Wayward Queen Attack - Noone finds mate in 1?!

Here is a game that is in the database where White mates on move 5.


Fool's mate is not in the database. That would be a mate by Black on move 2.
Continuing looking at the Bird opening, I see games where Black could have mated on move 4, but no game with that move in the database.

Of course, this does not mean that such mates were not found in some games; only that those games are not in the database. I think the hypothesis that very short games are not stored in the database is correct.

Also, as others have hypothesized, the cause of the missed mates that are in the database is probably because of premoving.

@Arminius4 said in #7:
> Cool! Any news about how frequent games ended with a checkmate in this position?
At the moment I'm grepping (=searching) the file but it takes forever :D

Results, finally @Arminius4
I found the following exact string 17 times in the August 2021 database of ~94 mio. games:

1. e4 e5 2. Qh5 Nc6 3. Bc4 Nf6 4. Qxf7#

That means there are in fact games with <5 moves in the database. What is used for the opening explorer is a different question.

@MieszkoTheFirst said in #13:
> Results, finally @Arminius4
> I found the following exact string 17 times in the August 2021 database of ~94 mio. games:
>
> 1. e4 e5 2. Qh5 Nc6 3. Bc4 Nf6 4. Qxf7#

Thanks for figuring that out. That are surprisingly few games, even if one considers that there may be different move orders to reach the position, given that the position appeared over 1400 times in the lichess opening DB.

Update @Arminius4
Turns out by searching for the string mentioned above I only got correspondence games, because timed games have the current clock in-between the moves in the data. I now used the following regular expression

1\. e4 .{0,20}1\.\.\. e5 .{0,20}2\. Qh5 .{0,20}2\.\.\. Nc6 .{0,20}3\. Bc4 .{0,20}3\.\.\. Nf6 .{0,20}4\. Qxf7#

and got 6738 results. That is August 2021. I'm running the search for 4. Qf3 now (1st move in the opening explorer) for comparison.

Edit: After that, just for fun, I will calculate the average time to reach 4. Qxf7#. Any other ideas?

In the 6738 games that went like this:

1. e4 e5 2. Qh5 Nc6 3. Bc4 Nf6 4. Qxf7#

the average White Elo was 1250.6 and the average Black Elo 1240.5.

I found 3139 games with

1. e4 e5 2. Qh5 Nc6 3. Bc4 Nf6 4. Qf3

Of those, the average White Elo was 1219 and the average Black Elo 1217.1.

Average time for games with 4. Qxf7#, ignoring the hour-field in the data (only minutes+secs):

314 seconds
29.4% of games were over 5 minutes
4.7% of games over 10 minutes
21.7% of games under 1 minute.

I might have made mistakes though.

...

@MieszkoTheFirst said in #16:
> In the 6738 games that went like this:
>
> 1. e4 e5 2. Qh5 Nc6 3. Bc4 Nf6 4. Qxf7#
>
> the average White Elo was 1250.6 and the average Black Elo 1240.5.
>
> I found 3139 games with
>
> 1. e4 e5 2. Qh5 Nc6 3. Bc4 Nf6 4. Qf3
>
> Of those, the average White Elo was 1219 and the average Black Elo 1217.1.

That is interesting. Because for the position after 3. .. Nf6 only 1426 games are counted in the lichess DB. If you counted 3139 games alone continuing with Qf3, only a small fraction of games seem to be included in the DB, regardless of length.

That also means, that the number 6738 can not be compared to the numbers found in the opening explorer and that it is unclear to me, how they constructed the opening db.

Maybe games shorter than a certain # of moves used to be included in the explorer, but not anymore?