Pro head to head
Two pro players or two teams, every map they met in. Each record sits beside what the ratings expected before each map and the range luck alone would give, because a record between two sides is mostly their strength plus luck.
Pick two teams to see every map they played against each other. For example, T1 against Gen.G →
How to read this
Expected is the rating forecast from before each map: it knows the players and the teams, not the champions. Added up over the maps, it is how many wins their strength predicted.
The luck range is where the wins land about 19 times in 20 when only strength and luck are at work. Measured over every pair of pro teams with ten or more meetings, records sit from the forecast about as far as luck puts them, so a record inside the range says who was stronger, not that one side has the other's number.
Ratings are the pro TrueSkill ratings the forecasts are built on. A team's is the average of its five players'. A card shows the rating now; each map shows the rating each side carried into it, never one that already includes that map's result.
The record counts finished maps only. The one forecast for a game not yet played is "If they played today" at the top: the pro model before the draft, from each team's latest five.