User talk:RolandOfGilead13141

How do you know all that skill data for level 18? are you testing it at the proper shrines? Foo 08:34, 9 September 2006 (CDT)
 * Oh ok, now that I see the attributes for the level 18 and 20, it seems you do. gj. Foo 08:36, 9 September 2006 (CDT)
 * No I don't gather hard data for the highest 2 levels. However, skill effects seem to follow a strictly linear curve depending on its linked attribute's level, taking GW's rounding method into account. For levels 0-16 I have tested this theory extensively and have not yet found an exception to it, so I have full confidence that the theory is correct and can be applied to levels 17/18 and 19/20, respectively. RolandOfGilead 08:49, 9 September 2006 (CDT)
 * ok, that sounds better then just guessing numbers. can you say more about it? Foo 09:06, 9 September 2006 (CDT)
 * Sure. The rounding method applied in GW is: x.5 and greater is rounded up, otherwise down. For example, 5/2 = 2.5 = 3 in GW. There are exceptions, but AFAIK they are alwas explicitely stated.


 * The linear curve I mentioned is the basic equation describing a line in a Cartesian coordinate system:


 * y = mx + c, or, for our purposes,


 * effect = effectIncrease*attributeLevel + offset


 * where effectIncrease is the increase per level and offset simply the value at level 0 (but before rounding, so offset can be a fraction, even though a fraction is never displayed to the player).


 * For example, if the skill says "target foe is poisoned for 5-17 seconds", then effectIncrease = 1.0 and offset = 5.0. This is trivial if the increase is a whole number. But what about fractions?


 * Example: Life Siphon effect duration, from level 0 trough 16, and the increase per level below it:


 * 12 13 14 14 15 16 17 18 18 19 20 21 22 22 23 24 25


 * N/A +1 +1 +0 +1 +1 +1 +1 +0 +1 +1 +1 +1 +0 +1 +1 +1


 * Do you see the pattern in the lower line? The increase is always 1/1/0/1/1. You could also say it is 0/1/1/1/1 or 1/0/1/1/1, as long as the pattern returns without fault. Either way, now we know that the effect will increase by 4 every 5 levels. Thus, effectIncrease = 4/5 = 0.8. The only thing left to do is to find out the exact value of offset, which can be determined by what my Maths teacher would've called "a sharp look". In this case, offset is 12.0. Now we can calculate the duration of Life Siphon for an arbitrary level. Hope that helped. RolandOfGilead 10:01, 9 September 2006 (CDT)


 * Hehe, maybe I should have said I'm a mathematician first. it is still a guess, an assumption that the line has nice and round values. how short should a pattern be so that you will consider it reliable? there was a discution about it already, and I think it said that we will only put in values confirmed in-game. (energy cost, after experties and other effects is always rounded down). Foo 11:40, 9 September 2006 (CDT)
 * As long as we don't get to see the source code of the game, every formula and every rule with predictive power the community deduces about game behaviour is an "assumption", unless confirmed by ANet, so "assumption" is not a very useful term to begin with here. But, staying with the term, all assumptions are not alike; there are good ones and bad ones. The good ones are founded on large amounts of data, can explain all data and are not contradicted by even one piece of data.


 * The theory of strictly linear skill progression is a good assumption in the above sense, because it fulfills all these criteria. Feel free to disagree, but to the best of my kowledge (and I have put a lot of time into this), every single one of the hundreds of skills in GW follows the linear curve in levels 0 to 16, which is 80% of the 0-20 range which currently matters. I do not exaggerate when I say that I have checked this behaviour for every (!) unique skill range in Prophecies skills, i.e. 5-17, 20-50, 1-10, etc. etc. ad infinitum. If you want proof, I can upload a csv-file which is readable by the spreadsheet program (excel, OOo calc, etc.) of your choice. As long as there is no shred of evidence contradicting the theory, it would be plain unreasonable to dismiss it in favour of dull data collecting without generalization, just because "it could be different just 1 level further". Yeah, but so what?. All evidence points towards linear progression without any deviation and none speaks against it, so why not go with it? If it should turn out wrong later, ok, let's dump the theory with or without replacement, that's how it goes, no problem at all. But until then, this theory provides an equation which not only correctly describes all known skill values, but also has potential predictive power for non-observed level ranges. Who is afraid of a good assumption which has a non-zero, but imho small, chance of being wrong? RolandOfGilead 13:03, 9 September 2006 (CDT)


 * Progressions are linear through the values for 0 and 15, as in a line through (0, f(0)) and (0, f(15)). Round non-integer results.  Using 0 and 16 doesn't work (Aura of Restoration is one example). See Talk:Skill_details/Archive.  --Fyren 15:46, 9 September 2006 (CDT)
 * Ok, I stand corrected, thanks for discussion. You could've saved me all that by just linking to the formula :) Anyway, at least I was right about the linearity of the function, because the one in Skill details is linear as well, only the coefficients are a ittle different. For example, for Aura of Restoration the equation is y = 152 + x*16.5333333, versus my y = 152 + x*16.5. Obviously the difference is very small but big enough to be noticed, if only in a very few cases. Also, if (v15-v0) is a multiple of 15, then my approach gives the same results, so my method is really a special case and an approximation. RolandOfGilead 17:42, 9 September 2006 (CDT)


 * I'm not Foo (if you're mistaking me for her, heh). I just happened to see your discussion with her so chimed in.  I also have some wiki templating written up that auto-generate progressions, but I don't know if we're going to use it.  See, for example, Sandbox/Symbiosis and the source for it.  The values for 0 and 15 are used to generate the progression table automatically, so hopefully we can ditch all this manual progression data crap.  (We didn't do this from the start since ParserFunctions are needed to do the math, and PF weren't installed till a few weeks ago.)  --Fyren 18:12, 9 September 2006 (CDT)


 * "her"? how did you get to that conclusion? Foo 18:20, 9 September 2006 (CDT)
 * If you insist, I'll use "it" in the future. --Fyren 18:22, 9 September 2006 (CDT)
 * :O Foo 18:25, 9 September 2006 (CDT)
 * Hm guess my brain went into auto-mode when it saw a short name starting with F hehe. Anyway, the Sandbox template looks good, if not complete. What's missing? BTW, yeah I know we didn't have that. I used to be user:Roland of Gilead, but my password got lost after a longer break...Skills have always been my favorite topic :) RolandOfGilead 18:43, 9 September 2006 (CDT)


 * The template is "complete," but see GuildWiki talk:Sandbox/Skill box. The gist is I changed some things to how I personally think it should be but I haven't discussed the changes with anyone, really, so I'm not going just to switch all the skills over.  --Fyren 18:53, 9 September 2006 (CDT)

New window
It's only temporary for the event; it'll be gone till Nightfall is out. So, I dunno. Don't make so many changes that you can't change it back easily. --Fyren 09:49, 22 September 2006 (CDT)
 * Do you know that for certain? Did they announce they will revert the interface changes? Seems pretty pointless to me, but if you got solid info, I'll revert it RolandOfGilead 09:52, 22 September 2006 (CDT)
 * It's in the patch notes. --Fyren 09:53, 22 September 2006 (CDT)
 * Ok, then I would leave the skills and attributes article as it is and let it stand as long as it is correct, i.e. for the duration of the NPE. When the NPE is over, I'll just bend the links back to the skill inventory article as it was 30 minutes ago. Then, when NF goes live, we can again just bend the links to skills and attributes window. Sound like a plan? RolandOfGilead 10:05, 22 September 2006 (CDT)
 * I'm okay with whatever, as long as you're doing the work! --Fyren 10:07, 22 September 2006 (CDT)
 * By the way, Skuld left a 'higher standard' notice on the Skills and Attributes Window page but apparently he doesn't want to tell me what's wrong. Could you comment on what's wrong then? Also see the discussion page. RolandOfGilead 10:16, 22 September 2006 (CDT)

Redirects
GW:REDIRECT says to not make redirects for different cases of article names. --Fyren 12:18, 26 September 2006 (CDT)
 * well sorry, didn't think this is against official policy. Still, it seems such a pointless policy, because searching for an article without capitalization yields unpredictable results. Sometimes it leads directly to the corrrectly capitalized article, which is fine. Usually you get to the search page with the correct article being the #1 hit; still acceptable, but requires another click and page load, and GW isn't the fastest site out there most of the time. Rarely, you won't even get that and the one searching has to assume the article doesn't exist for some reason or whatever else, which is pretty bad and counter-productive imho. At least the last case should be caught. RolandOfGilead 12:28, 26 September 2006 (CDT)


 * I'm just telling you why I deleted it, since you wouldn't have made it if you knew. --Fyren 12:43, 26 September 2006 (CDT)
 * No offense taken. But what do you say about the policy? After skimming GW:REDIRECT (edit:aha, capitalization error :), the discussion there seems to revolve mainly and repeatedly about capitalization, so wouldn't another discussion started by a de-facto newcomer just be annoying? RolandOfGilead 12:46, 26 September 2006 (CDT)


 * In my comment on the talk page there (that I posted while incognito as 68.142.14.19) I say that I'm ambivalent. For all the cases we don't allow, Wikipedia does in order to help searching.  Better searching is obviously good.  But it'll also mean tons of people will link (within the wiki) to those redirects not knowing which is the real article and that's a pain both because they'll use the wrong capitalization and because they should be linking to the real page and not the redirect (it just happens that both would be fixed at the same time with the right link).  --Fyren 13:09, 26 September 2006 (CDT)
 * Is this a pure style question then or am I still missing something? Assume someone links to signet of might instead of Signet of Might. If the former has a redirect to the latter and someone klicking on the former gets to the article, would that not serve the link's purpose, namely to provide a shortcut to the intended article? Obviously, forcing capitalization in wiki links makes the text more readable and is desirable, but it harms the search feature, so I can see a certain conflict of interest there. However, fixing style in articles (i.e. correcting capitalization even though redirects exist) seems easier and more to the wiki spirit than fixing the search function. RolandOfGilead 13:25, 26 September 2006 (CDT)
 * I suppose it's style issue where they decided to protect style and make searching a little more difficult. I didn't follow that discussion and vote closely (or the older ones, which I'm not sure where to find off the top of my head; this has been brought up a few times).  There's GW:ULC which says to use lowercase/normal English case for everything but names of things in game that ANet capitalizes, so "signet of might" should not be used anywhere in an article, whether link or just plain text.  That issue I'm not ambivalent about: ANet's capitalization sucks and it pains me to write things capitalized like they do, whether skill names or whatever. But more than a year ago when I first argued against it I got shot down.  If ULC didn't exist as it does I would want to allow all-lowercase redirects, at least.
 * Specifically about search problems, it only breaks in the case of a title that contains a word starting with a lowercase and a word starting with an uppercase that's not the first word. So if you search for "blacktide den (mission)" you end up at Blacktide Den (Mission), but stuff like "signet of x" breaks since the "x" is capitalized in the article name while "of" isn't.  In most cases, the right article is the first result.  I'm not sure why "signet of might" isn't, I don't know enough about how MW's search works at the moment.  --Fyren 13:39, 26 September 2006 (CDT)
 * Exempting only article titles from ULC for the sole purpose of lower-case redirects is something I would consider a small change but allow for fixing working around the search function misbehaviour. That would be a great benefit for a small compromise. But anyway, if the consensus says otherwise, well, it's still nonsense but I'll get over it. RolandOfGilead 13:56, 26 September 2006 (CDT)

R/D Toxic Spammer
Hiya there, I moved it out of Category:Untested builds for you as user pages shouldn't be in there. You may of course readded if you move it back into the main namespace when Nightfall comes out ;) Thanks. --Xasxas256 22:58, 26 September 2006 (CDT)

Dye icons
How'd you get such a clean cutout of the new dye icons?
 * simply took screenshots, followed by a quick & dirty MS-Paint job. --RolandOfGilead 13:33, 26 October 2006 (CDT)

Profession color templates
Not sure if you saw GuildWiki talk:Style and formatting/Profession colors‎, there's talk of the templates to use for profession colors. --- Barek (talk • contribs) - 13:28, 29 January 2007 (CST)


 * Just curious - what's the intended purpose of Template:M-color? It looks to contain the same codes as Template:Mo-color, or is "M-color" still a work in process? --- Barek (talk • contribs) - 17:44, 29 January 2007 (CST)
 * Never mind that, i thought that was the missing monk-template, I forgot that it was Mo, not M. Delete please. Then please watch Talk:W-color, I have noticed severe performance problems with the x-color templates. I'll have the write-up done in a couple minutes. --RolandOfGilead 17:47, 29 January 2007 (CST)

sig
I saw the sig. FYI: You may want to review GW:SIGN. The sig should contain either your name, a section of your name, or an image that conveys the meaning of the name (ie: an image of a rocket for a user named Rocket, or a user of a gem for a user named Gem, etc).

Also, to get the sig to show when you type ~, go to preferences, check the "Raw signatures" box, then in the box "Custom signature" enter.

Let me know if you run into any problems. --- Barek (talk • contribs) - 18:19, 30 January 2007 (CST)
 * Thanks. I was getting to that, just wanted to make everything work first --[[Image:Roland_icon.png]] (talk) 18:27, 30 January 2007 (CST)


 * Hm Could you tell me why the link to my talk page is fine in the /sig article, but not if I actually include the sig somewhere? Or is that a caching issue? --[[Image:Roland_icon.png]]Roland of Gilead (talk) 18:31, 30 January 2007 (CST)
 * It looks okay to me - what are you seeing is wrong? --- Barek (talk • contribs) - 18:33, 30 January 2007 (CST)
 * Oh, if you mean not being a link here, it's because you're on the talk page already, so the link just bolds. --- Barek (talk • contribs) - 18:34, 30 January 2007 (CST)
 * Ah yes, I post too fast. I remembered 3 seconds after hitting Save Page :) BTW, do you, for one, see what my image icon is supposed to be? --[[Image:Roland_icon.png]]Roland of Gilead (talk) 18:35, 30 January 2007 (CST)
 * I see a red dot over something that I can't see clearly enough to recognise - so not sure. --- Barek (talk • contribs) - 18:44, 30 January 2007 (CST)
 * Ah yes, I never was much of an artist. Maybe I'll try to improve upon it... --[[Image:Roland_icon.png]]Roland of Gilead (talk) 18:49, 30 January 2007 (CST)
 * Sadly, a limit of 19px in height makes it difficult to make much of anything recognizable. --- Barek (talk • contribs) - 18:56, 30 January 2007 (CST)
 * True enough. Making minimalism work is the true mastery, for every craft. --[[Image:Roland_icon.png]]Roland of Gilead (talk) 19:02, 30 January 2007 (CST)
 * Note: I saw your post on LordBiro's talk page. I was wondering if the one image was a gun of some sort - but the other object meshed with it a bit, so I wasn't certain if there were two objects (one being the revolver), or if it was actually just one object that I wasn't able to recognise.  --- Barek (talk • contribs) - 19:04, 30 January 2007 (CST)
 * Would you say that version is an improvement? --[[Image:Roland_icon.png]]Roland of Gilead (talk) 19:30, 30 January 2007 (CST)
 * Yes, I can make out both the dolphin and the boomerang clearly now.
 * Joking. Actually, yes, the revolver and rose are clear now. --- Barek (talk • contribs) - 19:33, 30 January 2007 (CST)
 * Heh, you almost got me. How despicable, undermining my feeble artistic (or was it sadistic?) efforts like that. --[[Image:Roland_icon.png]]Roland of Gilead (talk) 19:39, 30 January 2007 (CST)

Sorry
Didn't realize we used the 1..12 values in the skill descriptions. Guess I should have as I look at them all the time :P. Thanks for fixing it. Turk Nagona 19:40, 1 February 2007 (CST)

Word of Healing
FYI, I reverted a change you made to Template:Word of Healing, the previous edit was correct according to the guildwars update page. --Horizon Blue 22:53, 1 February 2007 (CST)
 * Thanks for the note, but my edit was correct indeed. ANet uses 0..15 attribute range, we use 0..12 instead. You can also check in-game. Meanwhile, I'll restore my version. --[[Image:Roland_icon.png]]Roland of Gilead (talk) 23:01, 1 February 2007 (CST)
 * Ah, my bad. Hmmm, I wonder if it would be possible to have that mentioned when editing a skill template, I imagine I'm not the first person to make that mistake. --Horizon Blue 23:29, 1 February 2007 (CST)
 * I wanted to answer that with Style and formatting/Skills or Skill templates, but I have to admit it doesn't seem to be in these articles or in their talk pages. It's probably considered knowledge too common to be mentioned explicitely, but I'll ask the admins whether that's official policy, written down somewhere. Thanks for the hint. --[[Image:Roland_icon.png]]Roland of Gilead (talk) 23:43, 1 February 2007 (CST)

Border vs. Background
I know you had an interest in the profession color topic. You comments on GuildWiki_talk:Style_and_formatting/Profession_colors would be appreciated if you have any objections. --Zampani 14:04, 11 February 2007 (CST)

Lockpick analysis on Luck titles guide
Hey. I stumbled upon your addition to this article, and can't for the life of me work out how you arrived at the "extreme" values for achieving the max lucky title. Could you please shed some light? "At Treasure hunter rank 0 and only picking Hard mode chests, 26,300 chests have to be opened and 24,450k have to be spent." Surely if you buy a single lockpick and it never breaks that would be 1.5k cost? If we're going for extreme. Additionally, are the calculations for the "new" max of lucky? Thanks in advance for any clarification you can provide! --BlueNovember 13:08, 28 April 2007 (CDT)
 * Yeah theoretically you could get max lucky title with a single lockpick, but that's most most most probably not gonna happen. In fact, even if you started picking chests at 98% retention rate, then the probability for retaining it 26,300 times is a number with 231 zeros after the decimal point. Spelled out, it would be a
 * 0.0000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000176%
 * chance.


 * If it does for you, you should consider playing the lotteries instead, you would probably win the next ten drawings in a row before you can use a single lockpick to open 26k chests :) After all, the other extreme value would be that you never ever reach any lucky points at all if your lockpick ALWAYS breaks, but that's not gonna happen either over a long time.


 * All analyses in the Luck titles guide, including the Boardwalk games, have to be taken as statistical statements. This means, virtually nobody will experience exactly these numbers; ca. 50% will be above them, the rest below, but if you take the mean value of the many ten thousand players going for chests or the Ring games, then that's about the values you will get. That's the power of large numbers for you.


 * In short, statistics of this kind are almost never true for the individual or over short time spans, but almost always for a large mass of individuals and a long time. That's the reason why places like Las Vegas work; an individual player might become a millionaire, but on average and in the long term, the house wins. If you ever have the opportunity to take a statistics course in school or whatever, take it. It's a most useful topic which can help you a lot in real life too, not just GW.


 * As for how I calculated it: I just wrote a simple java program that simulates opening chests. As I have written in the analysis, on average every chest gives you 250*R lucky points and costs 1500*R, where R is the chance to retain the lockpick, and every opened chest gives you 1 point to the treasure hunter title. From there, it's just a matter of keeping track of chests opened and lucky points gained and gold spent, and whenever the virtual character gains a rank in either treasure hunter or Lucky title, you update R and continue until you arrive at max lucky rank, at which point we're done. If you vary the starting conditions (See Lockpick for details of the calculation of R), you can get the most extreme statistically expectable values. And yes, of course the calculations are based on the latest max level "Blessed by Fate". --[[Image:Roland_icon.png]]Roland of Gilead (talk) 17:40, 28 April 2007 (CDT)

Image copyright problem with Image:Lockpick lucky point price.jpg
Thank you for uploading Image:Lockpick lucky point price.jpg. GuildWiki takes copyright very seriously, the image you have supplied is missing information on its copyright status. The image will be deleted after 7 days, unless the copyright status is determined for the license and the source of the image. Please add a copyright tag to the image description page.

Thank you for your cooperation. If you are the creator of this work, please indicate on the file page. 84.13.251.42 05:20, 3 May 2007 (CDT)

qr on skill lists
quick list looks better above the toc. -- Xeon 13:36, 20 May 2007 (CDT)
 * Yup, thanks, wanted to do that anyway but came out wrong at first. Inserted at the wrong line probably. --[[Image:Roland_icon.png]]Roland of Gilead (talk) 13:45, 20 May 2007 (CDT)
 * missed ranger :P -- Xeon 13:47, 20 May 2007 (CDT)

ehh, why did you move that back
you should of brought it up on the talk page before you moved it, also i moved it to sunspear skill because it has data and not just qr information. -- Xeon 04:52, 24 July 2007 (CDT)
 * Sorry, I didn't know it was already reverted. I have moved a good dozen or so QRs, so I thought I had somehow forgotten about that one. --[[Image:Roland_icon.png]]Roland of Gilead (talk) 13:05, 24 July 2007 (CDT)