Morra Gambit

After 1.e4 c5 2.d4 cxd4 3.c3, taking the pawn is dangerous for black. He can lose very quickly. So how about 3...d3, disrupting White's development? The King's bishop and queen's knight will be badly placed.

It is definitely not "!" by any means.  White can just take it one of two ways and get development at the same time.  ! is reserved for the best move.

There is almost no danger in dxc3.  White gets an open center but the fact that they get their knight out first would happen unless you went 1. e4 c5 2. d4 cxd4 3. c3 Nc6!? (no clue if this is any good, seems worse than the regular line).  I have always taken on c3 and responded with Nc6 after Nxc3 and never lost to the Gambit.  So I am inclinded to continue that play.

Why is it bad? Can you prove it? Can you lose material? If so, how?

You shouldn't take the morra too lightly. You can safely accept or decline the pawn, the latter by 3...d3, 3...Nf6 (which is what I would advise) or 3...d5. All options, except perhaps 3...d3, lead to approximate equality (says modern theory). If black plays 3...d3 white should play c4 at some point and then develop his knight to c3.This gives rise to normal looking sicilian maroczy positions. I'll give two games from the excellent book on the topic by H.Langrock, one with the accepted variation with e6 and a6, which shows how things can go wrong, and one in the declined variation with 3...d3.

Black has lost time by playing ...d3 too.

Like the BDG, the Smith-Morra is another gambit with which I have a good lifetime of experience in.  I even have a game against Tim Taylor from 1981 that is the stem game in his book "How to Defeat the Smith-Morra With 6...a6".  (LOL, he won of course, but theory has been revised recently with 7. e5!)

The Smith-Morra Accepted is definitely fun to play as white.  However, I find myself retiring it because too few players are accepting it, opting instead for 3...Nf6, where I believe black has almost immediately equalized.  Since ...Nf6 also serves black well against the 2. c3 Sicilian, he gets bonus economy out of the choice.

With respect to 3...d3, the immediate 4. c4 is correct as 4. Bxd3 allows 4...d5.  However, 4. c4 has a dark side that I have pointed out, namely the surprising 4...d2+! (D) where not only does white have to put something on that square he doesn't want to, but he still has to expend a move playing Bd3.

just play 3...Nf6 against the morra, that way you only need to learn c3 sicilian theory + it may take your opponent out of theory.

i dont think d3 is neccessary or good, after Bxd3 white is a pawn up and, as far as i can see, doesnt really have disrupted development. 

the point was made that white cant start a kingside attack easily but white doesnt need to attack, he doesnt need to proove anything, simple development will give white an edge.

d3 doesnt even create any real weaknesses that black can exploit. 

Good Morning.! I don't know this variation but white is very defensive and attack because of strong pawn structures which cannot be decline except any branches to defeat this variation. Maybe, the players decide to own decision as well. Hmm, it seems to be rare generate a new variation now. It depends. Thank you very much and appreciating your post.

Bxd3 is met with d5?? huh, well it loses a pawn to Bb5+ followed by Qxd5.

this is an excellent explanation to how white should play the lethal  gambit...

 excellent annotaion.

this is how white plays it

after d3 then c4,follow with Bxd3 and its hard for black to break with d5