates. --Flargen 15:40, 1 June 2009 (UTC)
              • I got the captchas on three consecutive edits to Outfit and that does not use the NeedsSpading template. It does use a bazillion Data: templates and these:
                1. {{!}}
                2. {{!!!}}
                3. {{Click}} (protected)
                4. {{Data1}} (protected)
                5. {{Element}}
                6. {{InlineItem}}
                7. {{Plural}} (protected)
                8. {{StackItem}}
                --Club (#66669) (Talk) 17:41, 17 June 2009 (UTC)
                • I bet it's {{click}}, since that works by inserting the full URL of an image into the page. --Quietust (t|c) 17:42, 17 June 2009 (UTC)
                • Click does trigger it: Click test page --Club (#66669) (Talk) 18:09, 17 June 2009 (UTC)

Hardcore + Freeing the King = Soft-Perm Skills?

Hello, I saw you reverted my edit to Hardcore, but as far as I can tell, you really don't get your soft-permed skills back after smashing the prism. There's a short discussion about the same thing on the Talk page that says basically the same thing. Is there some technicality I'm forgetting about? I didn't understand the comment you made in your revert. -- Xerxes314 02:30, 8 June 2009 (UTC)

If I may try to avoid stepping on Flargen's toes...all those discussions are prior to announcement on April 2nd which made the skills available.--Toffile 02:35, 8 June 2009 (UTC)
Oh! That's very interesting. Probably some text describing that should be on the page somewhere? -- Xerxes314 03:11, 8 June 2009 (UTC)

Template Requests

Hey! I was talking with TechSmurf about this, and he recommended putting a bug in your ear just in case he didn't get it done himself.

I recently updated the Clover Adventures page to look much more like the Semi-Rare Adventures and Bad Moon pages, but I found need for a couple of template-related changes, and as I'm rather template-inept, I thought I'd ask for some help.

First, could StackItem be given a default value of "something" (with the question mark image), similar to the default for an empty Acquire?

Secondly, I'd like to put in a request for a "StackMeat" template similar to StackItem-- something very simple, with the meat image, and in small font below, "X Meat". It would be very helpful for tables such as the one on said page.

Thanks! --Southwest 01:41, 31 July 2009 (UTC)

  1. What do you need a generic option for stack item for? It's just an image followed by an item name (unless something else is specified) with link, so there's no good context for default text. Acquire's comes from specific in-game text when fighting darkness.
  2. As for the meat template, that wouldn't be hard (it could almost be done with StackItem as is, long as you gave Meat a data page; the shortcoming being the lack of an amount parameter). How would you feel about a "short" form of {{meat}} that just spit out something like "Meat.gif 1000 Meat" (on the same line)? --Flargen 22:01, 31 July 2009 (UTC)

Slimeling

As a response to the message you sent me: I'm afraid I can't be more specific, for obvious reasons. However, I would appreciate it if you took my comment as a hint and not respond with snarkiness and hostility.--Ekeinos2 02:45, 31 August 2009 (UTC)

  • Disgorging works like pickpocket, except with a different base drop rate applied. If you scrolled to the bottom of the page, you would know. Now, what's this obvious reason you can't be more specific? Are the multipliers wrong? Is it wrong to say it uses pickpocket-like mechanics? It's not so much that Flargen is being hostile; you haven't started any discussion refuting the data that exists on the Talk page and so there is nothing to substantiate your claim that something needs spading. What is this something? --CG1:t,c,e 03:12, 31 August 2009 (UTC)
    • Stating that I lack common sense and can't scroll down is insulting. I am obviously well aware of the existence of the tables and the comments at the bottom, since I'm making comments on them. As it's obviously not appreciated, I will not make any more attempts to give you any hints. So, feel free to remove the "needs spading" tag again, and keep a page that not only ignores edits from dev team members, but also has raw data in the discussion that highly disagree with the predicted values from the formulae in the main page.--Ekeinos2 03:32, 31 August 2009 (UTC)
      • See, this is what Flargen and I were asking for; substantiation of evidence; if you said "the table assumes independence of pounds" earlier, this revert war wouldn't be happening. Now, what data in the discussion disagrees with the formulae on the page? Telling us "there is a problem" doesn't help us at all if we don't know what the problem is. Hinting that there is a problem also doesn't help much either. Also, grammatically, it would be "ignores edits from a dev team member" (it's singular, unless you happen to be two people). --CG1:t,c,e 03:46, 31 August 2009 (UTC)
      • If you think there is a flaw from the data in the talk page, especially one that "highly disagrees" with the predictions, then POINT IT OUT. You don't need to be a dev or not to see such a thing and comment on it. That and it would be fairly silly of me to assume "Ekeinos2 on wiki = Ekeinos in game". Or would you naturally believe that I am actually named "Flargen" in game? Put aside that I actually go to efforts to put in a disclaimer on my page about who I may not or may be. I know you don't want to be all abusive of secret dev knowledge, but if you can look at the data and point specifically at it and say "this doesn't seem to work right", then you would actually be doing something useful without violating some sacred dev sanctum. Heavens forbid you should stoop to that level. Once I know where this alleged flaw is, I will revisit the issue in my spading to see what I can find out. I'm just meat farming right now anyway, so it's not like I'd be particularly inconvenienced by doing more disgorging spading. --Flargen 04:31, 31 August 2009 (UTC)
        • And I see there are edits going on on the slimeling talk page, so hopefully this is just such the actually useful commentary we were looking for. But point being, you were a bit overly cautious of your super dev powers, and "I am dev" is not a particularly useful defense or argument for practical purposes --Flargen 04:33, 31 August 2009 (UTC)

Would you mind...

I was taking a look at Familiar abilities and realized that it's essentially an orphan. Do you think we should just insert a wikilink into the "Abilities" title of Template:Familiar? I mean as it stands now, most people aren't probably going to find it. (I'd go ahead and do this, but I'd rather be safe than break such a widely used template)--Toffile 03:21, 1 September 2009 (UTC)

  • I'm actually not so fond of the page myself. It's not very pretty-looking. I prefer Familiars by type. Still needs work, but the formatting's a little nicer and it's easier to parse and find what type of action you're looking for. It serves essentially the same function, as well. The Abilities page would need some substantial reworking to even make the table have a meaningful sort function, which is the minimum it would seem to need to be really be a worthwhile page. --Flargen 03:29, 1 September 2009 (UTC)
    • Well I don't mind the Abilities page. It serves a similar but slightly different purpose to the Type page. While both let you look up a familiar's abilities, the Abilities page at least lets you take a look at all of its abilities at once, rather than one-by-one. I mean, if I wanted to look up Volleywhelps, I'd first have to check the Type page for all the Volleys, then cross-reference them with the Whelps. The matrix gives it a better reference. Both pages should probably be given a facelift though.--Toffile 03:36, 1 September 2009 (UTC)

User != Talk

  • good call. many thanks, i blush. --Evilkolbot 21:29, 8 September 2009 (UTC)


Familiar Metadata for the win

I suppose that was inevitable, especially since the "use item" for familiar hatchlings shows the adult familiar's image. I was thinking the familiar name itself should also be in the data page (as it is for pretty much everything else), but given that it only actually shows up on the familiar page itself, you were probably correct to leave it out. --Quietust (t|c) 21:49, 2 October 2009 (UTC)

  • I wouldn't be opposed to the name; I just suddenly felt confused about what to call it, since there's a conflict (in my brain, at least) with the "default name" you get when hatching it. I wasn't entirely sure about the gender, either, since right now I think that's only relevant to the use item template. I just didn't feel like throwing in a second parameter to that template to deal with the matter. --Flargen 21:55, 2 October 2009 (UTC)
    • Looks like JRSiebz just found a good reason to add the "name" field - it allows using {{StackItem}} and {{InlineItem}} directly with familiars. --Quietust (t|c) 13:29, 6 October 2009 (UTC)

Jick braek teh tamplete!?

The new dolphin whistle is being listed as a spleen item due to {{item}}'s assumption that level requirements are for consumables... Jick needs to put another notch in his belt. --BagatelleT/C 01:47, 27 October 2009 (UTC)

  • Oh, nuts. That's something of a conundrum. How about a parameter to specify that it isn't a consumable, and without that specified at all (or appropriately) it'll just assume it's a spleen item? That should prevent us from needing to update every single spleen item page because of a single item (for now). --Flargen 01:51, 27 October 2009 (UTC)
    • I reckon it doesn't much matter how it's fixed, since Jick will only break it again later. 'Cos he's a clubby griefing meanie. --BagatelleT/C 22:43, 27 October 2009 (UTC)

redirects to subsections

  • something's changed. i thought they didn't work, too, so i tried it to to get proof. did seem to work, though. try it yourself. --Evilkolbot 16:18, 20 March 2010 (UTC)
    • Nope; they still don't work. I'm guessing you clicked the link on the actual redirect page, which works - however, the actual redirect doesn't go to the subsection. --CG1:t,c,e 18:51, 20 March 2010 (UTC)
      • That redirect to subsections was added in MW v1.14. KOLwiki runs on v1.13. -- sulfur 19:42, 20 March 2010 (UTC)
  • i stand corrected. i clicked the link on the diffs page. oh well. --Evilkolbot 21:33, 20 March 2010 (UTC)
    • I suppose it wouldn't hurt to just leave the subsection redirects as is. If our MW ever updates to allow it, they'd transition in seamlessly. It's not like they're problematic right now. Only issue is that if the name of the sections on the page change, we'd have to go look through the redirects and update accordingly. But that'd be an issue even with the MW upgrade. So perhaps I'll just leave future subsection redirects alone, long as they point to a correct subsection. --Flargen 00:14, 21 March 2010 (UTC)

Initiative for Briny Deeps

It looks like at some point you made a change for the Briny Deeps monsters to have 0 instead of ? initiative. I've been finding that a 0-initiative monster always allows one to run away, but that's not the case with any of the three monsters here. Did you have a specific reason for setting it to 0 instead of ?--Foggy 15:35, 14 June 2010 (UTC)

  • I'm not sure exactly what you mean, but is your question taking into account the initiative penalties the sea pressure gives you? And I did the usual PonE method for getting initiative values in the sea, after establishing what the pressure penalty was. And all of them in the deeps (and most everywhere else in the sea) came out to be 0. --Flargen 16:28, 14 June 2010 (UTC)
    • And if anyone's wondering how I know for certain it's a pressure penalty and not just that, say, everything in the trench has an initiative of 200: putty copies. --Flargen 16:37, 14 June 2010 (UTC)

Mother slime attack still wrong

After a while I had looked again at the slimetube page, just because, but I noticed that the coated in slime turns from the construct and colossus wasn't changed, nor Mother Slime's attack formula. I'm pretty sure her attack was confirmed as (10000 - turns_spent*10) + 1000 instead of 11000 - turns_spent*10, but I'm also sure the template for that is protected, could you maybe do it? But as I wrote this I changed the turns you get of coated in slime on the slime construct and colossus so 3 is the minimum (which it is) --Andrew3 19:13, 21 June 2010 (UTC)

  • I'm not sure what template you think it is needs to be edited (by an admin in particular). Pretty sure everything you're talking about is an input parameter to a protected template. But you can edit the inputs to a template call. The template that's actually protected is for formatting and uniformity purposes. But one is generally free to invoke templates on any page at all. Within reason. --Flargen 18:06, 26 June 2010 (UTC)
    • (Almost forgot about this) Yeah, I see now. I assumed when I saw (protected) near the bottom that even if I hit save it wouldn't do anything to it. A wiki noob is me of course. Changed now though, thanks. --Andrew3 02:34, 5 July 2010 (UTC)

Monster Defense

Back in early 2009 there was discussion about changing the monster infobox to display 90% of the input value, to reflect what was actually useful to players ingame. You seemed in favour of it at the time - would you mind making this change?

Should I edit Template:Combat/data to add a comment about the wiki's defacto policy of recording the coded defense? I can't find it written up anywhere else. If I were planning on writing this up, variance and all, what would be the best page to put it on? I was thinking of Spading, but I'm not sure it fits with the tone.--Ryo_Sangnoir 09:43, 26 June 2010 (UTC)

Edit: I just realised a potential problem with this - scaling monsters have "scales" or more precise notes written in for their defense. I'll go trawl parser functions to find a solution. --Ryo_Sangnoir 09:51, 26 June 2010 (UTC)

I've put together a terrible hack that works if the defense begins with "scales", which should be most of them (form of "Scales with..." or "Scales to..."). I can't get #iferror to work, and that would be a much better solution.

{{#ifeq: {{#expr:0.9*{{{defense|''unknown''}}}}}|Expression error: unrecognised word "scales"|{{{defense|''unknown''}}}|{{ceiling|{{#expr:0.9*{{{defense|''unknown''}}}}}}}}}

Reading over my initial request, I'd just like to note that I am in favour of recording the coded defense, and eagerly await those distant, halcyon days when we'll have the variance listed as well. --Ryo_Sangnoir 10:41, 26 June 2010 (UTC)

Deus Ex Machina found a solution that works better, so I'm recommending that one instead. Full code for the line is:

| {{#switch:{{{defense}}}|!=Indeterminate|{{#ifeq: {{#expr:{{#expr:0.9*{{{defense|''unknown''}}}}}}}|Expression error: unrecognised word "expression"|{{{defense|''unknown''}}}|{{ceiling|{{#expr:0.9*{{{defense|''unknown''}}}}}}}}}

--Ryo_Sangnoir 11:16, 26 June 2010 (UTC)

  • Here's the main hiccup I can think of: I'm not sure if people have consistently been using the same standard for adding defense values. Some people might be using the coded defense (before the 90%), others the "visible" defense (after the 90%). I know I always used coded, and that old discussion suggests others had typically been using coded, but I can't really speak for the other editors. If we figure that's acceptable and that you and whoever else is a spader active on the wiki will go through and double-check defense values for accuracy, that is fine. Also, I think it will be easier to just add a "Scales" case to the switch statement that then gives the entire phrase as the case output. Will involve adjusting some data pages, but thankfully most scaling monsters are easy to find. Will make sure the phrasing and capitalization are uniform across the wiki, too. I could perhaps add a temporary categorization for the "error" part of the code that will make it easy to hunt-down any of the scaling ones, and then remove that part when we've fixed them. --Flargen 18:00, 26 June 2010 (UTC)
    • I'm fine with double-checking defense values for accuracy; I don't think many people have been adding defense values and they will have consistently added either the coded or the visible, so a list of pages that need changing could be found by checking the most recent monsters and looking for the editors. The problem with the scaling is that some monsters give an exact value to how much they scale - the only ones I can find offhand are the Feast of Boris monsters -- although you are right in that making things consistent is better. Most scaling monsters aren't listed on Special Monsters, either. I also thought mother hellseals might be a problem, but I don't think the 90% reduction is applied to scaling monsters at all - this includes the mothers.--Ryo_Sangnoir 21:54, 26 June 2010 (UTC)
      • Done. Didn't add any extra categorization, but added two "default" scaling input options (Scales or scales). --Flargen 00:30, 27 June 2010 (UTC)
        • Looks good, thanks. I've just realised that Template:Combat also shows monster defense, so that should probably be changed as well.--Ryo_Sangnoir 07:48, 27 June 2010 (UTC)

Just chiming in a little here. I have added some monster defense values to the wiki when I see that they are wrong (which, they mostly are, across the board). By "wrong" I mean "not what you see in-game." Which is to say, they're usually not taking into account that they're (in a great majority of cases) 90% of Attack. Which says to me that they're not coded (in-game) as equal to attack, but coded that they should be 90% of attack unless overridden. It doesn't make much sense to me that they'd be displayed on the wiki with the value that you don't see in-game, since mafia also uses the wiki's values when someone updates mafia's data files, and if you think a monster has 100 Attack and 100 Defense, that's a good bit different than thinking it has 100 Attack and 90 Defense, even though the wiki might claim that it has 100 Defense, when it actually doesn't. --RoyalTonberry 08:17, 27 June 2010 (UTC)

Adding further onto this thought process, what you might do is code the combat infobox data stuff to display 90% of Attack unless there's a flag provided for the actual defense. Perhaps this is what Ryo was talking about? It seemed more like he was saying that he wanted it to display (1.11*defense observed in game), and then you'd scale it back by 90%, which seems like it'd cause all kinds of confusion when adding data to the wiki. I think it'd be a lot better if the wiki handled it the same way as the game probably does. Which is to say that it uses 90% of Attack unless the monster has an overriding defense value defined. --RoyalTonberry 08:22, 27 June 2010 (UTC)

  • The trouble here is that we think we are handling it the way the game probably does - that is, the game has a value which may or may not be equal to attack, and then the amount of variance is calculated based on that value, applied, and then the whole thing is rounded and then multiplied by 90%. Essentially, we're using QuantumNightmare's convention - there's some discussion on my talk page and Salien's talk page, and here's a link to the AFH forums. Take, for example, the sabre-toothed ferret. Should have a range of 194-203 according to QN; I guess you'd say it was a coded defense of 200 (as different from the attack + 5 variance). I'll check it. The mafia problem should be fixed if they just know to take 90% of the values - which should be the ones we're now displaying.--Ryo_Sangnoir 09:28, 27 June 2010 (UTC)
    • I guess what I'm trying to say here, without being able to find a better example, is that assuming QN's model makes the numbers nicer - a defense variance of 194-203 could mean a coded value of 199 or 198, or it could mean a coded value of 220 depending on the system. But for 220 that's the maximum range, whereas for 198 or 199 you'd expect to see a 193 or a 204, respectively.--Ryo_Sangnoir 09:42, 27 June 2010 (UTC)
    • I understand what you're saying. I don't think you're understanding what I'm saying though. I strongly doubt that the way you want to handle it is how the game handles it. I would think that the game says that if it is not provided with an overriding monster defense value, then that value is 90% of the monster's Attack. Any variance of Attack/Defense is calculated off of Attack in this case, which readily explains why you'd see a variance of 5 on an observed monster defense average of 90, when the monster attack observed average is 100. Then, beyond that, if the defense value is overridden to not be 90% of Attack, the variance is calculated off of Defense (probably). I am not aware off the top of my head of a monster that has a higher defense value than attack, as well as having the defense and attack value be in a different mod20 tier, as well as having them be less than 100. Which is what you'd need to test if hardcoded Defense varies on its own, or uses Attack's variance (though perhaps a Demoninja would work?). What you're asking for with this policy seems like it's a dirty dirty hack to fix something that ought to have been handled properly in the first place, and will cause far more confusion for people adding monster defense values to the wiki than it's worth. I suppose that if you can get Jick to come along and say that it works the way you think it does (super wacky) as opposed to the way I think it does (much cleaner), then I will let up; however, intentionally asking that people put data into the wiki that isn't actually what the game says is ridiculous, and is just begging for a load of confused edits and reversions. --RoyalTonberry 09:50, 27 June 2010 (UTC)
      • I believe that monster level additions add to the base defense, not the modified defense; so this crossing of the mod20 tier. I'll check out the Demoninja, thanks for the suggestion.--Ryo_Sangnoir 10:06, 27 June 2010 (UTC)
      • Okay, right. Working with hostile amoeba. Adventuring with no +ML gives a def range of 22-24 - consistent with a 25 by my reasoning and a 23 by yours. Adding +15 ML (ring of aggravate monster, detuned radio) gives a base 40 by my reasoning and (presumably?) a 38 by yours. Or perhaps a 36.5 by yours depending on whether you're 90%ing the ML addition or not. Range of 36-37 received, not consistent with anything (possibly more turns would get an answer). Adding +16 ML (ring of aggravate monster, detuned radio, tiny plastic 7-foot dwarf) gives a range of 36-38. Clearly I plan to put more turns in to this; I'm just sticking this here so you can see what I'm starting. Everything I've noted so far is consistent with the theory that Def is 90% of attack with the variance attack implies, unless stated otherwise; after the 90% is applied the value is rounded up; variance is 5% of the value given, at most 5, rounded down; after adding or subtracting variance the result is rounded up; monster level is multiplied by 90% and then rounded down before being added and then the variance is applied after this. With the massive amounts of rounding going on it's almost certain there's a better theory that involves less rounding, but I can't find it right now.--Ryo_Sangnoir 10:51, 27 June 2010 (UTC)
      • Also: I made a short trip out to the Demoninjas and am stuck with a range of 27-28. Tomorrow I'll note the amounts of 27s and 28s I get so we can have a look at the distribution: how many demoninjas do you think I should kill to be reasonably sure there isn't a 26 or 29 hiding somewhere?--Ryo_Sangnoir 11:17, 27 June 2010 (UTC)
        • Yeah, I think that the way it's actually calculated is as follows: base_defense = <either hardcoded amount or 90% of attack>. total_defense = base_defense + ceil(0.9*bonus_ML) + in_combat_modifiers. (in_combat_modifiers would be like.. de-leveling skills or whatnot, that affect the value and don't scale to 90%). So I would expect to see, in that case, 23 + ceil(0.9*15) = 23 + 14 = 37 defense. Since its Attack is 20, that's a variance of 1 (this variance is computed before additonal ML, just ask fluffy bunnies at +99 ML, they do not vary), which is precisely what you observed. With your model, you're saying that you get ceil(0.9*(25 + 15)) = ceil(36) = 36. You'd also expect a variance of 1, and you'd see 35-37, not 36-38 which you observed. --RoyalTonberry 11:18, 27 June 2010 (UTC)
        • Actually, the variance is also multiplied by 90% along with the base, so I'd expect 36-38 as I observed, if the variance works off base ML (thanks for that - I was unaware). Right now I'm working on the demoninjas and have 11:6 27s to 28s. These are exact figures. I guess I'll run 100 and see how it turns out. For explanation of my theory I'll just link to User talk:Salien, which contains examples and confusion.--Ryo_Sangnoir 11:41, 27 June 2010 (UTC)
  • firstly, sterling work getting the template to work this out. very impressive.
  • unfortunately, although i disagree with the policy generally, i'd have to agree with RT. we can't second-guess why the game really does things, all we can do is model how the game seems to work (most physicists make this mistake.) yiab's awesome work on item drops should IMNSHO be only a first step to working out what the actually coded drop percentages are. i was, of course, loudly voted down on that one, and 4.86% remains, even after jick has stated clearly that percentages are either integral or conditional. so, for consistency's sake we should report in the wiki what the game tells us, even when there is clear evidence pointing another way. (do you see what i did there?) --Evilkolbot 11:04, 27 June 2010 (UTC)
    • For most of the newer monster-stuff I add, I add the figures I get in the summary so that they can be changed if necessary. Unfortunately back when I was adding figures when this whole thing was a new idea nobody dissented QN's version of how defense worked so I assumed it was proven and done with, and now have to pay for that assumption. And on item drops I think we've drifted closer to integral percentages that fit with the data, now.--Ryo_Sangnoir 11:17, 27 June 2010 (UTC)
    • Hah, I do see what you did there. That (item drop rates) is a case where I think a reasonable exception can be made since we have been told how it works in the game. We have not been told how defense calculations work in-game, can just make educated guesses. But we have been told that drop rates are stored in the database as integers, and usually multiples of 5, especially past 10%. --RoyalTonberry 11:18, 27 June 2010 (UTC)
    • If you actually work out the math on Yiab's item drop data, you will see that the data is generally not strong enough to conclude, with reasonable strength, that the only valid integer rate that would explain the data is 5% (using your hypothetical 4.86%). The problem with what Yiab did is that he used very, very low item drop bonuses. This means he was dealing with the vicious fact that variances on a straight percentage tend to be large fractions of the base rate. To toot my own horn a bit, whenever I posted drop rates on monsters, I would use integers whenever the data would strongly suggest it is the only viable integer drop rate. However, even though I used high item drop bonuses to mitigate those bothersome variances, it was not always the case that I could get a single integer to lie within the probable range of true rates. In which case I went with a decimal to reflect that fact. --Flargen 19:03, 27 June 2010 (UTC)
  • So, let me see if I've got this right. Using beanbats because they're so convenient. Attack is 21, def is 90% of attack = 18.9. Adding 19 ML to get to 40 as I was before, you take 0.9*40=17.1 and round up to get 18. Then 18 + 18.9 = 36.9. Base attack of 21 implies a variance of one, which gives a range of 35.9 - 37.9, which is rounded up to give 36 - 38. Meanwhile, I think the new base is 40 which I multiply by 0.9 to give 36. The base defense is 21, so the variance is 1, which is then multiplied by 0.9 to give a variance of 0.9 and a range of 35.1 to 36.9. Rounding up gives a range of 36-37; this is the range I observe. --Ryo_Sangnoir 11:53, 27 June 2010 (UTC)
    • Duplicating your test, with beanbats: +0 ML for 50 turns, I observe: 18 defense @ 15/50, 19 defense @ 26/50, 20 defense @ 9 / 50. Then, adding +19 ML for 50 turns, I observed 36 defense @ 39 / 50 and 37 defense at 11 / 50. That is very strange, why would we see a variance of +/- 1 at +0 ML, and not see the same variance at +19 ML? I may have to get a character to examine this further, as that base variance looks an awful lot like it could be a triangular distribution. --RoyalTonberry 12:22, 27 June 2010 (UTC)
      • I would explain why we see a variance of +/- 1 at +0 ML like this: base defence of 21 multiplied by 0.9 gives 18.9. Multiplying the variance of 1 by 0.9 gives 0.9, and thus the range is 18-19.8. All values are rounded up, so the range is 18-20. The rounding makes figuring out what is happening even more difficult - I can't see what could cause such a high proportion of 18s by this reasoning. Taking 90% of variance does explain only getting two values for the defense of demoninjas, as well.--Ryo_Sangnoir 13:37, 27 June 2010 (UTC)
      • Yes, defense is triangularly distributed. I thought that was well-known. Triangular distributions seem to be one of their favorite ones to use when creating variable ranges. I have a lot of data from getting defenses from critters in the sea that make that fairly apparent. Or I should. I at least have a whole lotta logs from the sea when I was punching fishies to death; since I was already assuming the distribution, I didn't always make official recordings of repeated defense values (beyond the mafia logs). You can probably ask Eleron about it, too; fairly sure the triangularity is known to him, as well. And, to my knowledge, the "take 90% of a coded value" method covers all cases simultaneously. You can probably math out why whatever method you want to propose is logically equivalent to that one. And at that point it's easier just to assume that it is always using 90% of some value, be it a default to attack or a hard-coded value. --Flargen 19:03, 27 June 2010 (UTC)

Someone mentioned earlier they didn't think that any monsters had higher defense than attack, but quite a few of the Infernal Seals do. However, because the defense is only recorded as a range they're currently displaying negative numbers... you can see that, for example, at Hermetic seal.--Starwed 17:00, 27 June 2010 (UTC)

  • Ah, yes, that would be a problem. Could maybe code up the dash in a different way that makes the template see it as a string. Might be able to detect if the template thinks the defense is negative, in which case it can just display the listed defense and not try to calculate. --Flargen 19:03, 27 June 2010 (UTC)
    • Okay, adjusted the template so that it will not attempt to perform calculations on a range. Or, more accurately, will not attempt to perform calculations on a negative value. Might as well de-break that thing while the debate rages on about whether we should undo the changes altogether or not. --Flargen 19:21, 27 June 2010 (UTC)

As the frenzy of interest now appears to be over I'll post my last arguments here: if we can agree that the variance is also being multiplied by 0.9, then we are essentially arguing the same thing: I, that a value is stored on the servers, and this value and the variance are multiplied by 0.9, and RT that a value is stored on the servers (if different to 0.9*attack), the variance is multiplied by 0.9 and applied to this.

If we cannot agree that variance is multiplied by 0.9, then I can't see how you'll explain the Demoninja, with a range of 27-28. If we can, then the hostile amoeba has a range of 22-24, so you would have to allow for fractional stored values, as no integer works for that range.

I am interested in any work on the distribution, though I have no multis with which to contribute, as of yet.--Ryo_Sangnoir 18:35, 2 July 2010 (UTC)

Template:Combat still needs to be changed to reflect the new defense system - would you mind doing this, also?--Ryo_Sangnoir 20:59, 2 July 2010 (UTC)

  • Done. Should have essentially the same behavior that the infobox currently does, with one exception: it will default to using 90% of the attack value from the data page if there is one (so if there is no defense value in the data page, nor is there a defense score specifically passed in). It currently does not give any indication to the casual reader that it is simply guessing that this is valid. I can probably fix that, if desired. I can make the infobox do the same thing if people like that behavior. --Flargen 02:11, 3 July 2010 (UTC)
    • Sounds like a good idea in theory: many monsters do have atk=def and for those that don't, it's normally somewhere around there. But it's not working: all the data pages are created with "?" being the default value, so it just defaults to "unknown". See The Sorceress' Tower (The Stairs) for a current example.
      • I don't have a lot of problem with that. I could change the pre-load of the data page to leave defense blank by default, though. --Flargen 19:13, 3 July 2010 (UTC)

Template help

Hey, having problems with a template again. If you look at {{test2}} you can see I'm trying out a template for adventure sequences. It works ok if I just use text as the arguments, but when I use the {{adventure}} template it all goes to hell. If you've got the spare time could you eyeball it over and see what I'm doing wrong? --Starwed 21:24, 28 August 2010 (UTC)

  • This is probably a tabling issue. The pipe symbol | has overloaded meaning; sort of. In a table you use it to denote new rows and columns, but in template calls it is used to separate arguments. So most likely the pipes in the table code for {{adventure}} are being misinterpreted as argument separators when it's being invoked in the use of {{test2}}. I believe I've encountered this problem before. I don't think I found a way to truly resolve it. There is {{!}} and such that were made to help get around such parsing issues, but it can only do so much. Basically, you cannot wrap tables around tables through template calls. You should try using styled div tags or something, instead. --Flargen 21:58, 28 August 2010 (UTC)
    • I've found that building the table directly in html is also a work-around. --Fig bucket 22:03, 28 August 2010 (UTC)
      • Ah, yes, it is. I think one time when I encountered this issue was back when I was converting the wiki's html tables to wikitables, and found that doing so broke certain templates because of this issue. So, yeah, you should be able to do an html hardcode of the table in {{test2}} to get what you want. Or use other styled html tags. Basically, html will fix it. --Flargen 22:05, 28 August 2010 (UTC)
        • Replacing the outer table with HTML didn't seem to work, although I'm guessing changing the inner table would have. The problem was just that wikimedia tables apparently want to start on a new line. (Although this problem only appeared in conjunction with templates, so it's not universal, I guess.) Anyway, that problem is basically fixed now... thanks for the help! --Starwed 23:16, 28 August 2010 (UTC)
          • Yes, nesting tables in wikimedia format is full of bothersome little issues like that. --Flargen 01:55, 29 August 2010 (UTC)

superlikelys, types, etc

Hey, wish you'd brought this stuff up in discussion. I know the wiki-consensus moves at a glacial pace, but worth a shot occasionally.  :)

I don't like the one-time-superlikely tag. There are other adventures which are one time, so it makes more sense to have an independent field for that.

Also, I really don't want to have the general superlikely icon be in the same place as the icons for clovers/bad moon/etc. The whole point of those was to separate adventures which required you (in some sense) to go looking for them, that you would not normally encounter while adventuring in a zone. Superlikelys are often the opposite -- vitally important adventures which you'll encounter after just a few short turns in a zone.

My preferred way of dealing with this would be to get rid of the false combat/noncombat dichotomy, and I think stupac mentioned something similar in the main forums. Once there's a section for adventures which aren't part of either the combat or noncombat queue, there will only be a point on having icons on those which belong to a more specific class.

Lastly, I'd considered and rejected putting the type info directly in adventure data pages. This is pretty pedantic and I'm not super-tied to the idea, but my thinking had been that the actual game logic seems to be implemented as part of the zone, not the adventure. I'm sure if Jick wanted to, he could implement an adventure that was a semi-rare in one zone and a superlikely in another. Now to be honest, that's probably not really a good reason compared to the utility of adding it directly in data pages. --Starwed 10:24, 9 February 2011 (UTC)

  • Does adding the icon really detract from understanding in any meaningful way? I have a hard time seeing how it does. The other thing is that you're sort of right about encounters being able to be more than one thing, but you're wrong about it being tied to a zone. Jick has specifically said that the friar's noncombat have both superlikely and noncombat copies, so in some sense that adventure is both. However, it's the adventure that has that property, not that zone. My goal is to make this information is available as possible and as correct as possible, and that's sort of impossible if the adventure pages are agnostic about what kind of adventure they are. --Stupac2 16:04, 9 February 2011 (UTC)
  • Your statement on what the icons serve to do is contradicted by the Bad Moon adventures. Most of those do not require you to go out of your way to find them. They just happen if you spend a few turns in the zone. Just like superlikelies. Which is what BM adventures are in the first place. I'm open to other formatting ideas, and perhaps in a few days or less we'll have a cool idea that completely replaces the icons, but I wanted to get the ball rolling on a tangible change in the formatting, and this plus the separate categories for the "other non-combat" types was for that purpose. --Flargen 17:15, 9 February 2011 (UTC)
Ascending Bad Moon isn't going out of your way? Anyway, the vague plan I had in my head when I started the tagging was to have other icons (such as one-time) appear after the title of the adventure. --Starwed 04:06, 10 February 2011 (UTC)
No more going out of your way than ascending Teet or Oxy, and you can get special adventures in the Barrr for doing that, too. Or ascending as a moxie sign so you can do the Going Postal quest, and yet you've had those tagged. --Flargen 05:09, 10 February 2011 (UTC)

Combat Initiative

On Deux Ex Machina's page, you mention that runaway is now the way it should be...how were you able to ascertain that happening?--Foggy 14:11, 4 April 2011 (UTC)

  • /dev told me. Though it would be worth it to independently verify on a monster whose initiative was known through the bugged mechanic. I think Jick had a plane to catch when he pushed the change, so no announcement was made. I updated Combat Initiative back when it got fixed. --Flargen 14:41, 4 April 2011 (UTC)
    • I double checked in the menagerie level 3, and confirmed the change. I saw that you also updated Run Away, I updated Spading to reflect the change. I was just bummed since I had done quite a few sections.--Foggy 16:40, 4 April 2011 (UTC)

Re: Question Marks

Sorry, I've fixed up what I can. I'm really quire new to this whole wiki-editing business, thanks for the pointers.--FractalP 06:32, 17 May 2011 (UTC)

  • Happens with everyone pretty much. No real worries. --Flargen 06:39, 17 May 2011 (UTC)

You're too fast!

on fixing the double posteffect error in unfortunate dumplings. I saw the cause when I went to edit the C.H.U.M. chum page, went back to the dumplings, and got a edit conflict when I went to save the fix. Damn ninjas. (OK, you can delete this if you want. ;) ) --Terion 02:15, 14 June 2011 (UTC)

Next-generation Frat Boy‎ history

As this page has been "moved" manually, the history of the page should really be moved over with it. Just to keep the past history of changes and so forth. Best way is likely to copy the text back to the old page, move the old page to the new (overtop of the new), and then restore the revisions. That will get the history properly moved over (for both page and talk page), and still leave us with the correct text on the page. -- sulfur 14:02, 7 August 2011 (CEST)

  • I believe both pages have a non-trivial history, is the thing. But I'll look into it and see if deleting one and the moving is better. --Flargen 03:04, 8 August 2011 (CEST)

link?

Hey, just wondering what the purpose of the link field you added to some Data pages was. Couldn't see any discussion that led to it, sorry if I missed such! --Starwed 18:31, 8 August 2011 (CEST)

Battle template special attacks

Special attacks no longer appear on pages at all - e.g. gluttonous ghuol.--Ryo_Sangnoir 14:17, 10 August 2011 (CEST)

  • Huh, whoops. The fix for that actually required I add #default for the default switch case. Which usually isn't actually necessary. Currently it will display nothing if you pass in special=?, which perhaps it shouldn't do, but I don't think anybody does this, anyway. Specials usually aren't assumed to exist before they've been identified. If it would be desirable to have it display the red "not known" in that case, it can be made to happen. --Flargen 00:39, 11 August 2011 (CEST)

RE:Mafia and ballroom songs

I know, but when I was doing the spading, I only used mafia for those four battles. I seriously doubt that it would have caught it in those four battles. --Deus Ex Machina (Talk) 00:24, 25 September 2011 (CEST)

Imagemaps

Since I have no clue how they work, I was wondering if you could add the imagemap data to the Secret Tropical Island Volcano Lairs.--Toffile 07:05, 15 January 2012 (CET)

  • I'll see what I can do about that tomorrow. The basic way I'd do it is to get the in-game source for those (by visiting it myself or trying to find someone who can); that will contain the necessary data to form the image map, and then you either infer what to do by looking at ones I already put on the wiki, or do a proper lookup on imagemap usage. --Flargen 09:11, 15 January 2012 (CET)

On the 0 turn thing

Is it really useful to showcase that glitch on those effects pages? It has no practical effect in the game whatsoever; it's just a glitch that'll probably be fixed someday. IMO it's most sensible to leave it out. --Improv 08:46, 29 January 2012 (CET)

Template:Combat/data

Is the template Combat/data really needed anymore? I looked at some old discussion and talk pages, and I found that it was created to help pre-load data pages. However, it looks like it's broken right now, and it doesn't seem to serve any purpose. Plus, no major pages link to it, except itself. I was going to delete the template, but I thought that I'd check with you first. Should it stay, or can I nuke it? --Oneforfortytwo 17:54, 11 June 2012 (CEST)

  • It's used as a preload for a newly created monster data page. Whenever someone, say, starts creating a new monster page and uses {{battle}}, and battle doesn't find a data page, it will provide a link to create the data page and the link tells the wiki to pull in the code for {{Combat/data}} as a pre-load. Otherwise wiki editors (even ones highly experienced with monster data) would have trouble figuring out how to format the data page and what the proper data field names are. So it's important. You have to be a bit careful with templates that are meant for item/monster pages, as they have a fairly complicated--and sometimes self-referencing, which is what's happening here to make you think it's broken--overarching structure. Kind of a fun exercise to track them all down and understand what's really happening, though. --Flargen 22:32, 11 June 2012 (CEST)

Hot Egg .1% drop rate

Estimates in /hardcore were .02% to .01%. How'd you get .1%? --Raijinili 23:39, 22 June 2012 (CEST)

  • How did they get .01 to .02? That's 5 to 10 times less likely than any other drop in the kingdom. The melange and cursed pirate drops are .1 to .2. It's still a bit conjectural on my part, too, though. .2 is still pretty reasonable. Hard to get enough data points to sufficiently narrow down the value. Otherwise I base it on my drop rate spading. --Flargen 23:53, 22 June 2012 (CEST)
    • I think they were looking at the data people were posting. Though for something this unlikely, one person can't really generate enough data to get something solid, unless you run like +900% item or something. --Raijinili 08:00, 23 June 2012 (CEST)
      • The problem with the data that people are posting is that it's a biased source of data. They're more likely to report data when it's exceptional, and in particular don't declare their intent to collect and post data before actually doing so, and that makes it useless for rigorous and proper analysis. Also, most of that data lacks detailed reporting of the amount of +items run (if they say +400% items, they might really mean mafia said +403.80% with a fairy type leading to a decimal, and they're just rounding), which in many cases means no such information at all. The posts with both grass and egg data can be used without that to establish what the ratio of the drop rates are, but without the +item bonuses you can't actually analyze the accuracy of that data, since the second moment of that ratio depends on the +item bonus non-trivially. And the variation on the ratio grass/egg is going to be fuckhuge (or fucksmall for egg/grass) thanks to the tiny drop rate on the egg leading to a tiny variation, which makes the average ratio stupidly hard to pin down anyway. Probably harder to pin down the ratio accurately than it is to pin down the egg rate directly, but I've not run the calculations. I would not say that the .1% I posted is conclusive, though. As with most fractional drop rates, eventually it's just too much of a pain in the ass to be accurate (especially since you don't know what granularity to accept; only accept integer tenths of a percent, or what?), and the spaders just call it good enough when it looks kind of like, say, .1, even though technically .2 or higher is still tenable --Flargen

A few more days, a few thousand more turns. Now it looks like it really is in the .02% neighborhood. That's insane. --Flargen 09:48, 25 June 2012 (CEST)