No checkmating material left on the board still it was declared win to whom, has the Bishop,

@Toadofsky

> Perhaps more to the point... here's a game Nordlandia won on time (although Black very likely will play Rxf5 after which it is difficult but possible to imagine how Black could lose):

Suppose black will do random moves and white will do best moves. Then it looks like probability to win game 1 is 10000 times less that probability to win game 2. (Game 1 is game of the first post, game 2 is the game your refered to)

Or may be even 1 000 000 times.

More on probabilities in game2:

With probability > 0.9 black will move both rook and king away to let white eat black pawns. Then there exist probability < 0.5 that black rook will take white pawn, but then with probality > 0.5 white will promote his other pawn to queen. So the probability to win game2 is more than 0.1.

Now try to find the probability to win game1.

@Nordlandia It looks that probability to win your game2 is more than 0.1 if white do his best moves and black do random moves.

@Sarg0n

> That's why! (random example) How often do you think you see a blue moon here?

It is of comparable probability if weak winning side do his best moves and strong loosing side do random moves. In fact, KN+KB example may have probability 100 times less. But I need more calculations.

If we're talking about random moves, why not start talking about takebacks and illegal moves?

@Toadofsky because they are takebacks and illegal moves. // Capitain Obvious

I just created a small and dirty program and see that probability to mate in originating post may looks like 10⁻⁶ and looks like it is more than 10⁻⁸. It may be even 10⁻⁵, but I do not beleive in this.

Quite a big probability, but still much less than in your game2.

lichess.org/analysis/standard/qqqqkqqq/qqqqqqqq/qqqqqqqq/1qqqqqqq/q2qqqq1/qq1q1q1q/qqq1P1qq/4K3_w_-_- How do you evaluate whites chances here?

> lichess.org/analysis/standard/qqqqkqqq/qqqqqqqq/qqqqqqqq/1qqqqqqq/q2qqqq1/qq1q1q1q/qqq1P1qq/4K3_w_-_- How do you evaluate whites chances here?

I am too lazy to think about it. FIDE rules seems to be not appicable to this.

Ok how about this: lichess.org/editor/rnbqkbnr/qqqqqqqq/8/8/8/8/4P3/4K3_w_-_-

Looks like less than 10⁻⁸, but again I am too lazy to calculate it more precisely.