smash.watch

Understanding Metrics

Smash Watch uses two primary metrics to evaluate player performance. This guide explains how they're calculated and what they mean.

Weighted Win Rate

What It Is

Weighted win rate is not just total wins divided by total games. It's a sophisticated metric that emphasizes:

  • Set count - Players who compete more frequently
  • Opponent quality - Wins against stronger players count more
  • Recency - More recent results may be weighted higher (implementation-dependent)

Why It Matters

A 60% weighted win rate against strong players is more impressive than 80% against weak opposition.

Example:

  • Player A: 80% raw win rate, mostly against locals
  • Player B: 60% weighted win rate, regularly beating PR'd players

Player B likely shows on the chart as having both a lower win rate BUT higher opponent strength - indicating they're competing at a higher level.

Opponent Strength

What It Is

Opponent strength measures the average skill level of opponents a player has faced. It's calculated based on:

  • Historical performance of opponents faced
  • Placement results of those opponents
  • Iterative strength calculations across the scene

The system calculates strength recursively - beating someone who beats strong players increases your opponent strength more than beating someone who loses to everyone.

Interpreting Values

Values are relative to your dataset, but general guidelines:

RangeCompetition Level
< 0.3Local/regional level competition
0.3 - 0.6Established regional/national level
> 0.6Elite/top-level competition

Important: These ranges vary by region and game - use them as relative guides within your filtered data.

Why These Metrics Together?

Using both metrics together reveals player profiles:

Scenario 1: High Win Rate, Low Strength

Profile: Local dominator

  • Consistently winning against weaker competition
  • May need to travel to tougher events to test themselves

Scenario 2: High Win Rate, High Strength

Profile: Elite player

  • Winning consistently against top competition
  • Likely PR'd regionally or nationally

Scenario 3: Low Win Rate, High Strength

Profile: Challenger

  • Competing above their current level
  • Taking losses but gaining valuable experience
  • Watch for improvement over time

Scenario 4: Low Win Rate, Low Strength

Profile: Developing player

  • New to competitive play or struggling in current bracket level
  • May improve by focusing on fundamentals

Comparing Players

When comparing two players:

  1. Similar opponent strength? → Compare win rates directly
  2. Similar win rates? → Player with higher opponent strength is likely more skilled
  3. Different on both axes? → Context matters (see profiles above)

Limitations

What These Metrics Don't Show

  • Peak performance - Only averages over the time window
  • Consistency - A player might have wild variance
  • Character matchups - No matchup-specific data
  • Format - Singles vs doubles, best-of-3 vs best-of-5
  • Recency - 3-month data includes old results

Data Quality

Metrics are only as good as the underlying data:

  • Players who rarely compete have less reliable metrics
  • Small sample sizes lead to outliers
  • Regional isolation can skew strength calculations

Next Steps

Understanding the Chart

The scatter plot is the core visualization in Smash Watch. This guide explains how to read and interpret it.

View Types

Smash Watch offers two view modes optimized for different analysis scenarios: State View and Tournament View.