Sound Gambits?

I think the Latvian Gambit is the worst. I don't think I've ever won a game playing it. Elephant Gambit is good too.

Budapest but I never use gambits.

If we are talking philosophy, i' ll give You a question:

What does constitute a gambit for You? And what does it mean to be sound?

A gambit is an opening, in which one side sacrifices material. However, You seem do be defining gambit as an opening, in which the sacrifice is unsound, or at least dubious. It is IMHO not a definition of a gambit. If you want a sound gambit, play Queens Gambit it is 101% sound.

On the other hand, What does it mean "sound" to You? Do you want opening adventage? Equal game? There are people, who claim that Kings Gambit is "sound' as it leads to draw with best play. Seemed good enough for one Boris Spassky. The fact that it is white who is desperetly fighting for that draw is here just a tiny detail. In that school of though any gambit that doesn't lose ( like KG) is "sound".

Besides, "sound" does not mean "atractive" . Vienna gambit ( 1. e4. e5. 2.Nc3, Nf6, 3. f4) is clearly sound, as white's superiority is  clear after it is accepted. The problem is that black always declines, when while nothing horrible happens, the position tends to quicly dry out and simplify, proving the gambit to be drawish.  White risks nothing but also gets nothing.  The gambit shines only against the insufficently prepared, making it unatractive for profesionals. Note that no gambit that leads to better position of the gambiting side will be ever accepted.

Answering your question: Yes, there is no reason not to add a gambit into a repertoire.

A) it can be something sound like queens gambit B) It might be a tricky way to get another position through transposition. Agains Urusov gambit both Lasker and Alekhine recomended to decline ( a testimony of soundness!) and to transpose as black to two knights defence. Great for white if You like playing against two knights defence. 

C) It is nice to have such a suprise weapon. Even Kings Gambit pops up now and then on high level, when one tries his/her luck and tests opponents memory. You could add, say Scotch gambit - another "sound" but "unatractive" gambit ( to my humble knowledge)- well prepered black will draw against  any line, but white is not taking much risk.

Thank you for the in-depth response. Yes, definitions are core to Logic and Philosophy (which I tend to enjoy as well). We can all agree on gambits having their place (as a surprise weapon, versus the unprepared, for casual games etc.)

However, I am going from the premise of "if my opponent is well studied" (hence "best play" by human standards - computer "best" we won't get into, but let us say two human grandmasters versus each other with said gambit opening in question).

Here is how I think of a gambit:

"A position (entire opening variations in this case) where one player sacrifices material (typically a pawn or two, but sometimes even more) for compensation. The compensation usually takes the form of positional assets or the initiative." 

What makes a gambit "sound" by my perspective and definition then? I consider the gambit sound if and only if the compensation is long-lasting against "best play" (again, "best play" is objective, but thought of as prepared grandmasters in this case - not necessarily computers, but it may be computers if the entire game of chess was solved. Since it is not currently "solved", computers often change their "best line" evaluations, so that is why I think of human play here.)

By my definitions and opinion, I believe the King's Gambit is not sound. I actually like the King's Gambit a lot. One friend of mine is a USCF Class A player and he uses the King's Gambit as an integral part of his opening repertoire under classical time controls. The opening has a lot of theory present and there are enough complications in it that the better prepared player will likely succeed. For this, I believe the King's Gambit is a good opening that needs to be taken seriously; this is not the same as "sound" to me though. I believe that with best prepared play by Black, White does not have adequate compensation. I would anticipate seeing the King's Gambit played more often at the highest levels if it was sound, but I think it is rare nowadays because Black seems to prove White's attempts unsound.

This is a lot of technical thought; let me rephrase it another way:

Pick any opening gambit. Say the Staunton Gambit with 1. d4 f5 2. e4. Is there any single line that Black can play to demonstrate White's pawn sacrifice is dubious? Black must assume White is prepared for the opening (no surprise factor) and White must assume the same for Black (objectively assessing the merits of the gambit opening in question). If Black can play a line that White can't avoid AND Black ALWAYS gets a better position from it (enough compensation for the material or getting the material back with the better objective position), then the gambit is unsound. Is the Staunton gambit actually sound? Maybe not, but that is what I am trying to discover. 

Nothing personal against the Staunton Gambit though - maybe it will be revealed in the future that we all missed something and now the opening is considered sound again (chess has yet to be "solved" yet). Pick another opening: repeat the process...is this gambit sound? By the end of it all: which gambits are sound, if any?

Ideally, I would love to find a clear "refutation" to every opening gambit in existence. The few (if any) that appear to objectively have compensation would then be considered the only "sound gambit openings." I wonder which opening gambits are sound and how many there are. 

Ultimately, I would have nothing against adding a gambit to my opening repertoire if I knew it was objectively sound (then time control wouldn't matter objectively). However, soundness is critical because I don't want to be playing openings that are objectively dubious; that feels like the advanced version of "hope chess" - we would then be reliant on the opponent being unprepared and I prefer to give them the benefit of the doubt.

I was scrolling and reading these long comments but I have to cause the power is going to be out soon.

The Queen's Gambit is the only gambit I hear mostly

"Even Kings Gambit pops up now and then on high level, when one tries his/her luck and tests opponents memory. "

Yeah and that's when the World Champion loses a miniature...

Yes, that is correct - but if we are mentioning memory, then that explains Magnus Carlsen versus Wesley So with Carlsen playing 1. f3 2. Kf2 in a fairly serious game.  So laughed it off and took it well though - I know a lot of GMs would throw a fit if someone played that against them. 

Other than surprise, I'd say the reason is more "I can play whatever I want."

https://www.chess.com/article/view/why-do-people-play-garbage-chess-openings 

But yes, even if some refutation does exist, it only really matters if the opponent can replicate it OTB. Not remembering the critical line(s) years after studying them is part of it, but for this thread: I am assuming the opponent is well-prepared and remembers the line(s) if such line(s) exist.

True and ironically, the "Queen's Gambit" is no longer considered a "true gambit." Apparently thought this because it seems sound with best play. Clearly, it is still a gambit by my definition though (because 1. d4 d5 2. c4 offers to sacrifice a pawn even if adequate compensation exists). It is almost as if an underlining assumption for most people is that a gambit, by nature, is already unsound. Most players who call the Budapest, the "Budapest opening" probably believe it is sound lol. It is called the "Budapest Gambit." If it is sound or not is not relevant (I doubt it is sound, but the complications are certainly good enough for most to play it anyway) because Black is sacrificing a pawn (hence "gambit"). 

I presume that you're not counting pseudo gambits like the queen's gambit where white reclaims the pawn in qga? Look at GM games and see what gambits are used, I'd suggest. Can't see a solid refutation of the anti Moscow semi slav, and honestly it feels like white gets more than enough compensation. It's also the main way for white to sharpen up the opening, so if you play d4 and prefer sharp positions, or are at a grade advantage...

I came across this while reading David Smerdon's Scandinavian

While I personally feel the opening is sound, this is uncharted theory. At the time when David wrote his book 5...Bxf3 was a novelty and I could only find one game where it has been played since then. 

Why does the Queen's Gambit not count as a "true" gambit?

Because white can easily win the pawn back.

Very interesting history. My opinion is that the Scandinavian is probably unsound, but complicated enough to be taken very seriously with human play (like how I feel about the King's Gambit). Bobby Fischer himself considered 1...d5?! to be dubious after 1. e4. In fact, Fischer literally laughed at his opponent once when Fischer played 1. e4 and they played the Scandinavian Defense. 

In regards to this specific variation, it is unstudied so at least it has that going for it in practical play. The history you gave is interesting, but if I was White, I'd just play Be2 and castle Kingside swiftly. I think White is better there, but that is just me. 

In a real game, chances are the opponent would at least be unfamiliar with the line and psychology may add some merit to the line. Objectively, theory or study means nothing for soundness; but in practical play, these factors sometimes matter a lot.

One gambit I play that is sound and fits any definition of a gambit is the 2 knights defense vs the Italian. 

@KeSetoKaiba , the nxd4 variation in the max lange attack does have a gambit line

Yes, I am familiar with the Max Lange Attack. In fact, our DUNE team member @Arrakis09 and I have gone over that opening in great detail before.

I agree and I liked your personal anecdote from 5 to 10 years ago. I know exactly what you mean; I was like-minded to that when I began chess too. Interestingly, I've also grown in a maturity since then; I didn't attribute it to endgames though. However, that very well might have played a factor; I've always liked endgames and it is true that endgames can teach one a lot about position. A lot of successful chess players have studied "backwards" by essentially reverse-engineering chess. They study openings in relation to the middlegame and the middlegame in relation to the endgame, but as Capablanca has stated: "...but the endgames can be studied in and of themselves."

p.s. How did you quote within a quote like this? I see @MichalMalkowski with their flag and all inside the quotation as well. 

Scotch gambit