The Crystal Ball is Riot's set of tournament-wide prediction questions that sits alongside Worlds Pick'Ems. Instead of picking match winners, you predict things like who will average the best KDA or which champion will be most contested. The questions are the same for everyone, so the edge comes from understanding what each one is really measuring. Check lolesports.com for the official rules, deadlines and scoring.
Champion questions
Highest presence counts a champion every time it is picked or banned, so the strongest champions in the patch usually win this. Look at which champions were contested most at recent regional finals and in the Play-Ins, and favour champions that are strong in several roles or hard to ignore in draft.
Highest win rate and highest average deaths per game are both affected by sample size. A champion picked twice can top the win rate list, which is why the tracker only counts champions with a minimum number of picks. Think about which picks will be common but not overplayed.
Player questions
Average KDA rewards players who stay alive and join fights, so safe carries and supports on strong teams do well. Teams that win quickly also give their players fewer deaths. Damage per game favours carry roles on teams that fight for a long time. Kill participation usually goes to supports and junglers, who are involved in many fights.
Questions like Peyz vs Gumayusi total kills depend mostly on how far each team goes, because more games means more kills. Will Faker reach 700 career Worlds kills depends on how many games T1 play, so the tracker starts counting from his 616 kills before Worlds 2026.
One game with 15+ kills or 450+ CS are rare events. Ask how often either has happened in past Worlds before predicting yes.
Event and team questions
Unique champions picked tends to grow with the number of games and with how varied the meta is. Games before or after 32 minutes depends on how fast the dominant strategies close games. The questions about regions in Knockouts, an LCS team advancing, and HLE versus T1 depend on how the groups and Swiss stage play out, so it is worth watching early results before they lock.
Common mistakes
- Picking a player or champion off one great regional run without checking sample size.
- Forgetting that a team that goes deep plays more games, which pushes up totals like kills and damage.
- Ignoring how the patch changes which champions are strong.
- Locking everything on day one. Some picks may be worth changing after the Play-Ins if the rules allow it, so check lolesports.com.
Follow it live
Once the games start, the live tracker shows who is leading every question and refreshes automatically after each game.