返回提示词目录第 1460 页

设计竞技游戏排位体系

为竞技游戏设计隐藏分、段位、晋级、赛季重置及作弊处理等完整排位方案

完整提示词

<context>
You design ranked modes for competitive games. Most ranked systems separate a hidden skill rating used for matchmaking (Elo-family systems or Bayesian ones such as Glicko or TrueSkill that track uncertainty) from a visible rank that players see and chase. The visible layer is a motivation system and can be more generous than the rating, but if the two drift far apart players feel cheated. Design decisions that matter: how new players are placed without wrecking matches, how many points a win or loss moves, how team games credit individuals, whether seasons reset ranks, how inactivity decays, and how to handle smurfs, boosting, win-trading and leaving. Small player populations force trade-offs between match quality and queue time.

Game:
{{game}}

Population:
{{population}}
{{#problems}}

Current problems:
{{problems}}
{{/problems}}
</context>

<task>
1. Goals: what ranked should give players (a fair match, a sense of progress, a status symbol) and the trade-offs for this game and population.
2. Rating and tiers: the hidden rating approach with a plain-language explanation, the visible tiers and divisions, and how visible rank maps to the rating.
3. Placements and new players: starting rating and uncertainty, number of placement matches, and protection for new players from experienced ones.
4. Gains and losses: how points move per match, adjustments for uncertainty and win streaks, team versus individual credit, and tier protection rules, with example numbers.
5. Seasons and resets: season length, soft reset formula, rewards, and decay for inactive players.
6. Abuse and edge cases: smurfs, boosting, win-trading, deliberate losing, leavers and disconnects, duo or party queue gaps, and what to do when the queue is empty.
7. What to measure: match quality, queue time, rank distribution, and churn after losses, with targets to set and how to adjust.
</task>

<constraints>
- Mark numbers as starting values to tune with live data.
- Explain rating systems accurately at a high level and do not claim a specific published game uses a specific formula unless widely documented.
- Do not reward deliberate losing or exploit-prone behaviour; check every rule for abuse.
- For small populations, say explicitly what is traded away (match quality, tier count, region split).
</constraints>

<output_format>
## Goals
## Rating and tiers
Table: Tier | Share of players target | Rating band.
## Placements and new players
## Gains and losses
## Seasons and resets
## Abuse and edge cases
Table: Problem | Rule or detection | Trade-off.
## What to measure
</output_format>

可替换内容

  • {{game}}
  • {{population}}
  • {{#problems}}
  • {{problems}}
  • {{/problems}}

分类与标签

设计创意、游戏机制、排位系统、竞技设计、赛季规则

前往 AI 图片创作