Unipeg 稀有度模型
uPEG 采用的一种 NFT 稀有度算法。一份抛砖引玉的草案 —— 公式自洽、可独立 复算,欢迎指正与改进。1. 概述每个 Unipeg 是一个由若干部位(每个部位有一个 id)和若干颜色随机 生成的形象。玩家关心的「稀有」由两件事构成:哪些部位凑在了一起,以及 几个部位的 id、几处颜色撞在一起。 稀有度定义为该 Unipeg 图案出现的概率取负对数:稀有度 = -ln P(图案) 「图案」= 这个 Unipeg 的可观测配置:哪些部位出场 + id 的同值分组 形状 + 颜色的同值分组形状。P(图案) 越小(越罕见)⇒ 稀有度数值越大。2. 生成模型一个 Unipeg 逐槽位随机生成。 10 个部位槽位:body、eyes、hair、horn、wings、tail、accessories、 legsFront、legsBack、ground。 出场(presence):部分部位每次必出场,部分部位按固定概率随机出场。类型部位出场概率条件出场hair0.80tail0.80wings0.30accessories0.30horn0.20必然出场body、eyes、legsFro...
Unipeg 稀有度模型
uPEG 采用的一种 NFT 稀有度算法。一份抛砖引玉的草案 —— 公式自洽、可独立 复算,欢迎指正与改进。1. 概述每个 Unipeg 是一个由若干部位(每个部位有一个 id)和若干颜色随机 生成的形象。玩家关心的「稀有」由两件事构成:哪些部位凑在了一起,以及 几个部位的 id、几处颜色撞在一起。 稀有度定义为该 Unipeg 图案出现的概率取负对数:稀有度 = -ln P(图案) 「图案」= 这个 Unipeg 的可观测配置:哪些部位出场 + id 的同值分组 形状 + 颜色的同值分组形状。P(图案) 越小(越罕见)⇒ 稀有度数值越大。2. 生成模型一个 Unipeg 逐槽位随机生成。 10 个部位槽位:body、eyes、hair、horn、wings、tail、accessories、 legsFront、legsBack、ground。 出场(presence):部分部位每次必出场,部分部位按固定概率随机出场。类型部位出场概率条件出场hair0.80tail0.80wings0.30accessories0.30horn0.20必然出场body、eyes、legsFro...
Unipeg Rarity Model
A rarity algorithm for NFTs, as used by uPEG. A draft shared to invite scrutiny — every formula is self-contained and independently reproducible.1. OverviewEach Unipeg is a figure randomly generated from a set of parts (each part carries an id) and a set of colors. What players perceive as "rarity" has two ingredients: which parts came together, and how many part ids — and how many colors — coincide. Rarity is defined as the negative logarithm of the probability of the Unipeg's pattern:rarity...
Unipeg Rarity Model
A rarity algorithm for NFTs, as used by uPEG. A draft shared to invite scrutiny — every formula is self-contained and independently reproducible.1. OverviewEach Unipeg is a figure randomly generated from a set of parts (each part carries an id) and a set of colors. What players perceive as "rarity" has two ingredients: which parts came together, and how many part ids — and how many colors — coincide. Rarity is defined as the negative logarithm of the probability of the Unipeg's pattern:rarity...
