Talk:Luck titles guide

No reason for article
I don't see a reason for this article. To me, I think we should just migrate any new material found here to the Rings of Fortune, Nine Rings, and/or Titles articles, then delete this one. --- Barek (talk • contribs) - 11:08, 1 July 2006 (CDT)


 * I can see the merits of this article, but I agree that it should probably be merged into either Titles or the individual event articles  &lt;LordBiro&gt;/&lt;Talk&gt; 11:26, 1 July 2006 (CDT)
 * Biro, you said it better than I did: I can see merit in the contents, just not in having it as a unique article. --- Barek (talk • contribs) - 11:27, 1 July 2006 (CDT)
 * It can be removed on July 5th. Might as well keep it comprehensive while the event's still going on. --Arch Cuisinart 23:39, 1 July 2006 (CDT)
 * I don't think it should be merged because it's a pretty big table, I think it would add needless clutter to the titles article. I don't think it should be deleted either, since there's probably going to be a 2007 dragon festival (the lore describes it as an annual event), and there's a decent chance that they'll return then, making the information useful again. It's also possible that they'll add games of chance in other holiday events. -- Gordon Ecker 22:04, 4 July 2006 (CDT)
 * Everything together is more convenient, the page isn't big, and it helps to compare what each option brings/cost without having to flip pages. If anything the Rings of Fortune and Nine Rings articles could be removed (this page replaces both of them advantageously). 83.159.9.78 14:18, 20 October 2006 (CDT)

Math Redo
someone needs to redo the math, since since the center of nine rings actually give 25 as reward for winning while the corners give 55. Alexanderpas Talk 12:22, 1 July 2006 (CDT)


 * These numbers don't match up with mine. For one thing; Nine Rings is running at 19 games per cycle, not 20.  This may be new, but it's what I've observed over the last two days.  I'll post my numbers later this afternoon.  Also, there doesn't seem to be any mention of the fact that if you're going for one of the unlucky titles, then it DOES make a difference which ring you stand in (for Nine Rings). Dareya 13:11, 1 July 2006 (CDT)


 * Okay, I take that back. He does mention playing the corner in Nine Rings (for optimal losses). Dareya 13:38, 1 July 2006 (CDT)


 * What about for achieving both titles? Which ring would you stand in? (I'm currently spending the 9 hours in Rings of Fortune). EDIT: Nevermind. The net-gain in the long-run is the same regardless of which circle. I belive the guide for optimizing time-efficiency and cost requires the corner.--Arch Cuisinart 13:42, 1 July 2006 (CDT)

An anon keeps putting in false information. Do the math for nine rings if you have one friend stand in each circle: So no matter what, having one friend in each cicle will cost the collective group 5 tickets each game! --Thervold 17:07, 5 July 2006 (CDT)
 * corner:One player loses 10 and gains 55, two lose 10 and gain 15, six just lose 10 = 45+10-60=-5
 * side:One player loses 10 and gains 40, three lose 10 and gain 15, five lose 10 = 30+15-50=-5
 * center:One player loses 10 and gains 25, four lose 10 and gain 15, four lose 10 = 15+20-50 =-5

Having done the Math one day whilst bored at work (for nine rings), I worked out that in terms of net gain/loss of tickets, it didn't matter where you stood (assuming you never move rings), as the overall balance was exactly the same. However, the game does not go by ticket wins/losses or games won/lost, it goes by tickets won vs Games lost, so therefore the best place to stand was in the corner for both titles. Also I got 33 days for Max Lucky title, and 60 days for Max Unlucky. If you want to see my workings, search the trash bin at my place of work :P

I don't agree with the statement "This maximizes the time-efficiency and minimizes cost as well." in the Achieving Both Titles section. It does minimize cost, but it is not more time-efficient. I haven't done calculations, so I don't want to change this page, but simply by looking at the tables, one can see that you win tickets and lose games quicker by standing in the 9-ring corner than you do anywhere in the 16-ring game. Gwm 16:04, 20 October 2006 (CDT)


 * You'll lose 9/16 (0.5626) games on average in Rings of Fortune. Standing in the corner in 9 Rings, you'll lose 2/3 (0.6667) games on average. (In the middle, you'll lose 4/9 (0.4444), and in the side you'll lose 5/9 (0.5556).) I have no clue how to write that into the article though. - [[Image:Savio_3_mini.gif]] Savio  19:09, 20 October 2006 (CDT)


 * I think it should be enough to just revise the quoted sentence to read something like "This minimizes the cost of achieving both titles, however it increases the expected time investment since the Rings of Fortune game has a slower rate of compiling tickets won and games lost than the corner of the 9 Ring game." I don't think you need to add all the detail that you mention, since it's in the table.  But at least that one existing sentence should be corrected. Gwm 15:48, 21 October 2006 (CDT)


 * I added a more accurate description of what the table really shows. Which is the amount of time you need to spend on each game in order to reach the same level of titles at the same time. If you want the titles in the least amount of time, play Nine Rings. If you want them cheapest, play Rings of Fortune. If you're going for the top level in both, you will gain the Golden title alot sooner than the Hated title no matter where you play. (But 357 hours earlier, and ~4,580,000 more expensive playing Nine Rings) --Rydier 21:55, 21 October 2006 (CDT)

The center of the Nine Rings games only gives you FIFTEEN Tickets, not 25...should this be corrected?
 * No, it still gives 25 for a direct hit, and 15 (like all other rings) if it's adjacent to the winning ring. &mdash; [[Image:Fin_sig.gif|User:Kyrasantae]] kyrasantae  21:23, 21 October 2006 (CDT)

Ah, thanks for the clarification...Woops.

People are messing with the master title requirements it seems. Currently the cost for the unlucky master title with Rings of Fortune is now 1,250,000 gold and I think it should be 333,333 gold? 09:42, 24 February 2007 (CST)

The Cost in time is just fine...
I don't know where the author(s) went wrong, but I know the provided math is not correct. I started playing yesterday at 12:30 noon, now it's 11:10 AM (so just under 23 hours) and I have not played in the luck games all those 23 hours (did actual playing for 2+ hours). However, I moved from 150,000+ to 220,000+, that is 70,000 in less than 23 hours. So, the math provided is certainly flawed. (I always play Corner in Nine Rings). --Karlos 13:12, 21 October 2006 (CDT)
 * compared to what? that sounds about right... --Lemming64 13:47, 21 October 2006 (CDT)
 * 3,454 tickets/hour * 21 hours = 72,534 tickets. You're actually doing a bit worse than average. - [[Image:Savio_3_mini.gif]] Savio  13:56, 21 October 2006 (CDT)


 * Yeah, ignore me. I have no idea how I reached the conclusion that it was slower in the article. I thikn I was looking at the wrong column or wrong table. --Karlos 14:02, 21 October 2006 (CDT)

Gaining Both Titles
If Nightfalls is going to take titles into account for game play (as shown here: Title), would it make sense to add a note that it might not be advantageous to achieve both titles? I am wondering if lucky/unlucky will have an impact, and if so, might having both titles cancel out one another. If lucky improved your gold drop loot drop rate but unlucky decreased it, having both titles might be disadvantageous. I realize this is pure speculation at this point, but it might warrent mention.
 * I highly doubt they'd implement a negative effect for a title. Especially since: (1) You can't remove it. (2) It takes 3 friggin months of AFK'ing to get it in the first place, leaving an account gimped after that much gold and time spent would just be stupid beyond reason. --Rydier 07:44, 22 October 2006 (CDT)
 * Not to mention that the way it works is by equipped title gives the bonus. If it does have a "negative" (hard to call any effect negative in a game where a -50hp item is precious) effect, then just don't equip it.
 * The way what works by equipping it? Salvage titles don't. The only titles that require equipping are the Lightbringer ones. -76.166.23.65 21:48, 23 February 2007 (CST) (scyfer)
 * Since when does it take "3 friggin months of AFK'ing" to get the title? The page estimate is what, 437 hours?  There are 168 hours per week, so that's less than three weeks of total consecutive time.  That only adds up to more than three months if you do it in small (e.g. overnight) chunks ... and since the Boardwalk is only open if-and-when, that will take a slice of forever.  Besides, you'd be AFK while sleeping anyway so that doesn't even count.  Me, I figure when the Boardwalk is open I get some much-needed time away from the game.  (I discovered I was married.)  More to the point, let's talk in terms of weekends, because so far that's what they've done.  You can conceivably play 24/7 over the course of a weekend event, if you do nothing but gamble, run to the ticket seller during the speeches, and never lose your connection.  So, 437/72=about six weekends of AFK time... not that big a deal. Auntmousie 22:12, 23 February 2007 (CST)
 * They only run the events for 48 hours at a time. 12 pm friday-12pm sunday, and total time doesn't change no matter what time zone you're in, just start and end times. So 437/48 = 9-10 weekends under your assumed conditions.--69.3.204.230 22:37, 24 February 2007 (CST)
 * Actually, events tend to run for two and a half days (from Friday 12:00 PM until Sunday 11:59 PM), which is 60 hours per weekend. I'll leave it to you to calculate how many such weekends the max titles would require by that assumption. Quizer 12:29, 6 March 2007 (CST)

Gaile's Interest
Should it be noted that Gaile Grey has expressed interest in having permanent boardwalk games (stating she would enjoy whack-a-worm)?


 * I do not recall Gaile saying this. Besides which, the Shing Jea games would come back for Dragon Festival (if possible).--Dark Paladin X 10:39, 29 April 2007 (CDT)

Suggestion: Calculator
Would a calculator app be of any value? Eg, "Enter your current and target lucky/unlucky scores," crunch some numbers, throw out a few tables: fastest, cheapest, and the compromising point of interest (furthest point with the cheaper time-money tradeoff, assuming it exists. This idea is still in its inception). Tables contain expected time in each of the games, expected cost in time and money. Perhaps provide a target range for lucky/unlucky scores (at least X, at most Y), for people who want to have specific titles rather than as-high-as-I-can-get ( may require a little stats -- "I can guarantee with 99.9%+ certainty that you will not exceed lucky score X to obtain unlucky score Y" under whatever conditions are reasonable. But I don't want the project to explode in complexity. But, hey, if someone else wants to do it~ )

Implementation: Javascript (hosted local, nonlocal), VB.net, Java..?

I may or may not pick up the project myself--I was feeling abnormally productive for a minute here, but I'm sure that will pass. >_> --Bob III 19:00, 25 February 2007 (CST)


 * Brilliant idea. I like the novelty of it. You could probably program it pretty easily with Java or .NET frameworks, but I don't think the Wiki, in it's current form, could host an app like that. If you made it, I'd recommend posting it on a download site and posting a link to it on places like GWOnline or GWGuru. FlameoutAlchemist 00:13, 26 February 2007 (CST)

Presto! User:RolandOfGilead/Java/Luck Titles Calculator. Run it offline or make it a simple webform if you like. --Roland of Gilead (talk) 21:16, 9 March 2007 (CST)

Lock Picks
So lock picks add to the title now maybe some notes should be made on lock pick vs game chances of getting the titles
 * Since buying tickets advances your Lucky title by a flat rate of 0.882 gold per point, it can be calculated that to get the same economical efficiency out of lock picks, you need a retaining rate of 85.3%, only available for very highly ranked characters. But of course lock picks are there all the time unlike the boardwalk, and you get some (minimal) compensation from the low-level chests needed to reach such a rate.--Tmakinen 10:26, 27 April 2007 (CDT)
 * Coming from a functional standpoint (ie dont care about titles but now it effects gameplay) is it worth going out and getting the lucky title through tickets on a return basis? That's the crux of the table I'm working on below --CKaz 14:00, 27 April 2007 (CDT)
 * How to read this table - amount of lockpick usage attempts to make back luck title costs (nine rings) after making listed title
 * Basically every progression effectively saves you 30g a lockpick (1500x.02) - again pure plat cost effective look
 * {| border="1"

!Lucky !!colspan=2 | using Nine Rings
 * Charmed
 * style="background:#EFE" | 1470 lockpick uses|| style="background:#EFE" | after 44.1
 * Lucky
 * style="background:#EFE" | 1470 lockpick uses || style="background:#EFE" | another 44.1
 * Favored
 * style="background:#EFE" | 4413 lockpick uses || style="background:#EFE" | another 132.4
 * Prosperous
 * style="background:#EFE" | 7353 lockpick uses || style="background:#EFE" | another 220.6
 * Golden
 * style="background:#EFE" | 14706 lockpick uses || style="background:#EFE" | another 441.2
 * Blessed by Fate
 * style="background:#EFE" | 44120 lockpick uses || style="background:#EFE" | another 1323.6
 * }
 * Synopsis? This isn't critical at any level and reward tails off sharply as you go up.  First couple are the only ones that might make sense for the casual IMO, even hardcore if you're not into title collection for the fun or AFK of it --CKaz 14:07, 27 April 2007 (CDT)
 * [Also I didn't feel the need to re-visit time figures or consider lockpick adds to luck, covered elsewhere, wouldn't add much]
 * How do discounted (1.2k) lockpicks figure in? A profit can now be made by buying 1.2k lockpicks and reselling them for 1.4k. The seller profits and those who don't have access to discount merchants profit as well.   [[Image:jkr.jpg]]     18:19, 27 April 2007 (CDT)
 * Well these are terrific however you slice it. 1% better usage saves 15g a pick, saving 300g with the 20% sales discount is just as good as it sounds, it's as good as +20% on saving your picks.  Of course that's just for those with elite access, for those of us masses who don't, even better if you can barter for lockpicks for 1250/1300/1350.  --CKaz 18:50, 28 April 2007 (CDT)
 * style="background:#EFE" | 44120 lockpick uses || style="background:#EFE" | another 1323.6
 * }
 * Synopsis? This isn't critical at any level and reward tails off sharply as you go up.  First couple are the only ones that might make sense for the casual IMO, even hardcore if you're not into title collection for the fun or AFK of it --CKaz 14:07, 27 April 2007 (CDT)
 * [Also I didn't feel the need to re-visit time figures or consider lockpick adds to luck, covered elsewhere, wouldn't add much]
 * How do discounted (1.2k) lockpicks figure in? A profit can now be made by buying 1.2k lockpicks and reselling them for 1.4k. The seller profits and those who don't have access to discount merchants profit as well.   [[Image:jkr.jpg]]     18:19, 27 April 2007 (CDT)
 * Well these are terrific however you slice it. 1% better usage saves 15g a pick, saving 300g with the 20% sales discount is just as good as it sounds, it's as good as +20% on saving your picks.  Of course that's just for those with elite access, for those of us masses who don't, even better if you can barter for lockpicks for 1250/1300/1350.  --CKaz 18:50, 28 April 2007 (CDT)
 * Well these are terrific however you slice it. 1% better usage saves 15g a pick, saving 300g with the 20% sales discount is just as good as it sounds, it's as good as +20% on saving your picks.  Of course that's just for those with elite access, for those of us masses who don't, even better if you can barter for lockpicks for 1250/1300/1350.  --CKaz 18:50, 28 April 2007 (CDT)

I don't follow the above numbers. Presuming you achieve the title then use lockpicks, and the cost of the titles on the 9 rings is 44.1/88.2/220.6/441.2/882.4/2206, I get:

(Example calculation: Golden. 882,400 / 150 = 5883.) Am I doing something silly? Edit: Presumes 1.5k lockpicks. WTS Lockpicks, 1.3k ea xD --BlueNovember 15:01, 30 April 2007 (CDT)

further notes
I happened to be one of the individuals 'rolled back' while AFK. What bothers me about this are three-fold - I was active for hours in the lottery area but those gains were removed as well, there has been no communication or annoucement regarding the roll-back or reason, and if it in fact is due to being AFK what kind of crazy title/game makes you be present for a very boring lottery game indefinitely?

There is the option to go after normal mode key gains to gain Luck, basically to the tune of 200 key saves. The % to retain is much higher on the cheapest normal mode chests, but still the effort will cost you about 150p each for the first couple levels and obviously take some time and the loot won't be much help here to offset the costs, something also to weigh vs more expensive chests. At some point I'll look to provide more of this detailed analysis/table unless someone else offers it up first. --CKaz 18:58, 28 April 2007 (CDT)

Don't oversimplify the math
Please don't remove info on how certain values are reached, specifically how the cost per Unlucky point is calculated. Saying that every broken lockpick gives 25 points is true, and the conclusion that the cost per UL point is constant is true as well, but it's an incomplete argument, because neither the chance of breaking a lockpick nor the price per chest is a constant. Only from the complete calculation can we see that the price per Unlucky point is indeed constant because the "(1-R)" terms cancel each other out. Also, "1500 / 25 points" is not a statement which can be shown true or false but just a lonely number. Reverting on that basis. --Roland of Gilead (talk) 05:19, 29 April 2007 (CDT)


 * The fact is, you gain unlock points from lockpicks ONLY WHEN your lockpick break, regardless of the probability. Only broken lockpicks give you unlucky points, and every broken lockpick cost you exactly 1500 (I didn't decide to ignore discounted prices, the version before me did, and every broken lockpick gives you exactly 25 Unlock points.  Now, tell me, how does the probability of lockpick breaking or retaining have to do with the gold cost of unlucky point?  Nothing!  On that basis, I petition to re-revert to my edit.  Don't overcomplicate the math by introducing variables that never needed to be there in the first place. -User:PanSola (talk to the [[Image:follower of Lyssa.png]]) 08:27, 29 April 2007 (CDT)
 * I agree with PanSola. -Auron [[Image:Elit Druin.jpg|19px||My Talk]] 08:41, 29 April 2007 (CDT)
 * I agree with PanSola as well.--Stevo101 08:44, 29 April 2007 (CDT)
 * The problem is, the same argument goes for the Lucky points: Every retained lockpick costs you zero gold, but that doesnt mean that lucky points are free, obviously. You can't just take the price of a retained lockpick (0) and divide by 250, and you can't take the price of a broken lockpick (1500) and divide by 25. --[[Image:Roland_icon.png]]Roland of Gilead (talk) 09:05, 29 April 2007 (CDT)
 * The question is not "how much gold is needed for getting those specific un/lucky points?", but "how much gold will I lose in the process?". in the process of getting unlucky point, you lose exactly 1500g per 25 points, and in the process of getting lucky points, your lose depends on your 'R', cause your goal of not-breaking picks is tainted by mishappens in which you DO break them. Foo 09:37, 29 April 2007 (CDT)
 * I'm well aware of how to calculate the price of Un/Lucky points. My point is that this calculation can be presented most simply and most coherent by just sticking to the correct formula, which is Average price per chest divided by Averge Un/lucky points per chest. There is no reason to leave out the (1-R) term in the first place. I thought you, as a mathematician, could symphatize with a little precision and with starting a calculation at step 1 rather than step 2 (where some terms have already cancelled out).
 * I do appreciate the elegance of a variable naturally dropping out of the formula, but in this case, it is equally precise and convincing to take PanSola's approach, and for the sake of the causal reader of this article, I'd rather keep the mathematical beauty out of it this time, and keep it as simple as possible. Foo 11:15, 29 April 2007 (CDT)
 * I'm still not convinced that leaving out the terms is more clarifying. After all, if the first formula had them, why doesn't the second? Doesn't that disparity imply that the the second formula never depended on R in the first place? There is obvious symmetry between the formulas for Un/Lucky points per chest (a number multiplied by a probability), but the symmetry is seemingly broken for gold per Un/Lucky point. I consider that more confusing than leaving the term in and then let it cancel out in an comprehensible manner. If that helps, I will tex the formulas and take a screenshot, for lack of tex support on guildwiki. Give me half an hour. --[[Image:Roland_icon.png]]Roland of Gilead (talk) 11:27, 29 April 2007 (CDT)
 * Let me express it differently. Every single lockpick is capable of giving you exactly 25 unlucky points, no more, no less (assuming you keep using it until you can't use it anymore).  It is a simple fact that, with one single lockpick, which costs 1500 g, you can only get exactly 25 points, completely independent of the chances of the lockpick's retention.  Anyone who is capable of reading and understanding the probabilities of the complicated version should also be capable of grasping the above simple fact intuitively. On the other hand, with the number of lucky points, how many you get per lockpick depends on your retention rate.  Thus the math to calculate lucky points get complicated by probabilities.  The cost of unlucky point doesn't need to be complicated.  There's no need to cancel out something (the 1-R term)that we didn't have to introduce initially in the first place. -User:PanSola (talk to the [[Image:follower of Lyssa.png]]) 09:51, 29 April 2007 (CDT)
 * The average price per chest opening is 1500*(1-R) (unless discounted), and the average Unlucky points per chest are 25*(1-R), and the average price per unlucky point is Average price per chest divided by Averge Un/lucky points per chest. Do you want to dispute these quite simple facts? Introducing the (1-R) term is absolutely necessary, because in the calculation of price per Lucky point, they don't cancel out. Then we use the term again for Unlucky points, and they cancel out. How is that over-complicating things? I'm reluctant to abridge a perfectly correct formula which terminates after the exact same number of steps either way for the sake of "intuition" and hand-waving, which, although they may also give the correct result, don't exactly go well together with Maths. --[[Image:Roland_icon.png]]Roland of Gilead (talk) 10:54, 29 April 2007 (CDT)
 * I think that the point is that the (1-R) in both parts of the fraction includes the same R, meaning that it is not a "coincidence" that they cancel each other, but in fact, each time this R will cause the expense of those 1500g, it will also cause the gain of those 25 points, and vise versa. in the same way, I could add a variable for the probability that I'll play gw that day. Foo 11:32, 29 April 2007 (CDT)


 * The fact that you are even bothering with the price per chest in the first place is overcomplicating things for the Unlucky title track. It is necessary for the Lucky points, but is an unnecessary overcomplication for Unlucky points.  Just because my flight plans calling for a stopover at LA when I'm flying from Chicago to Hawii, doesn't mean I need to stop over at LA when flying to London.  I am opposed to making the extra stopover just so the number of transfers (and the location of the transfer) will work out the same whether I fly to Hawaii or London.  -User:PanSola (talk to the [[Image:follower of Lyssa.png]]) 01:47, 30 April 2007 (CDT)
 * If there's a fire extinguisher hanging on the wall and a fire breaks out, the normal guy and the mathematician will take it and extinguish the fire. The next time the extinguisher stands on the ground and another fire breaks out. The normal guy takes it and uses it, while the mathematician takes the extinguisher, hangs it on the wall, and then takes it off the wall to use it. That's the way of the mathematician, because it guarantees that the solution to the problem is correct. And as I wrote in reply to Foo above, I think breaking the symmetry between the formulas is more confusing than using the correct terms although not strictly necessary; and I think with the tex screenshot, things should have become easier to read and understand. Don't you agree? --[[Image:Roland_icon.png]]Roland of Gilead (talk) 02:21, 30 April 2007 (CDT)
 * I disagree. Keeping the symmetry just to guarantee the solution to the problem is correct is silly. I prefer wiki users to be treated as normal people, not non-flexible mathematicians. You are proposing to give a non-smart mathematical treatment, by insisting on always transferring to LA regardless of where you are going, and on always hanging the fire extinguisher on the wall first no matter if it's on the ground, or handed to you by someone else with the safty pulled and ready to use already.  The tex screen shot doesn't answer "why the heck does price per chest matter for unlucky points?", just like drawing the flightpath from Chicago to LA then to London doesn't answer why stopping we needed to stop at LA at first.  It causes unnecessary confusion by introducing the question "why did we need to do that?", when the answer is "we don't need to do that at all, but that's what we did when going to Hawii, so might as well do it again for the symmetry".  There's a difference between "not strictly necessary", and "strictly unnecessary".  -User:PanSola (talk to the [[Image:follower of Lyssa.png]]) 03:14, 30 April 2007 (CDT)
 * I question the assumption that every single reader will "intuitively", without any explanation, know that all terms involving probability will cancel out in the Unlucky case. My solution already involves the explanation, by saying "price per chest/UL points per chest = ...", while your single number didn't involve an explanation, not even an assessable statement. Can't you accept that this is an article heavy on Maths? You seem to disregard the prose explanation that follows the calculations as well, nobody is required to see and understand the Maths if it confuses him/her. --[[Image:Roland_icon.png]]Roland of Gilead (talk) 04:20, 30 April 2007 (CDT)
 * "every single user" does not need to intuitively know that the terms involving probability will cancel out, but just that each time he breaks a 1500g pick, he gets 25 unlucky points. it's as simple as that. the probability is forced here, and symmetry is not a good enough reason to include it. Foo 06:37, 30 April 2007 (CDT)
 * I love repeating myself, so I ask: is the prose explanation that follows the not sufficient to explain the supposedly confusing calculation? And if I follow the calculation, then including the (1-R) is absolutely necessary, because both sides of the fraction (price per chest/UL points per chest, which is the correct formula) contain it at first. I can't believe this is meeting such adamant resistance especially when your and PanSola's main (and pretty much only) argument consists of mere assertions of user confusion and their willingness to look at clean math in a math-based article. --[[Image:Roland_icon.png]]Roland of Gilead (talk) 06:55, 30 April 2007 (CDT)


 * The answer: it is insufficient to the people who can't intuitively figure out Anet has essentially told us: "Each lock pick gives you 25 unlucky points", and completely unnecessary to the people who can. The only reason including 1-R is necessary in your calculation, is because your calculation is doing something completely unnecessary -- trying to figure out the per-Chest cost of unlucky points.  It is as unnecessary as figuring out the per-character cost of lockpicks or the per-character chance of finding chest in the explorable area, for this Lyssa-loving per-account title track; and I don't see you writing prose explaining why those are unnecessary.  Those who can actually follow the explanation on probability should also be educated enough to understand that per-Chest computation is completely unnecessary for the computation of Unlucky points, whereas those who are not educated/intelligent enough will be just as confused whichever version we present them.  I can't believe this is meeting such adamant resistance especially when your main (and pretty much only) argument is for some arbitrary "symmetry", and "oh they will cancel out anyways".  You, sir, are the one messing with the cleanness of the math in a math-based article.
 * The answer: it is insufficient to the people who can't intuitively figure out Anet has essentially told us: "Each lock pick gives you 25 unlucky points", and completely unnecessary to the people who can Could we please in the foreseeable future stop arguing about what's intuitive? Everyone has different opinions of what is intuitively right. What about the humorless, basement-dwelling low-lifes who give a flying dirt about intuition and want mathematical proof, even if it's as basic as the one we're discussing? Just look at Talk:Nine Rings and read some of the demonstrated misconceptions people have about statistics. Watch people in the actual game who will for the most part stand in the center ring, because intuitively, it's obvious that you win more tickets in center than in corner, right? What I want to say with that: intuition can be and should be considered a bad advisor about mathematics, and since this a wiki, there's no guarantee that a quasi-mathematical argument in the article has not been mutilated by intuition. That's what traceable calculations are for, to make sure that intuition has not fooled the author. And if someone wants just the numbers, fine, just skip the maths (which are, conveniently, in a visually discernable area of the article) and look at the results, no harm done. I cannot fathom how anyone capable of browsing the web without getting his brain burned out could be overburdened with the task of skipping five lines of maths if they are considered too confusing.
 * because your calculation is doing something completely unnecessary -- trying to figure out the per-Chest cost of unlucky points. Just to make sure we are talking about the same thing, I was calculating the "price per Un/Lucky point" not "per-chest cost of unlucky points". I'm sure you realize at least that there's a difference between these two. I don't suppose you find it unecessary to calculate this value at all, so if you were as kind as to clarify your objection?
 * It is as unnecessary as figuring out the per-character cost of lockpicks or the per-character chance of finding chest in the explorable area and I don't see you writing prose explaining why those are unnecessary. I've never suggested doing so and implying I did is an insencere way of arguing, a straw man. But you're in company with Foo there, who said something about including the probability that he'd play GW that day. Same nonsense, same non-argument.
 * especially when your main (and pretty much only) argument is for some arbitrary 'symmetry' and 'oh they will cancel out anyways' My arguments are:
 * -Believe it or not, there is something called "mathematical beauty" or "elegance". Using one approach to solve two problems is mathematically more elgant than abridging that approach once but not the other time, especially when the 2 problems are basically the same. Saying that the "price per Un/Lucky point" equals "chest price per Un/Lucky point" means using the same approach each time; leaving out R once but not twice without explanation is a non-obvious variation of that approach. With some additional effort you can explain the variation, sure, but then it's additional effort for an approach that gives the same result. You have won nothing, except that you now have one calculation which seemingly does not depend on R to begin with while all others do. Yes, in the end it doesn't depend on R, but that's exactly what I want to show and which you want to avoid to show like the devil avoids holy water. Call this symmetry arbitrary if you will, but it is better than no symmetry with all other things being equal.
 * -It is better to separate the Maths and their explanations. Math is supposed to be exact and yes, sometimes it is painfully obvious. If that hurts anyone's reading flow or feelings or whatever, that one may use as much prose as is to his liking to alleviate the pain. If you want some formal terms: The maths are the result, the following prose is their discussion, what they mean, etc.. It's a basic principle to not let the discussion bleed back into the results, and by taking out the (1-R) when calculating the results, you are doing exactly that. You're using knowledge achieved in the discussion to simplify calculating the results which are the basis of the discussion, and that's what I object to.
 * -It is better, as I have written above, to choose the correct formula over an not necessarily wrong, but decidedly abridged version whose sole, dubious merit is that it's more "intuitive". You do concede that there are people who don't intuitively grasp the price per UL point by themselves; then why, of all things, do you prefer an "intuitive", abridged version of a formula instead of a step-by-step formula to explain it to those un-intuitive people? That just doesn't make sense. Intuitively speaking.
 * So now that I've again offered my arguments, if I asked you to stop misrepresenting them, would you tone down on the straw men? --[[Image:Roland_icon.png]]Roland of Gilead (talk) 10:10, 30 April 2007 (CDT)


 * Here's the other misrepresentation issue: Just because something is shorter, doesn't mean it is abridged. It was never necessary to stop over at Los Angeles if you are flying to London from Chicago.  My proposed flight plan takes me directly from Chicago to London, and it is a representation to call it an abridged flight plan just because you can come up with one that makes a transfer stop at Los Angeles first.  Lucky and Unlucky points are NOT essentially the same problem.  One is a cosair ship and the other is a fly.  Your claim is that there is elegance in using a single cannon to sink the ship AND kill the fly.  I heavily disagree.  The claimed "elegance" of using one approach to solve two different problems is marred by inelegance of bringing out a cannon to kill the fly.
 * My formula is "Price per lockpick divided by Unlucky points per lockpick", and that is as correct as the formula "Price per chest divided by Unlucky points per chest".
 * Problem A:
 * Given (by Anet): Price per lockpick (1500)
 * Given (by Anet): Unlucky points per lockpick (25)
 * To find: Price per Unlucky point (P_unlucky)
 * Problem B:
 * Given (by Anet): Price per lockpick (1500)
 * Given (by Anet): Lucky points per retention (125)
 * Given (by Anet): Probability of retention
 * To find: Price per lucky point (P_lucky)
 * Problem A and B are completely different in nature. Problem A is a simple exercise that any kid with decent background in math but who never took probability can figure out, come up with a formula, and prove its correctness.  "Price per lockpick divided by Unlucky points per lockpick" is by no means an "abridged" formula.  It is the orthodox approach for problem A that any math, physics, or chemistry teacher would have told you to use.  Just like anyone with any sense of geography would prefer a direct flight from Chicago to London, as opposed to making a random transfer at LA just because his travel agent also handled someone else's plans for Hawaii which called for a stop over at LA. -User:PanSola (talk to the [[Image:follower of Lyssa.png]]) 10:52, 30 April 2007 (CDT)

Even though none of my Maths teachers have taught me so, I really know what you're getting at. If I bought a lockpick and used it until it breaks (no matter how often that is), then I would have paid 1500g for 25 Unlucky points and therefore 60g per UL point. That's a perfectly legitimate argument, I never contested that.

However, simply writing "Price per UL point = 1500g/25" is hardly self-explanatory without prior knowledge (i.e. a prose explanation why no dependance on R) or actually thinking about why, especially if the only workable argument for Lucky points required R, in the line above. That's what I mean with abridged.

Now if I approach the problem from the chest's perspective rather than the lockpick's perspective (so to speak), then the solution becomes self-explanatory, because it becomes evident from the formula that and why UL point price does not depend on R. Additionally, the chest-based approach is mandatory for lucky points.

Or do you suggest that the values necessary for the formula are useless? I would strongly object to this notion:

For example, if someone knew "I want to farm 100 chests today and I have a retention rate of R", then he can also calculate about how much money he must expect to spend on keys. Or he wants to know how many chests does he probably have to open until he reaches the next tier in the Un/Lucky rank, then he can use the "Un/Lucky points per chest" value.

And if we have already provided these 3 values because they are useful in themselves, it's not exactly far-fetched to just continue using them for the price calculation. It may be like shooting with cannons at flies, but if you have sunk the ship with the first cannon and you already have the second cannon loaded and ready to fire, why go out and make the extra effort to roll up a newspaper and hunt down the fly yourself? Granted, it's a small effort, but still larger than the zero effort of igniting the second fuse. --Roland of Gilead (talk) 11:54, 30 April 2007 (CDT)
 * Ugh. If I'd known I'd start such an argument, I'd have left it alone. Especially since now there's a picture with formulas inside rather than something readable :( Would you mind returning to a text version? Obviously I prefer the simpler-but-still-correct way to deduce the cost per unlucky point, and I think taking shortcuts where allowed IS mathematically elegant, but I don't care much here, it was just an edit. I DO feel strongly about information supposed to be read being text rather than a picture, however. Thanks. 134.130.4.46 00:38, 2 May 2007 (CDT)
 * Um, I believe I started this particular argument, not you (whoever you might be). As for the tex picture, I think that's a totally different and at most weakly-related subject. -User:PanSola (talk to the [[Image:follower of Lyssa.png]]) 03:51, 2 May 2007 (CDT)
 * Do you find the picture more readable than text? I do, because the fractions are just ugly and require extra parentheses in plain ASCII text. I sure wish we had Wikipedia-grade tex support here. Or any at all. --[[Image:Roland_icon.png]]Roland of Gilead (talk) 07:35, 2 May 2007 (CDT)
 * Actually, I personally found the text perfectly readable. Readability was never part of my concern.  My issue has always been the killing the fly with a cannon. -User:PanSola (talk to the [[Image:follower of Lyssa.png]]) 17:39, 2 May 2007 (CDT)

Roland, I would prefer that instead of talking of text vs picture, you could accept the fact that after this long discussion you are still disputed by five other users. please don't force your opinion on the majority. Foo 16:30, 2 May 2007 (CDT)
 * Now come on. How and when was I forcing my opinion on anyone? That would mean I had reverted edits in the knowledge the majority was against it, which did not happen. The only edit I made during the discussion was replacing text with the picture, which was part of my argument. You won't hear me objecting to the consensus now that the discussion is over. --[[Image:Roland_icon.png]]Roland of Gilead (talk) 03:39, 3 May 2007 (CDT)

Mathematical incorrectness here

 * At Treasure hunter rank 7 and only picking Ascalonian chests on Normal mode, on average 10,600 chests have to be opened and 887 have to be spent.
 * At Treasure hunter rank 0 and only picking Hard mode chests, on average 26,300 chests have to be opened and 24,450 have to be spent.

But here's the catch, since Ascalonian chest have an inherent +55% success bonus, you might be better of hunting ascalonian chests if you have no ranks on lucky and treasure hunter titles. --Dark Paladin X 10:41, 29 April 2007 (CDT)
 * How is it incorrect? With "extreme" I meant both ends of the spectrum, i.e., starting at zero lucky points, what is the expected minimum and maximum of gold spent/chests opened? For the minimum, the conditions are Treasure hunter r7 and hunting only Ascalonian chests, the maximum is treasure hunter r0 and hunting only har mode locked chests. Change any of the conditions and you get something between these extrema. --[[Image:Roland_icon.png]]Roland of Gilead (talk) 10:54, 29 April 2007 (CDT)


 * However, there's a flaw here, still. The rank7/pickingAscalonian is not achievable no matter how you cut it.  It should include the progression, as in: it takes so much money to go from 0 to 1 picking Ascalonian chests, then so much money to go from 1 to 2.... and so on.  Then sum all that up and that will give the accurate "cheapest" price, which will be well over 887plat.  The 887 number is not at all achievable in the game, since you can't start at rank 7.
 * Now, same goes for the rank0/picking hard mode. You rank is guaranteed to increase, so you cannot end up picking 26,300 chests and still be at rank 0.  It whould include the same progression as I explained above.
 * THEREFORE: I propose to remove the two completely unachievable and meaningless numbers, and actually do the same math it took to get those numbers, and get some more useful figures. Either the straight up "takes this much money to go from 0 to 7 picking Ascalonian/hardmode chests," or do a table with columns being Ranks (0-1)(1-2)...(6-7)(Total 0-7) and rows being (Ascalonian chests) (Hard Mode chests).
 * I know you're trying to say that "Your price will lie somewhere between 887plat and 24,450plat," but that's like saying that 1,500 lies somewhere between zero and 1,000,000. Useless.
 * And, if the more useful figures cannot be calculated by some dedicated person (I'm not gonna be the one doing the necessary math, as I am not willing to spend the time learning all the formulas, I'm just a reader in this case), this paragraph should be removed from the article entirely. The information offered, while abstractly correct, provides nothing of value to the reader.  The previous analyses show more than enough figures to give a much better guess at what it'll cost.  Correct me if I'm wrong? RoseOfKali 16:03, 17 May 2007 (CDT)
 * Correct me if I'm wrong?
 * No sweat.
 * I know you're trying to say that "Your price will lie somewhere between 887plat and 24,450plat,"
 * I was actually saying that your expected price for this method will lie between these figures. Nitpick if you will, but we are talking about statistics here.
 * but that's like saying that 1,500 lies somewhere between zero and 1,000,000.
 * I'm not exactly sure what you were trying to say with that. It's not useless to calculate the lower and upper bound and therefore be certain that the expectable actual value will lie between these two. Unlike your numbers seem to be, mine are not arbitrary but correct, to the best of my knowledge.
 * The rank7/pickingAscalonian is not achievable no matter how you cut it.
 * Why should it be impossible to have r7 treasure hunter and zero lucky points? You can get TH without lockpicks. I conciously decided to not take into account the amount of gold and chests opened necessary to get to r7 TH. It's just the premise I chose and I was calculating from there. If you insist we include the minimum price to get to r7 TH without lockpicks, fine, we can talk about that.
 * Now, same goes for the rank0/picking hard mode.
 * That's clearly not true. There is nothing speaking against having zero TH / lucky points in hard mode, and AFAIK I didn't forget any hidden costs. For this premise, the number is definitely correct; as I have written in the article, both numbers do pay attention to the "break points" when the lucky or treasure hunter title (and therefore the retention rate) increases. Certainly you haven't missed that statement?
 * And, if the more useful figures cannot be calculated by some dedicated person (I'm not gonna be the one doing the necessary math, as I am not willing to spend the time learning all the formulas, I'm just a reader in this case), this paragraph should be removed from the article entirely.
 * I indeed am this dedicated person and I'm quite certain that my numbers are correct, thank you very much. Since you a) seem to request an approach I already provided and b) are reluctant to calculate your own numbers (which, by the way, doesn't require any formulas at all), I would suggest we let these numbers stand until someone comes up with better numbers and an explanation how to reach them or a reason why this piece of information should be considered irrelevant in this article. --[[Image:Roland_icon.png]]Roland of Gilead (talk) 18:35, 17 May 2007 (CDT)
 * Roland, you seem to have missed my paragraph about the table that I proposed, you didn't comment on that at all. Statistics are statistics.  The chance of anyone dying someday is 1, the chance of someone dying today is... who knows?  I'm just saying that these brackets are too far out from where the actual numbers will lie, I didn't say they were wrong.  All I was saying is that you cannot start with maxed out Lucky on one end, and cannot stay at 0 forever on the other end.  Very few people have maxed Lucky at the moment.  The explanation that you provided assumes just that (at least it sounds like it does).  Include a progression table, and the numbers will become much more useful for everyone else, especially for people who are somewhere inbetween, and want to know how much more they have to go to reach a certain rank, or max the title.  I guess I was too wordy in trying to say this one thing.  If I still sound confusing, I guess I'll spend some time some day and actually compile the said table to show exactly what I'm trying to say.  I just didn't want to deal with the math if someone else has done it already.  (Yes, that does mean I'm lazy sometimes.)
 * And I may have not been clear on exactly what the numbers mean, but that should say something as well. I don't consider myself a genius, and I don't consider myself an idiot, either.  I just think that the table I proposed would be useful for someone who has partial or no progression in these titles, which would be true for probably a lot of players. RoseOfKali 15:05, 29 May 2007 (CDT)

The Math for the 9 Ring and 12 Rings are correct assuming every play each circle has an equal chance of winning. Basic Stats class will give any doubters the ability to do your own math with the tools given through the gift of education. Everyone knows that in computer programming there is no wuch thing as an equal chance. As if not A then B, and if not B then C.... etc etc. The programming is much better and more complicated. But I refuse to believe that every square over time will yield the same as all else at one point. But the point is not they shall all yield the same. It is the fact that the statistics are correct, and gamble for your title with the knowledge presented. Thanks AbsoluteEminence

Is this close aproximit like if your lucky it could be way shorter
i'm just wondering, because i'm waiting for the next event(saving up plats from my chest runs.
 * Assuming this question is "Are these numbers approximate? Could the times be shorter?" and relating to 9 Rings/Rings of Fortune... These numbers are the statistical averages.  While it is possible that you could get the Lucky Title and only spend 1 plat, it's extremely unlikely.  It's like how you can flip a coin 100 times and have it land Heads every time, but it's just really unlikely.  You may have to spend more, you may have to spend less, but this guide gives you a really good idea of how much you'll have to invest if you want a certain title.  Perrsun 20:44, 3 July 2007 (CDT)

Even second graders are taught to always reduce to simplest form. I'd hope a self-proclaimed mathematician could match their wits.

Wrong value of Tickets per hour losses in the table
That is based on the festival 6-8 Jul 2007

I've standed in the center of the Nine Rings for 7 hours. I've started with 10k (666 festival tickets - each costs 15 gold). I've gathered about 24k lucky points, and i've lost about 400 tickets.

So simple calculation shows that:

Wins per hour: 24000/7=3429 which match the theory.

Ticket losses per hour: 400/7=57. According the theory is must be 203!

So, if thetheory is ok, I don't agree that circles are equal. There were many calculations in the discussion, showing that rings are equal, but they aren't. And most profitable is to stay in center (i didn't make so complex calculations, i've just used calculator). Standing in the corner will kick you out of the game with speed of 203 ticket per hour, standing in center will keep you 4 time longer.

Aerial Starlight 21:26, 6 July 2007 (CDT)
 * I do agree that the math in the table is off (by a significant amount), but your own interpretation is also flawed. Based on what you've written above, you won close to 24000 tickets and ended the day with 266 of 666 tickets left.  If this is true, then the part you missed was that the total amount of tickets you lost was the net loss minus your total winnings in the period.  In other words, you did win 24000 extra tickets, but you don't have them anymore.  Your total loss is therefore on the order of 24400 tickets, not 400, like you thought.  Based on my own preliminary data (I'll be doing this all weekend), the ration of tickets won to tickets lost is right around 98%.  This fits perfectly with your data and is close to what I've got now.  I'm sure this table will be updated on Monday with new results.
 * Wanna help out? Record the exact times (as close as you can get), exact number of Lucky/Unlucky points and exact number of tickets with which you started and ended, in the handy table below.
 * &mdash;DaveK 07:13, 7 July 2007 (GMT)
 * {| border="1"

!style="background:#EFE"| User !! Start Time !!style="background:#E8E"| End Time !! Start Tickets !!style="background:#2AA"| End Tickets !! Starting Lucky !!style="background:#99A"| Ending Lucky !! Starting Unlucky !!style="background:#A9E"| Ending Unlucky !! Nine Rings? !!style="background:#FAA"| Position!!
 * - align="center"
 * style="background:#EFE"|Bustas|| 00:00? ||style="background:#E8E"| 07:00|| 666 ||style="background:#2AA"| 266? || 0? ||style="background:#99A"| 24,000? || ? || style="background:#A9E"|? || Yes || style="background:#FAA"|Center
 * - align="center"
 * style="background:#EFE"|DaveK|| 00:00 ||style="background:#E8E"| 02:45|| 500 ||style="background:#2AA"| 405 || 51,191 ||style="background:#99A"| 61186 || 2,771 ||style="background:#A9E"| 3,447 || Yes || style="background:#FAA"|Corner
 * - align="center"
 * style="background:#EFE"|Avisotin|| 00:00 ||style="background:#E8E"| 08:30|| 3000 ||style="background:#2AA"| 300 || 0 ||style="background:#99A"| 29,167 || 0 ||style="background:#A9E"| 2,131 || Yes || style="background:#FAA"|Corner
 * - align="center"
 * style="background:#EFE"|Rcollins779x|| 00:00 ||style="background:#E8E"| 12:00|| 6666 ||style="background:#2AA"| 4085 || 528,750||style="background:#99A"| 598,750|| 31,750 ||style="background:#A9E"| 35,750|| Yes || style="background:#FAA"|Side
 * - align="center"
 * - align="center"
 * style="background:#EFE"|Kain|| 17:00 ||style="background:#E8E"| 16:00 || 6000 ||style="background:#2AA"| 460 || 261.168 ||style="background:#99A"| 359.198 || 18.013 ||style="background:#A9E"| 24.766 || Yes || style="background:#FAA"|Corner
 * - align="center"


 * }


 * I'd just like to say that to anyone who wants to get either the lucky, unlucky, or even both titles in this Dragon Festival - it's a lot more fun to just stand in a circle and see what happens. Remember that all of this mathematical prediction can mean nothing... depending on how lucky, or unlucky... or both, you are. :) Avisotin 03:09, 7 July 2007 (CDT)


 * Assuming all circles have an equal chance of being selected, all circles should, on average have the same rates of ticket consumption and the same number of tickets won over a given period of time lost ratio and the same rate of ticket consumption, however the middle circle should have the most stable results, while the corner circles should have the most erratic results. -- Gordon Ecker 03:12, 7 July 2007 (CDT)


 * Unlucky is based on number of games lost not tickets. Calculations in the guide are based on standing on a corner circle as you will lose 2/3 of games played.  (As opposed to the centre where you only lose 4/9).--JP 03:20, 7 July 2007 (CDT)


 * That may not be true. If Unlucky points are awarded for any loss (including losses where you're a runner-up and get the consolation prize), then the odds are always 1/9 in Nine Rings.  If you standing in a ring that gets awarded tickets never counts as a 'lost game,' then you're right.  And I know quite well what the odds should be, but I'm much more interested in what the data shows to be true. :)  &mdash;DaveK  08:27, 7 July 2007 (GMT)
 * Watch your title bar when you are a runner up, it doesn't change for unlucky. However the page does assume that each circle has an equal probability of winning.  Also depending on how the time was worked out there may be space to fine tune that.( ie record the number of games in a particular time span and average that out.--JP 03:43, 7 July 2007 (CDT)


 * Yes, but the mean net ticket losses per hour should be theoretically identical for the same game regardless of circle, and all circles in the same game should be equal for the lucky title, which seems to be the only one of the two titles that the original poster is interested in. -- Gordon Ecker 17:25, 7 July 2007 (CDT)


 * True, but that's the definition of a mean. If that's the way things really do work then the data will bear that out. As I'd mentioned, I'm less interested in how things should be and more interested in how they are. But someone went and killed my project. :<
 * But the original premise has been satisfied. Aerial Starlight was right, and the data in aggregate indicates that the odds for corner rings are consistent and as expected (for games and tickets won/lost). I'll do the center ring next to see if there are any surprises.
 * &mdash;DaveK 22:40, 7 July 2007 (GMT)


 * The math from the table is quite solid, look at the original guru post for it. I spotted nothing wrong with it.  However, there is a difference between statistics and probability.  The idea behind the posts is that everything is based "on average".  On average, you'd expect to show this many wins or losses over that period of time.  Every time you do this, you could expect to either go out early, or to stay longer and win even more.  I can cite my experiences over the weekend as well, each showing a different conclusion.  The first night, I stood around for 2 hours playing 9 rings, standing on a corner.  Over the two hour period, I won and lost.  But at the time I stopped, I was exactly up 5 tickets.  So, I could say the data's wrong +5 tickets/ 2 hrs = I should be up 2.5 tickets an hour.  If you claim that maybe I wasn't on long enough, I did the same thing today, except I was playing for 6 hours.  And just as it seems, I was up 5 tickets again (corner still).  But maybe I was just lucky.  However, I failed to also mention, that after sleeping for 6 hours yesterday, I was down about 3000 tickets.  There's a lot of variation to consider, because it is a game of chance.  Unless someone records a lot of the wins to show that there's a statistically significant difference in the chance of winning in each of the 9 rings, I don't see a problem to the theory.  Because it is a theory, and it is probability, everybody's results can be different.  In other words, Aerial Starlight, I'm basically saying, you were just lucky.  I probably could do more math, and try to show you what your the chances of not losing so much over time is, but it's quite a bit more complicated than what has been down to show the averages.  164.67.234.129 06:15, 7 July 2007 (CDT)


 * You said it better than I could. Removing the note from the Nine Rings page because the averages are right, and can easily be shown so. --Buzzer 11:56, 7 July 2007 (CDT)


 * Except for the fact that you're wrong. I see where the original poster got the numbers used ([5/9 tickets lost per game] x [3600 seconds in an hour/(187 seconds/19 games)]) but the number 5/9 is incorrect.  The amount of time for a game of Nine Rings is approximately (187 sec/19 games) 9.84 sec.  This yields about 365 games per hour.  These numbers are static and unchanging.  No matter where you stand, there're always 365 games an hour.  Standing in a ring with tickets will cause you to lose 10 tickets for every game.  This also is not dependant on where you're standing or whether you win.  You always lose 10 tickets a game.  Thus, every hour, you lose 3650 tickets - no matter where you stand.  These losses are offset by the winnings, which do average out to 3454 tickets an hour.  This is shown by the data and internally consistent unlike the original numbers.  Seriously, as written, an average loss of 200 tickets an hour compared to an average gain of 3400?  No one would ever run out of tickets, and we know that's not true.
 * This is what Aerial Starlight was trying to point out, above. I'll give the guy credit, because he saw something that defied common sense and ran with it.  And he was right.
 * &mdash;DaveK 22:15, 7 July 2007 (GMT)

By the way... something that could slightly skew the odds is the issue of inventory slots available. While I was watching, someone in a corner ring had two groups of fifty-five tickets drop on the ground. They were reserved for the player, but since they were AFK the reservation expired and they got snagged. The lesson here is that in order to be a big winner you must have enough space to hold your winnings!