How We Verify Pokémon GO Data

Last updated: May 2026 · Author: Aria T., Data & Stats Editor

This page describes the methodology behind every recommendation on Pogodex: raid counter lists, GO Battle League tier lists, “is it worth investing” stardust analyses, event ticket assessments, and the numerical content of the Pokédex itself. The goal is to make the methodology open and checkable: a reader who wants to know how a recommendation was arrived at should be able to find that here.

Test accounts and trainer levels

We test across multiple trainer accounts and trainer levels. The accounts are real (no mass-account farming, no GPS spoofing, no third-party software) and have been built up through normal play since 2020:

Marcus’s account — Trainer level 50, deep PvP roster, full type-coverage raid roster, multiple Mega Evolutions at maximum levels, used for primary PvE and PvP testing.
Aria’s account — Trainer level 48, PvE-focused, full type-coverage raid roster, used for raid testing and stat verification.
Kai’s account — Trainer level 47, event-focused, secondary raid testing, used in particular for testing weather-boosted scenarios and regional event variations.
Regional raid Discord — ~200 trainers across all level ranges, used to verify duo/trio raid claims with multiple teams and to surface edge cases we would not encounter on our own accounts.

Raid counter testing

Counters that we recommend in a raid guide have actually beaten the relevant raid boss in our testing. The protocol:

For Tier 1 and Tier 3 raids, we test solo and small-team setups; recommendations note the minimum trainer count that succeeded.
For Tier 5 raids, we test trio (where possible duo) success with documented setups: Pokémon level, IVs, fast move, charged move, weather conditions, time-to-kill, deaths.
For Mega Raids, the same protocol with attention to whether Mega-boosted partners are part of the recommended team.
We test in clear weather as a baseline. Where weather boost is part of the strategy, we test in the boosted weather and report the difference.
We document failures as well as successes. A counter list saying “this team won, this team lost” is more useful than one saying “use these ten Pokémon.”

Counter recommendations carry a “tested on [date], game version [x.x]” note so readers can see whether the analysis is current.

GO Battle League testing

PvP tier lists are based on actual ladder play in the current season. Specifically:

Tier-listed Pokémon have been laddered by Marcus or Aria in the current season at relevant rank ranges (typically Veteran through Legend, depending on the league).
For each tier-listed Pokémon, we report a sample size: minimum 50 ladder games before tier-listing a new addition, more for borderline cases.
Matchup spreads (MUS) are documented: “this Pokémon beats X, Y, Z; loses to A, B, C” with the move-set assumed.
Where a Pokémon’s role is matchup-specific or strongly meta-dependent, we say so rather than giving it a flat tier ranking.
Where a tier ranking is changed mid-season due to a balance update, the previous ranking and the reason for the change are documented.

Tier lists carry a “current season: [season name and number]” tag and are refreshed within 48 hours of any move-pool rebalance.

Stat data verification

Numerical content on Pokédex pages — base attack, base defense, base HP, max CP, type, available moves, evolution requirements — is verified against the PokeMiners game master archive. The verification flow:

After each game version release, the latest game master file is pulled.
An automated script compares the new file against the previous version and surfaces every changed value.
Each change is logged: which Pokémon, which value, before vs. after.
Changes are reflected on the corresponding Pokédex pages within 24 hours where the change is significant; minor changes (a 1-point base stat tweak) are batched into the next regular review.
Aria reviews the complete diff each cycle, even when “no notable changes” appears, because the script can miss edge cases (newly introduced fields, schema migrations).

“Is it worth investing” testing

Whether a Pokémon is “worth powering up” depends on the player’s stardust, their existing roster, the meta, and what they want to use the Pokémon for. Our verdict articles test specific claims:

We power up the Pokémon to the recommended level using stardust earned through normal play.
We use it for at least two weeks across the relevant content (raids, PvP, gym defense) before publishing a verdict.
Where the Pokémon’s role depends on a specific moveset, we use Elite TM or wait for community day to acquire the right move; we report which path we used.
We compare against the obvious alternative the reader would have if they did not power this one up — a “you have this and this; if you have neither, here’s the better choice” framing.

Event coverage testing

For ticketed events (GO Fest global, GO Fest regional, special research with paid tickets, ticketed timed research), Kai tests the event in person where possible:

For GO Fest, Kai has attended five regional and global events in person since 2021.
For ticketed timed research, the ticket is purchased on a real account and the research is completed during the event window.
Reward structures, shiny rate observations (where the sample is large enough to be defensible), encounter pools, and field research are documented as observed.
Where observations diverge from Niantic’s announcement, both are reported with the divergence flagged.

What we do not test

To be clear about the limits:

We do not test exploits or third-party tools that violate Niantic’s Terms of Service. Where the community is debating a possible exploit, we describe the debate without facilitating the exploit.
We do not test the upper limits of paid acquisition. We do not buy raid passes in bulk to verify rare encounter rates; instead, we observe across the team’s normal play.
We do not test multi-accounting strategies that exceed Niantic’s family-account allowance.
We do not test region-restricted content from outside the relevant region; where regional content is covered, we rely on community reports from trainers in those regions.

Reproducibility

Where possible, we publish enough detail that another careful trainer could reproduce our finding:

Raid counter setups list specific Pokémon, levels, IVs, and moves.
PvP matchup spreads list the move-set being assumed for each side.
Stat claims include the game version they were verified against.
Event observations include the date and the trainer’s region.

If a reader’s experience contradicts our finding, please email info [at] pogodex [punto] space with the specifics. We treat reproducibility seriously and adjust our recommendations when independent testing suggests we got something wrong.