lichess.org
Donate

Opening explorer mystery, possible bug

If you enter the moves
1. e4 e5 2. Qh5 Nc6 3. Bc4
(of course the alternative move order is also possible) into the analysis board and open the opening explorer (set on lichess, everything selected), the move 3...Nf6, which has been played 280 times, weirdly has a winrate of 52%. In fact, according to the database, only one player has ever found the incredibly difficult Qxf7#! Even excluding Bullet games and 1600 players there are still 30 games with a winrate of 57% for black. How is that possible?
3. ...Nf6 wasn't played in any master games, and if you include the lichess games it will show all the games on the site and people can troll to get the high % of the worse moves and also notice the one game with the 4. Qxf7 is from a 2200+ player well most of the rest are sub 1800.
It is kind of fascinating to see that less than one percent of lichess player who were in that position played the mate in one. Maybe it's mostly bullet games and white didn't expect to actually be given the opportunity.
@ZenoZero It is to be expected that chess masters know Scholar's Mate, but I wasn't talking about the masters' database anyway. Also it doesn't take a rating of 2200+ to find it. My point is that they cannot all be trolls and there is probably a huge chunk of games missing from the database.

@Callidous Excluding Bullet there are still 234 games.
The answer is that very short games are not included in the opening explorer, leading to this effect.
@revoof Very plausible explanation. I personally think that this should be changed, though. In my opinion traps and quick mates are part of opening theory (especially at my level) and should be treated as such. Of course you'd still have to exclude quick draw agreements and sandbagging, but - for simplicity - just adding checkmates would be a step in the right direction.
Yeah, agreed. I'll try to change it in the next version of the explorer, but it might take a while.
#2
5. Qxf7, actually. The game started 1. e4 e6 2. Qe2 e5 3. Qh5, explaining how it was long enough to be included.

This topic has been archived and can no longer be replied to.