The Muzio, as most folks know, is a gambit involving the sacrifice or one or more entire pieces in the opening. There is both the Muzio and the Double (or Wild) Muzio. Most theorists consider the Double-Muzio refutable and Muzio itself rather tame at high level play, since, as David Bronstein put it in his book on the 1953 Zurich International, "after a short fight in the center, the pawn structure simplifies, and the fight which follows becomes flat and featureless." But, at the same time, Bronstein always encouraged the practice of playing gambits to improve ones' understanding of sharp positions and the interplay of material vs. time. Concerning the Double Muzio, Bronstein, in his book, 200 Open Games, maintained that "If one was to pick out at random only one variation from the multifarious King's Gambit (giving the main line of the Double Muzio), that in itself would be sufficient to earn for the opening the eternal gratitude of chess-players."
Whether or not such openings as the Muzio are suitable for high-class tournament play, they do help train chess players in certain aspects of the game while providing canvases for the creation of artistic and interesting games.
This article doesn't intend nor pretend to be a treatise on the Muzio, but rather a brief exploration of its history and application.
First a little fascinating history. The opening moves 1. e4 e5 2. f4 exf4 3. Nf3 g5 4. Bc4 g4 5. O-O were, as far as is known, first explored by Giulio Cesare Polerio around 1580. Five years earlier, Polerio had accompanied Leonardo da Cutri to Madrid where Leonardo defeated Ruy Lopez in their famous match. In 1606 Polerio, in turn, was beaten in a match with a man named Don Geronimo Cascio. This was at the home of Polerio's backer, Giacomo Boncompagne, who also employed Cascio after his victory. Cascio isn't well known today, but was described by Pietro Carrera, who knew him, as a beautiful youth of great talent, a virtuoso and lover of truth as well as a gambler with backers. Carrera succeeded in defeating Cascio at the court of Francesco Branciforte, the Prince of Pietraperzia, in the feudal town of Militello val di Catania.
Alessando Salvio, one of the best chess players of that time, wrote about the chess exploits of Leonardo da Cutri in his 1634 book Il Puttino. In this book, Salvio also mentions a third-rate player named Alexander Mutio (Signor Mutio d'Alessandro) who apprised him of Cascio's use of the opening first displayed by Polerio. Over a century later (1813), Jacob Henry Surratt, an English player/writer, authored a book named A New Treatise of the Game of Chess. A mis-translation of Salvio's work led Surratt to claim that the gambit presented by Polerio was the invention of Senor Muzio (a misspelling of Mutio) rather than Cascio, the practioner (or Polerio, the probable inventor). Since then, the opening has been called the Muzio Gambit even though shortly after, other writers recognized Sarratt's mistake and started calling it the Muzio-Cascio Gambit. In the latter part of that century, the great historian, Antonius van der Linde, systematically studied the actual works of Polerio and the gambit earned the name of the Polerio Gambit or the Polerio-Muzio Gambit. To further compicate this convoluted tale, Salvio gave the game played by Cascio with Italian or free castling. In free castling, " the Rook could be placed on any square up to and including the King's square, and the King could be moved to any square on the other side of the Rook." [wiki] In the Cascio game, the Rook ended up on f1, while the King landed on h1, giving White an irresistible attack. But even Polerio in his later codex offered the traditional castling sequence (King on g1).
In the laboratory the gambits all test unfavorably, but the old rule wears well, that all gambits are sound over the board. - William Ewart Napier
I've also seen it expressed that Theory favors Black, while Practice favors White. Generally, most games are determined more by the relative skills of the adversaries than by the particular opening, although sometimes surprises and preparation in openings can neutralize disaparities in skills. Gambits are often won on these points, but those reasons are really not recommendations. For a gambit to be recommended, at least at the higher levels of play, it must be either totally sound with winning chances following good play or the theory must be so convoluted that, even with suspected unsoundness, in practice the winning chances are high.
A short stroll through the Chess.com master games database gives us a glimpse into the practical chances of the moves leading into the Muzio Gambit. Below is a summary of those moves with some of the data. The key to the summary is-
% White wins : % draws : % Black wins (for reason I don't know, some don't add up to 100%)-
1. e4 e5 2. f4
40.3 : 19.5 : 40.3
King's Gambit Accepted
1. e4 e5 2. f4 exf4
38.4 : 18.8 : 42.8
KGA - move 3 white
1. e4 e5 2. f4 exf4 3.Nf3 -main line for Muzio
39.6 : 17.9 : 42.5
1. e4 e5 2. f4 exf4 3. Bc4
35.6 : 20.6 : 43.8
1. e4 e5 2. f4 exf4 3. Qf3 (Breyer Gambit)
36.7 : 13.3 : 50
1. e4 e5 2. f4 exf4 3. Be2 (Tartakower Gambit)
39 : 22 : 39
KGA - move 3 black
1. e4 e5 2. f4 exf4 3.Nf3 g5 - main line for Muzio
42.7 : 13.2 : 44.1
1. e4 e5 2. f4 exf4 3.Nf3 d5
40.5 : 24.8 : 34.7
1. e4 e5 2. f4 exf4 3.Nf3 d6
34.5 : 15.4 : 50.2
Other options that look favorable for black, but with much fewer games include 3...c5 in which black won all 3 games; 3...Ne7 in which black won 47.8 and white only won 28.9 of the 90 games. and 3...h6 in which black won 47.4 vs white's 37 in 135 games.
KGA - move 4 white
Here the sampling begins to shrink substantially, so we'll examine what we have:
1. e4 e5 2. f4 exf4 3.Nf3 g5 4. d4 (Rosentreter Gambit)
52.3 : 0 : 38. 6 - 44 games
1. e4 e5 2. f4 exf4 3.Nf3 g5 4. h4 ( King's Knight's Gambit)
42.1 : 14.7 : 43.2 - 373 games
1. e4 e5 2. f4 exf4 3.Nf3 g5 4. Nc3 (Quade Gambit)
40: 20 : 40 - 5 games
1. e4 e5 2. f4 exf4 3.Nf3 g5 4. Bc4 - main line for the Muzio
41.8 : 11.4 : 46.8 - 220 games
KGA - move 4 black
Here it gets crazy:
1. e4 e5 2. f4 exf4 3.Nf3 g5 4. Bc4 h6
16.7 : 0 : 83.3 - 6 games
1. e4 e5 2. f4 exf4 3.Nf3 g5 4. Bc4 Nc6 (Blachly Gambit)
23.5 : 23.5 : 52.9 - 17 games
1. e4 e5 2. f4 exf4 3.Nf3 g5 4. Bc4 Bg7
37.1 : 14.7 : 48.3 - 116 games
1. e4 e5 2. f4 exf4 3.Nf3 g5 4. Bc4 g4 - main line leading into the Muzio
48.3 : 0 : 44.9 - 89 games
KGA - move 5 white
1. e4 e5 2. f4 exf4 3.Nf3 g5 4. Bc4 g4 5. 0-0 (Muzio Gambit)
47.4 : 0 : 47.4 - 57 games
1. e4 e5 2. f4 exf4 3.Nf3 g5 4. Bc4 g4 5. Nc3 (MacDonnell Gambit)
62.5 : 0 : 37.5 - 8 games
1. e4 e5 2. f4 exf4 3.Nf3 g5 4. Bc4 g4 5. Ne5 (Salvio Gambit)
42.9 : 0 : 47.6 - 21 games
1. e4 e5 2. f4 exf4 3.Nf3 g5 4. Bc4 g4 5. d4 (Ghulam Kassim Gambit)
50 : 33.3 : 16.7 - 6 games
KGA - move 5 black
1. e4 e5 2. f4 exf4 3.Nf3 g5 4. Bc4 g4 5. 0-0 gxf3 - main line Muzio Gambit
42.5 : 0 : 52.5
KGA - move 6 white
1. e4 e5 2. f4 exf4 3.Nf3 g5 4. Bc4 g4 5. 0-0 gxf3 6. Qxg3 - main line Muzio Gambit
42.5 : 0 : 52.5
Now, while no real conclusions can be drawn from such limit input there are several interesting thoughts and observations that arise from the above data.
1. the KG itself seems to be a dead draw percentage-wise, with actual draws quite low.
2. the KGA offer slightly less chances that the KGD.
3. White's move 3. Nf3 (which can lead to the Muzio) seems to offer White the best chances.
4. In this dbase, 4. d4 is substantially more successful than 4. Bc4
5. Although the sample is tiny, Black seems to have great success with 5...h6. in fact, 5...Nc6 and 5....Bg7 all show much better success for Black than 5...g4.
6. White's has 4 reasonable choices for move 5 after 1. e4 e5 2. f4 exf4 3.Nf3 g5 4. Bc4 g4 .
All but 1, the Salvio Gambit -5.Ne5 - gambit the Knight.
Let's wrap part I up by looking at games involving the 4 move 5 choices just with an eye to the potential of each opening:
Salvio Gambit (5. Ne5)
MacDonnell Gambit (5. Nc3)
Ghulam Kassim Gambit (5. d4)
Muzio Gambit (5. 0-0)