Form at publish
3 – 1
Set by set
Form at publish
Martin Mastera to win
54.0%
Sealed before the first serve and never touched. The record is only checkable because of that.
Read
Rightnot sent
«Right» is not «we won». No pick was shipped, so this is a read that turned out correct, not a call that won.
Set 1 · 11-8
Set 2 · 12-10
Set 3 · 6-11
Set 4 · 11-7
Point 76 set 4, 11-7 · our read held Martin Mastera at 54.0%
76 points played in total.
H2H Martin Mastera 4–12 Tomas Beran 2000 Martin Mastera has won 4 of their 16 meetings.
Same-day rematch · 1h before this match
Mastera beat Beran 3-1.
Same-day rematch winners hit ~50% historically. The market typically prices this in, which is why it moves nothing in the read below: it is context, not a signal.
54.0%Our read54.6%to break even at 1.833
short by 0.6pp54.6% (what it has to beat, and what the price implies: at 1.833 they are the same number)
Not whether we were right, but which read failed. One match proves nothing: the calibration is the aggregate of rows like these, and that aggregate does not exist yet.
This table lists the interface catalogue: the one you already see in ALIGN / CONFLICT. The engine weighs its own subset, and they are not the same set. It says so here instead of hiding it.
3 for, 1 silent, and still no pick. The gate cuts it all the same.
This room is closed. Messages clear 24 hours after they are sent.
Today's board has its reads written and frozen, the same way this one was at 17:39 UTC. They stay hidden until each match finishes: that is the only reason this record can be checked at all.
Today's board · How the record is graded