why does the engine appear to contradict itself?

When I play Bxd5 the engine says it is a mistake and Bd3 is the best move.
When I play Bd3 the engine says it is a mistake and that Bxd5 is the best move.
???

[Site “Chess.com iPhone”]
[Date “03/13/2021 11:50AM”]
[FEN rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1]
[White “ChesssAddict (1262)”]
[Black "nequanb (1265)"]
[Result "White Wins"]
[Termination "White Wins"]

1.e4 {10:00} e5 {9:59} 2.Nf3 {9:53} Nf6 {9:58} 3.Nc3 {9:35} Nc6 {9:55} 4.Bc4 {9:31} Nxe4 {9:53} 5.Nxe4 {9:28} d5 {9:53} 6.Bxd5 {9:28} Qxd5 {9:51} 7.d3 {9:18} Bg4 {9:40} 8.h3 {9:15} Bh5 {9:36} 9.g4 {9:13} Bg6 {9:34} 10.Nfg5 {9:04} Be7 {9:15} 11.Qf3 {9:01} f6 {8:52} 12.Nxf6+ {8:58} Bxf6 {8:44} {White Wins}

For me it just says that Bd3 is the best move

Yeah I have this same problem all the time! The strangest part is the actual evaluation number is different when the move is suggested versus when it is played. My theory is that when a move is suggested the engine analyzes it to a shallower depth than when the move is played. This would mean the results you get when you play the move are more accurate and that it's more important to look at the evaluation number than the descriptive word.

 

I typically play on my phone, so maybe it's an issue with the mobile version of analysis, or in general analysis on weaker settings

Horizon effect.

There's nothing useful to be learned from the engine output at such an early phase of the game- this position has been played thousands of times.

For the record, the right move is 6.Bd3- the move 6.Bxd5 gives Black the bishop pair for no good reason.

 

The right move is 4 Bb5, or also 4 d4.

There is nothing wrong with your move.  Its all going to depend on what type of game you want to play.

 

 

May I improve your line a bit?

 

 

Please do.  I like to learn.

I dont think 6.Bd5 is a 'bad' move.  I mean that in relationship to the OP's skill level.