primadog

my POH dungeon for the upcoming player-made quest event, The Spoils of War. (map's custom made by slayer_shunt)

Let's see if we can come to an agreement, silver. Even though the drop rate is random, there's still certain things you can say about the amount of drops you get. For example, if you're making a green dragon run, you'll probably get more gp drops than clue drops. Now let's say green dragons drop gp half the time ( 50%) and clues drops 1%. if you killed 100 dragons, you'll probably expect more gp drops than clue drops. Furthermore, you'll probably more likely to to see between 40 to 60 grops of gp than 0 to 10 drops of drop, or 1 clue drop than 100 clue drops (assuming you done each clue). Now there's a way in statistics to measure just 'how likely' 40 to 60 drops of gp is going to be, compare to 0 to 10 drops. I won't go into the details, you can find it here. What I meant from that part you quoted is if you done 300 drops, the average drop should be 30 items (this is the average value, aka expected value or mean), with a standard deviation of 5.5. This means there's a 95% chance ( 2 standard deviation from mean) to have drops between 19 to 41 items. Since there's only 5% chance to be beyond that range, below 19 drops or above 41 drops is good reason to say that my hypothesis of the drop rate to be 10% to be wrong. What you said regarding the second wheel is completely true. The second wheel portion is still speculation, mainly because I can't find a good formula that I am satisfied with. Of course, having more runs would be give us more accurate data, thus better estimate of the drop rate, but looking at the # of previous experiments that stopped midway , I can only take what i can find and what people's willing to spare (thanks mattyb3!).

converted the three main data files i used into spreadsheet: http://www.nd.edu/~xhong/runescape/beechworth.xls http://www.nd.edu/~xhong/runescape/mattyb3.xls http://www.nd.edu/~xhong/runescape/newbiei.xls hope you find it useful! PS Mattyb3, your data was very important, ty! Let me know if you have more for me.

updated to reflect some problems brought up by earlier posts. Thanks for the help everyone.

mattyb3, I'll really appreciate it if you can provide me with the total rune/gp/bolt drop you receive. That's the main item I need as for now. (if possible, i'll like to receive a copy of your excel file) Mc90123, to make it easier for me to interprete your data, I have put your barrow log through my script. I hope you don't mind. arkwelded, what you mention is a possibility, but it's not something that we can test in anyway, so I will have to assume that drop rate is not player-dependent.

http://www.stat.yale.edu/Courses/1997-98/101/binom.htm I am currently in desperate need of more barrow logs to research though, if anyone can contribute links, it'll be greatly appreciated.

working on that right now. I'll report when i have a good size count. EDIT1: i been doing 6 kc for the past ten runs, it appears that indeed you don't get blood & bolts at that, i'll have to modify the hypothesis about this, but i'll need further data.

You're not guaranteed to get 30 item out of 300 chests, probability don't work that way. However, if you get less than 19 items or more than 41, it's evidence that my conclusion is wrong. If so, let me know. reason: In binomial distribution: average drop = number of runs * brothers killed * chance deviation = ( number of runs * brothers killed * chance * ( 1- chance ) ) ^ .5 There's a 5% chance that your drop count is larger than average drop +/- 2 * deviation Based on yours: average drop = 30 deviation = 5.4 there's a 95% chance that your drop count will be within 19 to 41 items.

converted the three main data files i used into spreadsheet in case someone else want to give the data a try: http://www.nd.edu/~xhong/runescape/beechworth.xls http://www.nd.edu/~xhong/runescape/mattyb3.xls http://www.nd.edu/~xhong/runescape/newbiei.xls LOG as evidence: others used for summary purposes: 19-20-574-40948124 4 out of 45 trips 19-20-272-39981415 13 out of 162 trips 19-20-685-40892430 8 out of 72 trips 19-20-605-34726195 35 out of 360 trips 19-20-626-25865127 14 out of 140 trips Newbiei Detail: http://www.nd.edu/~xhong/runescape/blog/newbiei.txt This is newbiei's barrow log, ouput on Thu Mar 29 04:38:33 2007 Total Trips: 112 Total Worth: 14135072 gp || Average per trip: 126206 gp || Average per hour: 575064 gp Total Time Spent: 24.58 hours || Average per trip: 13.17 min ~~~~ =============Summary============= 27851 Mind rune 13615 Chaos rune 6298 Death rune 2988 Blood rune 2779 Bolt racks 127756 Gp 2 Guthans helm 1 Dharoks platelegs 1 Torags helm 1 Torags hammers 1 Ahrims robetop 2 Ahrims staff 2 Karils crossbow 1 Karils leathertop 1 Karils coif 14135072 value 1 Half key 112 run ~~~~ beechworth's blog, origin: 19,20,516,35296675 Detail: http://www.nd.edu/~xhong/runescape/BLog/beechworth.txt This is beechworth's barrow log, ouput on Mon Mar 26 13:10:59 2007 Total Trips: 379 Total Worth: 46845999 gp || Average per trip: 123604 gp || Average per hour: 292787494 gp ~~~~ =============Summary============= 46845999 value 379 run 1 Guthans platebody 3 Veracs helm 2 Veracs brassard 1 Veracs flail 1 Veracs plateskirt 1 Dharoks greataxe 2 Dharoks platebody 1 Dharoks helm 1 Dharoks platelegs 3 Torags hammers 1 Ahrims robeskirt 1 Karils crossbow 1 Karils leatherskirt 3 Karils coif 2 Dragon med helm 15 Half key 94917 Mind rune 40933 Chaos rune 27811 Death rune 13645 Blood rune 12076 Bolt racks 356250 Gp ~~~~ mc90123's barrow log, based on his post on the second page Detail: http://www.nd.edu/~xhong/runescape/BLog/mc90123.txt This is mc90123's barrow log, ouput on Mon Mar 26 14:17:22 2007 Total Trips: 204 Total Worth: 29739797 gp || Average per trip: 145783 gp || Average per hour: 1486989850 gp ~~~~ =============Summary============= 29739797 value 204 run 1 Guthans warspear 2 Guthans helm 2 Veracs flail 1 Torags platebody 1 Torags helm 1 Ahrims robeskirt 1 Ahrims hood 4 Dragon med helm 11 Half key 51937 Mind rune 22406 Chaos rune 12574 Death rune 7192 Blood rune 6851 Bolt racks 202617 Gp ~~~~ mattyb3's blog from post http://www.nd.edu/~xhong/runescape/BLog/mattyb3.txt This is mattyb3's barrow log, ouput on Tue Mar 27 11:06:15 2007 Total Trips: 73 Total Worth: 6196807 gp || Average per trip: 84888 gp || Average per hour: 619680700 gp Total Time Spent: 0.01 hours || Average per trip: 0 min ~~~~ =============Summary============= 6196807 value 73 run 1 Ahrims robeskirt 2 Ahrims hood 18934 Mind rune 10302 Chaos rune 5772 Death rune 1473 Blood rune 966 Bolt racks 91827 Gp ~~~~

Requesting more barrow logs /w kc! Premise: Barrows drop is not rare, nor is it affected by kc, time spent, or style, with evidence. Summary 6 brothers = 6 drops, the chest holds all the drops. each drop has 1/60 chance to be barrows item. If not barrows/d med/half key, the drop can be the following: |~300 mind | ~150 chaos | ~75 death | ~50 blood | ~35 bolt | ~500 gp 6-7 KC | 1 | 1 | 1 | 0 | 0 | 4 8 KC | 1 | 1 | 1 | 1 | 0 | 3 9+ KC | 1 | 1 | 1 | 1 | 1 | 2 Myth: Barrow drops are rare. FALSE Evidence: the average rate for barrows items is 1 per 10 trips, or 1/60 for each barrows brother. This is assuming you kill all six brothers. ref: (only counting barrows drops as for now) my own data 12 out of 122 trips 9.7% 19-20-574-40948124 4 out of 45 trips 8.9% 19-20-272-39981415 13 out of 162 trips 8.0% 19-20-685-40892430 8 out of 72 trips 11.1% 19-20-605-34726195 35 out of 360 trips 9.7% 19-20-626-25865127 14 out of 140 trips 10% 19-20-516-35296675 25 out of 382 trips 6.5% mc90123's log 9 out of 204 trips 4.4% --------------------------------------------- total: 120 out of 1487 : 8.07% (requesting more barrow log counts) acceptable range: 120 +/- 24 (at 95%) expected at 1/60: 148.7 items Acceptable as for now? YES! but barely Myth: Drop rate depends on kc(kill count)/method/time taken. FALSE Evidence: look at the above examples, the drop rate is incredibily steady, regardless of personal style. Myth: You can't figure out the chances. FALSE Chances: Each brother you kill give you one drop (clones don't count). You get maximum of 6 separate drops from the six brothers you kill. If the drop isn't a barrows or half key or d med, then it is dropped by the following ratio: |~300 mind | ~150 chaos | ~75 death | ~50 blood | ~35 bolt | ~500 gp 6-7 KC | 1 | 1 | 1 | 0 | 0 | 4 8 KC | 1 | 1 | 1 | 1 | 0 | 3 9+ KC | 1 | 1 | 1 | 1 | 1 | 2 ie, if you have 8KC, you have 1/7 chance to get ~300 minds, ~150 chaos, ~75 deaths, ~50 bloods, and 3/7 chance to get ~500 gold for each brother you killed, assuming you didn't get a barrows drop. Evidence: 1) Look at my drop log, Looking at the default values and compare to each drop, you will notice that every single one of them can be counted as six different drops by the six different brothers. Example: Trip 57 Loot- 299 mind, 37 bolt, 1010 gp, 128 chaos, 70 death From this data and looking at the average drop table, we can conclude that the brothers had 1 mind rune drop, 1 bolt racks drop, 2 gp drop, 1 chaos rune drop, and 1 death rune drop ( 1+1+2+1+1 = 6 brothers! ). 2) Looking at the drop summary from my blog: From only 72 trips ( or 72 * 6 = 432 drops), look at how close that resembles the drop rates I posted above. 14007 Mind rune / ~300 = 46.69 / 432 drops = 10.8% 7629 Chaos rune / ~150 = 50.86 / 432 drops = 11.7% 4217 Death rune / ~75 = 56.23 / 432 drops = 13.0 % 2399 Blood rune / ~50 = 47.98 / 432 drops = 11.1 % 2672 Bolt racks / ~35 = 76.34 / 432 drops = 17.7 % 71256 Gp / ~500 = 142.51 / 432 drops = 33.0 % 7 barrows = 7 / 432 drops = 1.6% 1 half key = 1 3) Add up the drop counts from above: 46.69+50.86+56.23+47.98+76.34+142.51+7+1=428.61 ie 428 drops This is data from 72 trials, so I killed 72 * 6 = 432 drops 429 v. 432, pretty close for something random, isn't it? 4) Summary of data from other players, summarized by KC alone: The data is rather quite large and complicated. The basic idea is in random variables the average value approaches expected value, so if we have average a large number of chest drops based on KC, we can expect the average chest drop to be similar in ratio to that we expect (as above). The summary of chest drop by KC using beechworth, mattb3, and my data: http://www.nd.edu/~xhong/runescape/OutputSum.htm . The more runs we have at a particular kc, the closer is it to the ratio we expect. If we group the KC by 6-7, 8, or 9+, the relationship is ever clearer: The ratios are very clear now. BOTTOM LINE: WHAT THE CHEST HOLDS IS NOT AFFECTED BY KC**, TIME SPENT, METHOD! *In case you're wondering why the sum is not 100%, here's the details: There's a 1/10 chance that you can get a chance to get a rare drop. This is called a 'rare drop wheel' or a second 'wheel' (my current data is that detailed enough for me to decipher the drop rate of this second wheel. Thus, all of below are SPECULATION! If you can link me to more data, it will be very helpful) 1/10 rare drop wheel as follow: any of the 24 barrow pieces ~1/200? d med ~1/200? half key ~7/200? ------------------------- 1/200 * 24 + 1 / 200 + 7 / 200 = 33 / 200 = 1 / 6 Thus, a rare drop has a 1/6 * 1/ 10 = 1/ 60 chance The rest of the rare drop wheel is equivalent to the normal drop ratio: 300 MIND 1 : 150 CHAOS 1 : 75 DEATH 1 : 50 BLOOD 1 : 35 BOLTS 1 : 500 GP 1 **One interesting calculation we can do: we can find the average worth of a chest at 6-7 KC, 8, or 9+. mind runes - 20 gp chaos runes - 135 gp death runes - 300 gp blood runes - 550 gp bolt racks - 330 gp 6-7 KC: 1/7 * 300 * 20 + 1/7 * 150 * 135 + 1/7 * 75 * 300 + 4/7 * 500 8 KC: 1/7 * 300 * 20 + 1/7 * 150 * 135 + 1/7 * 75 * 300 + 1/7 * 50 * 550 + 3/7 * 500 9+ KC: 1/7 * 300 * 20 + 1/7 * 150 * 135 + 1/7 * 75 * 300 + 1/7 * 50 * 550 + 1/7 * 35 * 330 + 2/7 * 500 6-7 KC: 7250 * 6 = 43500 8 KC: 11107 * 6 = 66642 9+ KC: 12686 * 6 = 76116

sounds great! Hope you have the whole server to yourself, cause you'll need it with all the attendies!

jock, i'll most definitely update the post to reflect you and your team's contribution (right now our more immediate concern is correcting our errors). I agree with you that there was allot that went wrong in the trial, but that's why we done this trial -- to solve these kind of errors before we go forth with the event. Thanks for coming, we couldn't fix the quest for the better without you :).

fun? yes messy? Absolutely Are we going to do another? Hunt me down if we didn't!

can you feed them tomatoes or strawberries? If so maybe a basket of them would be a better choice?

Warriors should have names that instill fear on their enemies and awe on their friends. Ahrim the Unreachable Karil the Inescapable Torag the Invincible Guthan the Indestructable Verac the ... Dharok the ... (still working on the last two)

jamyj2 you are ridiculus!

Once you got to lvl 74, you can build oak doors in the dungeon, which uses 10 oak planks & are slightly faster. ranging pedestals (8 teak) are not good experience, unfortunately, because it have a very long loading screen.

finally a high-level skiller update and people still complains? God, if I'm in jagex i'd cry every night.

8/10 I did thought fight cave's hard

7/10 AOW material if +1 you explain more about how to endure the grinding +1 infiltrate OoC +1 and destroy it from within :thumbsup: wonderful guide otherwise

crafting, it's simply the most diversed skill. Just look the length of the Crafting Skill guide!

^the animation's awesome, gonna get a few asap! I can't seen to find that construction goblin, do anyone have a pointer for me?

Where is this, can you show it on a map for me? I'm quite interested.

Let's keep the conversation on topic, shall we? Look at the facts, strider showed a feature that the official client is missing, now we have such a feature - everyone wins. I for one disliked the idea of 3rd party clients, but this is a benefit to everyone, no one can deny that.

:( i screwed up the US time because of daylight saving. It's fixed now, if there's any future screw-up, please remember the time is 10PM GMT!