FooSW RTA Analytics
Real stats from real Guardian games. See what actually gets picked, banned, and won.
— replays · 542 monsters · Guardian G1–G3 only
Real stats from real Guardian games. See what actually gets picked, banned, and won.
— replays · 542 monsters · Guardian G1–G3 only
Last 14 days in the dataset
Combo rankings and counter-search can use the full Guardian mix or narrow to G1-only or G3-only games from the replay tags, so stats match the bracket you care about.
Use case: You mostly play G1 and want cores that win there, without G3 one-tricks inflating pick rates — flip to G1, then sort by smoothed WR.
Use Counter: opponent has and add the monsters they’re on (or you expect in RTA). The list switches to combos that actually won into that matchup in Guardian replays — real draft data, not guesswork.
Use case: They lock a scary core (e.g. strip + control + threat) and you need team ideas that beat that comp in G1–G3 — filtered to what still shows a strong win rate in those games.
Hand-pick the units you own in a searchable monster box view (saved in your browser) or load SWExporter JSON — ownership filtering is one click away.
Use case: Quickly hide every combo you can’t build so you only see realistic cores for your account.
Toggle whether pick rates, bans, gems/traps, and counters use all Guardian games or top-league only — same UI, two sample sizes.
Use case: Rare monsters need the bigger pool; pure endgame meta questions (e.g. first-pick wars) are clearer on G3-only.
Lock your side and the enemy’s picks slot by slot; the tool surfaces what Guardian players actually chose next in matching situations (plus model suggestions where trained).
Use case: “They took Seara + Oliver after my Ereshion — what do strong players pick third?” Practice concrete branches instead of guessing.
All tools are in the left rail — hover to expand on desktop, or tap 🛠 on mobile.
I got tired of guessing. The tier lists floating around are either paywalled, based on one person's ranked experience, or just copied from each other season after season. I wanted to know what actually wins at Guardian — not what a content creator thinks should win. So I started parsing replays.
Every number comes from real games. No opinion polls, no votes, no vibes. 348k+ Guardian matches, parsed and crunched. Win rates are smoothed to avoid small-sample flukes, and you can always check the game count behind any stat.
It's built for your box, not a whale's. Upload your SWExporter file and every tool filters to what you actually own. Core Teams shows combos you can draft. Build Advisor tells you what to build next. The whole site reshapes around your account.
⚠️ Heads up: data updates manually when new replays are exported — not real-time. Monsters with very few games will have unreliable win rates. Stats on niche picks should be taken with some skepticism.
Built by a Guardian RTA player. Started as a script to answer one question: what are the highest win-rate draft cores at the top of the ladder? Grew from there into something usable.
The data pipeline is manual — I export replays, run the analysis, and regenerate the pages. Not fancy, but the numbers are real G1–G3 games.