Research:Jewellery
From Discworld MUD Wiki
Durability
Observations (damage caused by JPCT casts):
item full_value value_drop times # platinum crescent moon ring (penultimate): 12400 wibi, 1154p, 3100 "plat. cres. moon (7)" 12400 0 404 "plat. cres. moon (7)" 12400 744 1 # 50p, 6/100 "plat. cres. moon (7)" 12400 868 1 # 59p, 7/100 "plat. cres. moon (7)" 12400 1488 1 # 100p, 12/100 "plat. cres. moon (7)" 12400 1612 1 # 109p, 13/100 "plat. cres. moon (7)" 12400 1736 1 # 117p, 14/100 "plat. cres. moon (7)" 12400 2480 1 # 167p, 20/100 "plat. cres. moon (7)" 12400 2976 1 # 201p, 24/100 "plat. cres. moon (7)" 12400 3100 1 # 209p, 25/100 # gold crescent moon ring (penultimate): 4700 wibi, 580p "gold cres. moon (7)" 4700 0 437 "gold cres. moon (7)" 4700 94 1 # 9p, 2/100 "gold cres. moon (7)" 4700 329 1 # 30p, 7/100 "gold cres. moon (7)" 4700 470 2 # 42p, 10/100 "gold cres. moon (7)" 4700 564 1 # 51p, 12/100 "gold cres. moon (7)" 4700 611 3 # 55p, 13/100 "gold cres. moon (7)" 4700 658 1 # 59p, 14/100 "gold cres. moon (7)" 4700 705 1 # 64p, 15/100 "gold cres. moon (7)" 4700 940 1 # 85p, 20/100 "gold cres. moon (7)" 4700 987 1 # 89p, 21/100 "gold cres. moon (7)" 4700 1081 1 # 98p, 23/100 "gold cres. moon (7)" 4700 1222 1 # 112p, 26/100 "gold cres. moon (7)" 4700 1316 1 # 122p, 28/100 "gold cres. moon (7)" 4700 1410 1 # 133p, 30/100 "gold cres. moon (7)" 4700 1551 2 # 149p, 33/100 # gold crescent moon ring (below penultimate): 4700 wibi, 580p "gold cres. moon (6)" 4700 0 406 "gold cres. moon (6)" 4700 141 1 # 13p, 3/100 "gold cres. moon (6)" 4700 329 1 # 30p, 7/100 "gold cres. moon (6)" 4700 470 2 # 42p, 10/100 "gold cres. moon (6)" 4700 799 1 # 72p, 17/100 "gold cres. moon (6)" 4700 846 2 # 76p, 18/100 "gold cres. moon (6)" 4700 940 2 # 85p, 20/100 "gold cres. moon (6)" 4700 987 1 # 89p, 21/100 "gold cres. moon (6)" 4700 1034 1 # 93p, 22/100 "gold cres. moon (6)" 4700 1269 1 # 117p, 27/100 "gold cres. moon (6)" 4700 1410 1 # 133p, 30/100 "gold cres. moon (6)" 4700 1457 1 # 138p, 31/100 # rose gold crescent moon ring (penultimate): 5800 wibi, 679p "rose gold c. moon (7)" 5800 0 609 "rose gold c. moon (7)" 5800 58 1 # 5p, 1/100 "rose gold c. moon (7)" 5800 174 2 # 16p, 3/100 "rose gold c. moon (7)" 5800 348 1 # 31p, 6/100 "rose gold c. moon (7)" 5800 406 1 # 37p, 7/100 "rose gold c. moon (7)" 5800 464 1 # 42p, 8/100 "rose gold c. moon (7)" 5800 754 1 # 68p, 13/100 "rose gold c. moon (7)" 5800 870 1 # 79p (674 to 595), 15/100 "rose gold c. moon (7)" 5800 1276 1 # 115p, 22/100 "rose gold c. moon (7)" 5800 1334 1 # 120p, 23/100 "rose gold c. moon (7)" 5800 1566 1 # 141p, 27/100 "rose gold c. moon (7)" 5800 1624 1 # 146p, 28/100 "rose gold c. moon (7)" 5800 1682 2 # 152p, 29/100 "rose gold c. moon (7)" 5800 1914 1 # 172p, 33/100 # rose gold crescent moon ring (below penultimate): 5800 wibi, 679p "rose gold c. moon (6)" 5800 0 604 "rose gold c. moon (6)" 5800 290 1 # 26p, 5/100 "rose gold c. moon (6)" 5800 406 1 # 37p, 7/100 "rose gold c. moon (6)" 5800 464 1 # 42p, 8/100 "rose gold c. moon (6)" 5800 580 1 # 52p, 10/100 "rose gold c. moon (6)" 5800 638 1 # 58p, 11/100 "rose gold c. moon (6)" 5800 754 2 # 68p, 13/100 "rose gold c. moon (6)" 5800 1102 1 # 99p, 19/100 "rose gold c. moon (6)" 5800 1160 2 # 104p, 20/100 "rose gold c. moon (6)" 5800 1218 1 # 110p, 21/100 "rose gold c. moon (6)" 5800 1276 1 # 115p, 22/100 "rose gold c. moon (6)" 5800 1392 1 # 125p, 24/100 "rose gold c. moon (6)" 5800 1450 1 # 131p, 25/100 "rose gold c. moon (6)" 5800 1508 1 # 136p, 26/100 "rose gold c. moon (6)" 5800 1624 1 # 146p, 28/100 "rose gold c. moon (6)" 5800 1682 1 # 152p, 29/100 "rose gold c. moon (6)" 5800 1856 2 # 167p, 32/100 "rose gold c. moon (6)" 5800 1914 1 # 172p, 33/100 # black gold oak leaf ring (7), 8000 wibi, 857p "black gold oak (7)" 8000 0 402 "black gold oak (7)" 8000 480 1 # 32p, 6/100 "black gold oak (7)" 8000 560 1 # 38p, 7/100 "black gold oak (7)" 8000 720 1 # 49p, 9/100 "black gold oak (7)" 8000 1040 1 # 74p, 13/100 "black gold oak (7)" 8000 1360 2 # 102p, 17/100 "black gold oak (7)" 8000 2240 1 # 182p, 28/100 "black gold oak (7)" 8000 2400 2 # 196p, 30/100 "black gold oak (7)" 8000 2560 3 # 210p, 32/100 # gold: 2000 wibi, 302p, 500p "gold ring" 2000 0 673 "gold ring" 2000 20 3 # 2p, 1/100 "gold ring" 2000 60 2 # 7p, 3/100 "gold ring" 2000 80 2 # 9p, 4/100 "gold ring" 2000 100 1 # 11p, 5/100 "gold ring" 2000 120 1 # 13p, 6/100 "gold ring" 2000 140 1 # 16p, 7/100 "gold ring" 2000 160 2 # 18p, 40p, 8/100 "gold ring" 2000 180 1 # 20p, 9/100 "gold ring" 2000 200 1 # 22p, 50p, 10/100 "gold ring" 2000 260 1 # 30p, 13/100 "gold ring" 2000 280 2 # 33p, 14/100 "gold ring" 2000 340 3 # 41p, 85p, 17/100 "gold ring" 2000 360 2 # 44p (or 45p, 2.98 to 2.53), 90p, 18/100 "gold ring" 2000 400 1 # 49p, 100p, 20/100 "gold ring" 2000 420 1 # 52p, 21/100 "gold ring" 2000 460 1 # 57p, 23/100 "gold ring" 2000 480 2 # 60p, 24/100 "gold ring" 2000 520 1 # 65p, 26/100 "gold ring" 2000 540 2 # 68p, 27/100 "gold ring" 2000 560 2 # 71p, 140p, 28/100 "gold ring" 2000 580 1 # 73p, 29/100 "gold ring" 2000 600 1 # 76p, 30/100 "gold ring" 2000 640 1 # 81p, 32/100 "gold ring" 2000 720 2 # 92p, 36/100 "gold ring" 2000 740 1 # 95p, 37/100 "gold ring" 2000 800 1 # 103p, 40/100 "gold ring" 2000 820 1 # 106p (3 to 1.94), 41/100 "gold ring" 2000 840 1 # 108p, 42/100 "gold ring" 2000 880 1 # 114p, 44/100 "gold ring" 2000 940 1 # 122p, 47/100 "gold ring" 2000 980 1 # 127p, 49/100 "gold ring" 2000 1000 1 # 130p, 250p, 50/100 "gold ring" 2000 1020 2 # 133p, 51/100 "gold ring" 2000 1060 2 # 138p, 53/100 "gold ring" 2000 1080 1 # 141p, 54/100 "gold ring" 2000 1100 2 # 143p, 275p, 55/100 # rose rhodonite: 15000 wibi, 1312p, 3750p "rose rhodonite ring" 15000 0 1004 "rose rhodonite ring" 15000 150 1 # 7p, 38p, 1/100 # 20p (1269 to 1249, 3525 to 3450), 75p, 2/100 "rose rhodonite ring" 15000 300 1 "rose rhodonite ring" 15000 600 2 # 27p, 150p, 4/100 # 43p (1269), 3525p, 94%, 6/100 "rose rhodonite ring" 15000 900 2 "rose rhodonite ring" 15000 1350 1 # 74p, 338p, 9/100 "rose rhodonite ring" 15000 1500 1 # 84p, 375p, 10/100 "rose rhodonite ring" 15000 1950 1 # 114p, 13050 wibi, 13% "rose rhodonite ring" 15000 2100 3 # 124p, 525p, 14/100 "rose rhodonite ring" 15000 2250 1 # 135p, 12750 wibi, 15% "rose rhodonite ring" 15000 2400 1 # 144p, 600p, 16/100 "rose rhodonite ring" 15000 2550 1 # 155p, 638p, 17/100 "rose rhodonite ring" 15000 2700 1 # 165p, 675p, 18/100 "rose rhodonite ring" 15000 3450 1 # 218p, 893p, 23/100 # small ruby ring: 5000 wibi, 607p, 1250p "small ruby ring" 5000 0 512 "small ruby ring" 5000 100 1 # 9p, 2/100 "small ruby ring" 5000 150 1 # 14p, 3/100 "small ruby ring" 5000 350 1 # 32p, 7/100 "small ruby ring" 5000 950 1 # 86p, 19/100 "small ruby ring" 5000 1050 1 # 95p, 21/100 "small ruby ring" 5000 1450 1 # 131p, 29/100 "small ruby ring" 5000 1500 1 # 138p, 30/100 "small ruby ring" 5000 1600 1 # 147p, 32/100 "small ruby ring" 5000 1650 3 # 153p, 33/100 "small ruby ring" 5000 1700 1 # 159p, 34/100 "small ruby ring" 5000 1750 2 # 165p, 35/100 "small ruby ring" 5000 1850 2 # 176p, 37/100 "small ruby ring" 5000 1900 2 # 181p, 38/100 "small ruby ring" 5000 1950 1 # 187p, 39/100 "small ruby ring" 5000 2250 2 # 221p, 45/100 "small ruby ring" 5000 2500 1 # 249p, 50/100 "small ruby ring" 5000 2650 1 # 266p, 53/100 "small ruby ring" 5000 2700 1 # 271p, 54/100 "small ruby ring" 5000 2800 2 # 282p, 56/100 # scarab: 2500 wibi, 358p, 625p "turquoise scarab ring" 2500 0 462 "turquoise scarab ring" 2500 25 1 # 3p, 1/100 "turquoise scarab ring" 2500 75 1 # 8p, 3/100 "turquoise scarab ring" 2500 100 1 # 11p, 4/100 "turquoise scarab ring" 2500 200 1 # 22p, 8/100 "turquoise scarab ring" 2500 425 1 # 48p, 17/100 "turquoise scarab ring" 2500 450 1 # 51p, 18/100 "turquoise scarab ring" 2500 500 1 # 56p, 20/100 "turquoise scarab ring" 2500 625 1 # 70p, 25/100 "turquoise scarab ring" 2500 750 1 # 85p, 30/100 "turquoise scarab ring" 2500 800 1 # 91p, 32/100 "turquoise scarab ring" 2500 850 1 # 99p, 34/100 "turquoise scarab ring" 2500 900 1 # 105p, 36/100 "turquoise scarab ring" 2500 975 1 # 115p, 39/100 "turquoise scarab ring" 2500 1000 1 # 118p, 40/100 "turquoise scarab ring" 2500 1100 1 # 132p, 44/100 "turquoise scarab ring" 2500 1125 2 # 136p, 45/100 "turquoise scarab ring" 2500 1175 2 # 142p, 47/100 "turquoise scarab ring" 2500 1200 1 # 145p, 48/100 "turquoise scarab ring" 2500 1225 1 # 149p, 49/100 "turquoise scarab ring" 2500 1275 2 # 156p, 51/100 "turquoise scarab ring" 2500 1300 1 # 159p, 52/100 "turquoise scarab ring" 2500 1325 1 # 163p, 53/100 # black gold: 60000 wibi, 2627p / 15000p "black gold ring" 60000 0 2000 # 2627, 15000 "black gold ring" 60000 600 1 # 2614, 14850, 1/100 "black gold ring" 60000 1200 2 # 2600, 14700, 2/100 "black gold ring" 60000 5400 1 # 2506, 13650, 91% # 2614 to 2425, 14850 to 12750, (/ (- 14850 12750) 150.0) = 14 "black gold ring" 60000 8400 1 # 2600 to 2384, 14700 to 12300, 2400 / 150 = 16 (/100) "black gold ring" 60000 9600 1 "black gold ring" 60000 10200 1 # 2398, 12450, 83% "black gold ring" 60000 13200 1 # 2330, 11700, 78% # 2600 to 2263, 14700 to 10950, (/ (- 14700 10950) 150.0) = 25 "black gold ring" 60000 15000 1 # diamond ring: 60000 wibi, 2627p / 15000p "diamond ring" 60000 0 2000 "diamond ring" 60000 1800 1 # 2587p, 14550p, 97% "diamond ring" 60000 9600 1 # 2411p, 12600p, 84% "diamond ring" 60000 12000 1 # 2357p, 12000p, 80% "diamond ring" 60000 13800 1 # 2317p, 11550p, 77% "diamond ring" 60000 24000 1 # 2087p, 9000p, 60% "diamond ring" 60000 26400 2 # 2033p, 8400p, 56% # 1966p, 7650p, 51%, though it's in "fairly poor condition" "diamond ring" 60000 29400 1 # classic gold engagement ring: 28200 wibi, 1906p, 7050p "gold engagement ring" 28200 0 1020 "gold engagement ring" 28200 564 1 # 27636 wibi, 98% "gold engagement ring" 28200 1692 2 # 1830, 6627, 94%, 6 "gold engagement ring" 28200 1974 1 # 26226 wibi, 93% # 27636 to 22842 wibi, 4794, -17% "gold engagement ring" 28200 4794 1 # 1589, ~5287.5 (21150 wibi), 75%, 25 "gold engagement ring" 28200 7050 1 "gold engagement ring" 28200 7614 1 # 20586 wibi, 73% "gold engagement ring" 28200 8460 1 # 1525, 4935, 70% "gold engagement ring" 28200 9024 1 # 19176 wibi, 68% # ruby serpent ring: 50000 wibi, 2402p, 12500p "ruby serpent ring" 50000 0 1513 "ruby serpent ring" 50000 3000 1 # 2335, 11750, 94% "ruby serpent ring" 50000 4500 1 # 2301, 11375, 91% "ruby serpent ring" 50000 5500 1 # 2278, 11125, 89% "ruby serpent ring" 50000 6500 1 # 2256, 10875, 87% "ruby serpent ring" 50000 8500 1 # 2211, 10375, 83% "ruby serpent ring" 50000 11000 2 # 2155, 9750, 78% "ruby serpent ring" 50000 13000 1 # 2110, 9250, 74% "ruby serpent ring" 50000 13500 1 # 2098, 9125, 73%
R code (just a bit of processing and plotting for now):
raw <- read.table("value-drop.txt", header=TRUE)
itemNames <- unique(raw$item)
itemValue <- frame()
for (item in itemNames) {
itemData <- raw[raw$item == item,]
itemValue[item] <- itemData$full_value
}
raw$damage = raw$value_drop / (raw$full_value / 100)
raw$full_value <- NULL
raw$value_drop <- NULL
expanded <- raw[rep(row.names(raw), raw$times), 1:3]
nonZero <- expanded[expanded$damage != 0,]
damageFreq <- frame()
meanDamage <- frame()
maxDamage <- frame()
durability <- frame()
for (item in itemNames) {
itemData <- raw[raw$item == item,]
damageFreq[item] <-
(sum(itemData$times) - itemData$times[1]) / sum(itemData$times)
meanDamage[item] <- mean(expanded[expanded$item == item,]$damage)
# durability is for how many casts an item is expected to last
# without being fixed
durability[item] <- 100 / meanDamage[item]
maxDamage[item] <- max(nonZero[nonZero$item == item,]$damage)
}
par(las = 1, mar=c(6, 10, 2, 2))
boxplot(damage ~ item, nonZero, width=damageFreq[order(itemNames)],
horizontal=TRUE)
text(x = 0, y = 1:length(itemNames),
labels = round(meanDamage[order(itemNames)], 2), col="red")
title(main = "Observed damage values",
sub = "Box widths represent relative damage frequency, numbers in red are for mean (expected) damage.")
summary(lm(durability ~ itemValue))
coef(lm(durability ~ itemValue - 1))
## summary(lm(1 / damageFreq ~ itemValue))
## summary(lm(maxDamage ~ meanDamage))
## plot(data.frame(itemValue, durability, meanDamage, damageFreq, maxDamage))
## library(fitdistrplus)
## descdist(durability / itemValue)
I have checked before this research that skills don't seem to affect damage notably (while they do affect misportal rate).
Ochrion's results with rose rhodonite rings are not far off: 58 rings were destroyed in less than 24569 casts, making it less than 423.6 casts per ring, while the approximate value-to-durability coefficient in the listed samples is 0.02246196; 0.02246196 * 15000 = 336.9294.
Durability (defined as a multiplicative inverse of observed mean damage) appears to depend on item value linearly. Damage frequency changes linearly with mean damage (though the former is one of the latter's components), while maximum damage doesn't seem to be correlated with them as notably, but doesn't look random either. Basically, more expensive rings tend to imply lower mean damage, lower damage frequency, and lower maximum damage too, though it is said (and likely) that it's not all just a function of value internally.
Interestingly, each tested ring's deduced durability divided by its value yields results fitting very well into a normal distribution: this could be just a testing artifact, but could also indicate that, among other options, an AHSB cast adjusts blorp's durability with a random value, with normal distribution. The results listed above are from individual rings (repeatedly fixed).
