Showdown TierShowdown Tier Methodology
How we collect data and calculate competitive statistics
About This Methodology
Showdown Tier combines usage data with team-level performance data to describe the competitive environment. This page explains the collection scope, filters, metric definitions, Tier rules, and limitations behind the reports published on this site.
The figures are descriptive estimates from our sample. They should be used as context for team building and metagame analysis, rather than as a guarantee of an individual battle outcome.
1. Data Collection
We collect battle logs from publicly spectatable battles on Pokémon Showdown. We collect a sampled subset of those battles to respect server capacity; this is not a complete census of every battle taking place on the platform.
The current ingestion pipeline includes only records with a confirmed winner and loser. Draws are not included in the statistics.
Collection is limited to major competitive formats. Our daily pipeline follows four broad stages:
- Collect publicly observable battle logs.
- Store cleaned battle records in the analysis database.
- Aggregate usage, win-rate, rating, and team data once per day.
- Generate and publish the format reports and detail pages.
2. Aggregation Window & Filters
Aggregation window
Reports use a rolling window covering the most recent 14 days of available battle data.
Rating range
The analysis includes all rated battles available in the collected sample.
Pokemon inclusion
A Pokémon is included in the report when its usage rate is at least 1.0% within the filtered sample.
Update frequency
The collection, aggregation, and report generation pipeline runs daily.
3. Metric Definitions
- Usage Rate
- Recorded team appearances as a percentage of the two team slots per unique analyzed battle: Usage Rate = (W + L) / (2 × unique analyzed battles) × 100. W and L count wins and losses for teams containing the Pokémon. In a mirror battle where both teams include it, one battle contributes two recorded team appearances, one win and one loss.
- Win Rate
- The percentage of recorded team appearances that are wins: Win Rate = W / (W + L) × 100. This is a team-outcome measure, not a claim that the Pokémon's isolated contribution caused the win.
- Lead Rate
- The percentage of appearances in which the Pokémon is sent out as one of the team's leads.
- Selection Rate
- The percentage of times the Pokémon is selected for the battle from the registered team. This metric is shown only for formats where the battle data records team selection.
- Average Rating
- The appearance-weighted mean of the post-battle ladder rating of the player whose team contains the Pokémon. Each recorded team appearance contributes that player's post-battle rating once. Rating bands instead classify each battle by the mean of both players' post-battle ratings.
- Pair Observations
- Partner and opponent cards count recorded pair observations. Their displayed frequency is pair observations / unique analyzed battles × 100, expressed as pair observations per 100 analyzed battles. One battle can contribute two observations when the same combination is recorded on both sides, so this value can exceed 100. It is not the probability that a pair appears in a battle.
4. Tier Calculation
Each Pokémon receives one of six Tiers, from S through E. The calculation first places its usage rate into one of eight bands, then applies the corresponding win-rate thresholds below. Thresholds are inclusive at the lower boundary: for example, 50% usage belongs to the “50% or more” row, and a 53% win rate belongs to the “53% or more” column.
| Usage rate | Win ≥60% | Win 58–60% | Win 55–58% | Win 53–55% | Win 52–53% | Win 51–52% | Win 50–51% | Win 49–50% | Win 48–49% | Win 47–48% | Win 45–47% | Win <45% |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 50% or more | S | S | S | S | S | S | A | A | B | B | C | C |
| 40% to <50% | S | S | S | S | S | A | A | B | B | C | C | D |
| 30% to <40% | S | S | S | S | A | A | B | B | C | C | D | D |
| 20% to <30% | S | S | A | A | A | B | B | C | C | D | D | E |
| 10% to <20% | S | A | A | B | B | B | C | C | D | D | E | E |
| 5% to <10% | A | A | B | B | C | C | C | D | D | E | E | E |
| 2.5% to <5% | A | B | B | C | C | D | D | D | E | E | E | E |
| Less than 2.5% | B | B | C | C | D | D | E | E | E | E | E | E |
The design intentionally requires a stronger win rate from less frequently used Pokémon. For example, a Pokémon at 20% usage and 53% wins is Tier A, while a Pokémon at 2.5% usage and 53% wins is Tier C. This helps distinguish popularity from evidence of performance in the sample.
5. Team Archetype Analysis
An archetype is a group of teams that share a common three-Pokémon core. When an additional Pokémon appears in at least 80% of the teams belonging to that core, the displayed core expands to include it. The same rule can expand a core beyond four Pokémon when the data supports it.
Display conditions
- The candidate pool is the top 20 Pokémon by usage or every Pokémon with at least 1.0% usage.
- An archetype must occur in at least 20 teams to be displayed.
- Archetype usage and win rate describe the teams in the group, not an isolated Pokémon.
6. Limitations
- Sampling introduces estimation error, and small samples can produce noisy rates.
- The ladder population is not evenly distributed across ratings; the middle rating range may be overrepresented.
- Pokémon Showdown ladder trends can differ from official battles played on a game console.
- A team-level win rate does not perfectly measure an individual Pokémon's contribution to a victory.
- Historical formats and data availability can change, so comparisons across periods require care.
7. How We Differ
Many statistics sites focus primarily on usage, moves, items, or abilities. Showdown Tier makes Pokémon win-rate tracking the center of its analysis and combines it with usage, rating, matchup, and team-synergy context. Our goal is to make it easier to ask not only which Pokémon are popular, but also how the teams using them perform in the sampled environment.