What square on a chessboard is the least visited square?

ichabod.  Very cool.

So for me, it's a3 then.

A few years back, I ALWAYS played the ruy lopez, I usually played those squares.

all squares are used in most games

All 64, on mine.

I would not define "visited" as the same as "number of moves occupied." I would define it as "number of times a square is entered." By this definition, I think the least visited would have to be a corner square, a1, a8, h1, or h8. Pawns would never go to these squares except to queen and knights and bishops would rarely go there because their movement is so constrained there. That just leaves the heavy pieces and the king. Kings usually castle kingside, and when they castle queenside, they are still 2 squares from the corner, so kingside corners would be visited more, either by kings seeking safety or by pieces involved in checkmating them. Since white moves first and thus on average must vacate a1 sooner than black vacates a8, thus making it available for visitation, I would say a8 is probably the least visited.

Great work!

I totally called a2!

I did another analysis. It turns out that all 64 squares are used in only about 1.4% of games. The average number of squares used is just under 57. 50% of games use 58 or more squares. These are based on the 2009 World Cup and WCC Qualifier games, and the figures count the starting squares as being "used." Of those, the least number of squares used is 43.

Given the rationale behind en-passant does a pawn passing through a square on the third (or sixth) rank count as its being visited?

What about the squares traversed while castling since a King can't castle through check?

If so what about the squares en-route to the final destination for Rook, Queen and Bishop moves?

h5

z98. Called.

No, actually h5 like mprhchess said.

i cant think of an opening where h5 is commonly played

wait, z98 is actually a good idea. . .

:)

You can also investigate if you make more horses or bishop moves during games- another stupid, undeveloping thing to do.

Knights. From the 2009 games:

1. Pawns Moves: 10,883
2. Rook Moves: 8,871
3. Knight Moves: 7,286
4. Bishop Moves: 6,811
5. King Moves: 5,175
6. Queen Moves: 4,800

Is that per bishop and per rook and per knight?  It's not fair that the Queen is only one...know what I mean?

Oh for crying out loud

haha. 

Good point, those are totals. The per piece counts are:

1. King moves: 2,587
2. Queen moves: 2,400
3. Rook moves: 2,185
4. Knight moves: 1,822
5. Bishop moves: 1,703
6. Pawns moves: 680

Those per piece counts can't be from the same data as post #56 above.. right?

Sure, but I divided by the total number of pieces on the board: 16 pawns, 4 rooks, 4 bishops, 4 knights, 2 queens and 2 kings. Double checking, there seems to be a typo on the rook moves (should be 2195), and maybe I rounded wrong on the king moves. Good enough for government work.