This is due to transpositions.
If you played say, 1. e3 e6 the position would have been rarely reached. But if you continued on to 2. e4 e5 you would get the classic kings pawns game, a position that has been reached hundreds of thousands of times.
You are correct.
The most obvious improvement to white is Bg5 where he avoids the poisioned pawn variation completely, and goes into a line considered the most likely to bust the sicilian if it wasn't for that one line.
A good database should show transpositions with differant numbers of moves by the way, chessbase has this feature.
I have frequently encountered a characteristic of chess.com’s Game Explorer that puzzles me. This is the most recent example. I was using Game Explorer to explore a particular line of the Sicilian that went:
1. e4 c5; 2. Nf3 d6; 3. d4 cd; 4. Nxd4 Nf6; 5. Nc3 Bd7; 6. Bc4 Nc6; 7 f3 a6
At this stage Game Explorer lists 2 games, one won by white and one by black. If you then click on the link to 7. .....a6 you get White’s reply as 8. Be3, but 5 games are listed at this point, 1 won by White, 2 drawn and 2 won by Black. And it continues in this vein. For instance, click on the 9. Be3 link and there are 3 responses listed from 7 games, 1 won by White, 2 drawn and 4 won by Black.
So what’s the story? Is the number of games at any particular point those games on the data base that used that precise order of moves but, when you click the link to get to the next stage, you get other games on the data base that got there by different orders of moves (by transposition, in other words)? If not, what is the explanation?