Methodology

Where the data comes from

Results are aggregated from the World Beyblade Organization's public Winning Combinations thread, where tournament organizers report the combos used by each event's top three finishers, supplemented by the community event database at MetaBeys (duplicate events are removed). Events are primarily 3on3 / deck format. We normalize part spellings (Takara Tomy and Hasbro names, letter codes like “GB” for Gear Ball) into one canonical part database sourced from the community wiki.

This is top-cut data - read it accordingly

Organizers report the decks that finished 1st, 2nd and 3rd - nothing below that. So every deck on this site already made a top 3. That has three consequences:

The tier list compares proven finishers against each other, not against the whole field. Even a C-tier blade here reaches top cuts - it just converts them into 1st-place finishes less often than the meta picks.
A blade that's missing isn't rated "bad" - it simply has no reported top-3 finishes. Absence is the strongest negative signal in this data, stronger than a low tier.
Usage means the share of top-cut decks, not of everything played on the day. We never see the decks that went 0-2.

Why there's no "winrate" here

Nobody publishes match-by-match results for Beyblade tournaments - organizers report which decks finished top 3, and that's it. And since a 3v3 deck rotates three beys within a single match, even a match result couldn't be credited to any single bey. So any per-battle "winrate" for a blade would be made up. Everything on this site is built from what's actually knowable: which beys were in the top-3 decks, and which of those decks won.

1st Rate, and the 33% baseline

Since everything here already top-cut, the fairest question left is: once a blade reaches the top 3, how often does its deck go on to finish 1st? That's 1st Rate. Throughout the site, "1st" means exactly one thing: the deck that took 1st place at the event. Finishing 2nd or 3rd counts as being present in the top cut, not as a win.

Three decks make each top cut and exactly one of them finishes 1st, so pure chance gives any top-cut deck about a 33% shot. That's the baseline to read the number against: 35% over thousands of events means the blade turns top cuts into wins more often than chance; 28% means it usually rides along in 2nd and 3rd place decks. The colors do this reading for you.

Two things to keep in mind:

Very popular blades get squeezed toward 33%. A blade in all three of an event's top decks "wins" no matter which deck takes 1st, and scores exactly 33% there. So a blade like Wizard Rod holding 35% while sitting in most top cuts in the game is a real, sustained edge, not a rounding error.
Small samples swing wildly. A blade with 3 top cuts and 1 lucky win shows 33%; with 2 wins it shows 67%. The Top Cuts column is the sample size behind every number on the row. Below 30 top cuts we treat a blade's stats as tentative: it keeps its spot in the ranking (small samples are pulled toward average first), but its tier shows as ? until there's enough data to judge.

The stats, and how to read them

Tier - the one-glance verdict, blending popularity and performance (see below).
Win Share - the share of all 1st-place decks that included it. “70% of tournament champions ran Wizard Rod.” It counts wins without dividing by appearances, so it blends how much a part is played with how often it wins - one intuitive number where bigger is simply better.
1st Rate - of its top-3 finishes, how often its deck took 1st place. Chance baseline ≈ 33%; the section above explains how to read it.
Usage (detail pages) - the share of all reported top-cut decks running it. In placement-only data this is the strongest signal there is: it's thousands of competitive players voting with their decks.
Top Cuts - the number of top-3 decks featuring it: the sample size behind everything else on the row.

All stats are computed inside whatever filter scope you set (region, date range).

How tiers are calculated

Like u.gg, we blend performance with popularity, and we keep small samples from making big claims:

Each part's 1st place rate is blended with 60 imaginary "perfectly average" appearances before scoring. A part with thousands of real appearances is barely affected; a part with a lucky 3-for-3 weekend gets pulled close to average - so it can't instantly outrank the established picks.
A popularity bonus rewards consistent presence (each extra appearance counts a little less than the one before). This deliberately weighs heavily: since the data only contains top cuts, reaching them again and again is itself success - a blade that top-cuts constantly but converts to 1st slightly below average is still a strong blade.
Each part's score is then measured by how far above or below the established field it sits in the current filter scope, and cut at fixed thresholds into X, S, A, B, C. X is the meta. C is the floor, not a failing grade: everything listed already reaches top cuts, so there is no "bad" tier here.

Blades with fewer than 30 top cuts (combos: fewer than 3) are ranked with the same smoothing, but their tier shows as ? until they have enough results to judge. You can hide them from the lists with the checkbox above each table.

Part names

Takara Tomy and Hasbro release the same parts under different names - Dran Sword is sold as Sword Dran in the West, Hells Scythe as Scythe Incendio. The TT / Hasbro switch in the header changes every name on the site and remembers your choice. Search understands both.

Affiliate links

Product links may include affiliate tags that support the site at no extra cost to you.