Magic Bounce
Pokémon that run it
| Pokémon | Teams | Share of its teams | Win rate |
|---|---|---|---|
 Hatterene | 1,103 | 97.10% | 44.41% |
 Espeon | 49 | 92.45% | 40.00% |
 Clefable | 3 | 0.26% | 38.10% |
 Mega Clefable | 2 | 40.00% | 45.45% |
 Absol | 1 | 0.27% | 50.00% |
 Mega Absol | 1 | 100.00% | 50.00% |
 Mega Sableye | 1 | 100.00% | 0.00% |
 Sableye | 1 | 0.03% | 20.00% |
All-time figures across every format. Method
Magic Bounce FAQ
All-time, whole corpusIs Magic Bounce good in Pokémon Champions VGC?
Magic Bounce is ranked #63 of 189 abilitys by usage, carried by 1,155 teams (1.28% of teams). Those teams have won 44.73% of 11,239 games, below the 50% line. Usage says how many players trusted it; the win rate says how those games went.
Which Pokémon run Magic Bounce the most?
Hatterene, Espeon and Clefable are the top species carrying it, by teams. The win rate beside each one is that species' own record, not this ability's.
| Name | Share of its teams | Win rate | Games |
|---|---|---|---|
| Hatterene | 97.10% | 44.41% | 10,751 |
| Espeon | 92.45% | 40.00% | 465 |
| Clefable | 0.26% | 38.10% | 21 |
| Mega Clefable | 40.00% | 45.45% | 11 |
| Absol | 0.27% | 50.00% | 6 |
| Mega Absol | 100.00% | 50.00% | 2 |
| Mega Sableye | 100.00% | 0.00% | 3 |
| Sableye | 0.03% | 20.00% | 5 |
Is Magic Bounce spread across the metagame, or concentrated on a few Pokémon?
Concentrated: Hatterene alone carries 95.50% of the 1,155 teams running Magic Bounce, and its top three species together account for 100.00%.
How reliable is Magic Bounce's win rate?
The 44.73% figure is a posterior mean, not a raw average; the interval behind it is 44.00%–45.46%, from 11,239 games. That is a narrow band for a figure with this many games behind it.
Where do these numbers come from?
Tournament decklists and real match records for Pokémon Champions, ingested from Limitless, Victory Road and the official circuit. These figures are all-time over the whole corpus, not scoped to a format or window the way the boards above the search bar are. Win rates are shrunk toward the field average rather than printed raw, so a thin sample cannot top a board.