Italian Game VS Ruy Lopez

I use a database of master games (2000-2012 years, 110.000 games aprox. all players +2400 FIDE).

After 1.e4 e5 2.Nf3 Nc6  3.Bc4

white obtain 54,3% of total points and the draws are 52%.

After 1.e4 e5 2. Nf3 Nc6 3.Bb5

white obtain 54,9% of total points and the draws are 52%

the results are very similar, I don't understand why the Italian Game hasn't a good reputation. 

It does have a good reputation. I don't understand why you would think otherwise.

Just because 2700+ rated GMs don't play it as often as another opening, doesn't mean it's bad. Most of them probably grew up playing the opening against their friends over and over and over again.

People at the top level tend to prefer the Ruy Lopez, I think, because there are a whole variety of different plans involved and there's a huge mass of opening theory to it, making it slightly more difficult, I would imagine, to prepare against the Spanish Opening than the Italian one.

However, that's not to say that you can't play very good games with the Italian Game. You can. People do.

There are more chances for advantage in the Ruy Lopez. Anyway that won't be a problem for a majority including 2000+.

At the amateur level the Italian Game has a fine reputation. At the professional level the Italian Game I think suffers from a combination of being very very thoroughly explored (it was one of the most popular openings of the 19th century) while lacking strategic depth.  Ruy Lopez is every bit as old as the Italian and also has a ton of theory attached. But it has a strategic depth which has GMs thinking there are still things left to discover about it. 

"I don't understand why you would think otherwise."

I read bad things for the Italian in many books. For example the "Claves del ajedrez práctico" (Secrets of practical chess, John Nunn).

"There are more chances for advantage in the Ruy Lopez" yes, In theory but why this more chances doesn't traduce in a best results?

I play the Ruy but If I can change for the Italian and play games with similar succes and less study of opening theory I can concentrate my time in other parts of the game. 

A lot of the 'quiet' d3 lines in the Italian are similar to Ruy lines. I don't think there's any reason that they aren't sufficiently complicated. Kramnik has played then, as have many Super GMs, I'm sure.

I play the Italian and tried to pick up on the Ruy. The Ruy is far more complicated. And as Black, I'd rather see White play Bc4 than Bb5 (for when I don't play the French ).

52% of points includes 50% of every draw...

After 1.e4 e5 2.Nf3 Nc6  3.Bc4

in 100 games white wins 28 (28.3) games, draw 52, lose 20 (19.7) games

obtain 54,3% of total points and the draws are 52%.

After 1.e4 e5 2. Nf3 Nc6 3.Bb5

in 100 games white wins 29 (28.9) games, draw 52, lose 19 (19.1) games

white obtain 54,9% of total points and the draws are 52%

ah, I get it now. Thanks.

How could those percentages be used to choose an opening repertoire?

It's the name. Only the name. The only way for white to get interesting positional play is to play Giuoco Pianissimo, which sounds like pessimism, which is bad. Therefor, the ruy lopez.

The ruy lopez is fun because almost all structures can happen, and grandmasters can outplay opponents and computers, all moves are logical and you can change from system to system without many problems. Hard for white to prepare.

How could those percentages be used to choose an opening repertoire?

With statistics a player can view if a opening o variation has a good score in the practice of masters. You need a sample with a significant size. In this case I talk about 1.340 italian games. Normally I search for openings that white obtains a better score without a great % of draw. For example I view the Four Knight's Scoth line, and this line has a good score for white 53,5% but 76% of the games are draw (in 100 games white wins 15.5 games, draw 76 and lose 8.5. 

The specific selection of openings depend of the player but I think the statistics may help.

In my opinion, The Italian is more aggressive as White (compared to Ruy and Scotch) but also allows counterplay from Black.

There's a reason why the Ruy has a long history...

Almost every opinionated pronouncement on these subjects is nonsense.

Most of all, these opening statistics that seem to make every beginner think they can provide an authoritative source for the relative merits of different openings are just complete nonsense, for at least five reasons:

1. The sample size for any opening - and therefore the relative strength of statistical information - is biased by fad and fashion.
2. The relative strengths of the players of the games making up the statistics is not taken into consideration. Because of point number one, this is actually a real problem.
3. At any given move number - especially the earlier moves - the overall statistics include data from clearly inferior lines, some of which have been played more often than the better lines.
4. An entire tree of variations may have been refuted in one branch, but the statistics do not reflect this, because only a few games essaying that refutation may have been played before good players will have finally given up on a once-popular tree of lines.
5. It is very likely that chess is a theoretically drawn game, like checkers, as opposed to a theoretically won game, like Connect Four. Because of this, ALL playable openings, when examined far enough out, will result in equality in all main lines.

The proclamation that the Italian is somehow worse than the Spanish (or vice versa) is nonsense. As a matter of fact, the main line Spanish and the Giuoco Pianissimo reach pretty much the same type of position, except that in the Italian version, Black has not overcommitted on the Queenside or sacrificed a pawn for immediate equality and a kingside attack, and White may not have engaged the tension in the center, so that in the Giuoco Pianissimo, there are more strategic choices or both sides, and more choices means more tension, and more tension means more chances to err (and thus to capitalize on errors). With proper play by both sides, either situation will resolve itself in an equal position.

The reason Nunn and his parrots say that the Italian is "busted" and serious players should move to the Spanish right away is because of an underlying assumption, that has been treated as a law by the best players since Fischer showed the way. That assumption is that the fight for the initiative in the opening is best carried out by the sharpest means possible. At this point in time, this is indeed the most practical approach, because technique has reached a point where all other known methods of fighting to maintain the initiative have been blunted.

If you buy into that premise, and also want to maximize your early winning chances (as opposed to maximizing your overall winning chances), drop the Italian and learn the Spanish.

Reason: the sharpest lines of the Italian are either unsound or avoidable, whereas the sharpest lines of the Spanish are the main lines, and cannot be avoided unless Black wants a clearly worse position. On top of that, these sharp main lines are perfectly logical, and are therefore easily learned.

But as I implied, that is not the only way to fight for the initiative, nor is it objectively better than any others, and in fact it comes with a great deal of risk. The word "sharp" implies not only the point of a sword, but its edges as well; even the most skilled swordsman risks being cut by his own sword in a swordfight. Sharp play implies a relatively simple directness when successful, or double-edged complications when not; if you prefer to be less direct, or prefer less double edged complications, feel free to ignore the biased pronouncements of Nunn and friends, who assume that their dogma is the one true way.

Each opening has its own advantages and disadvantages. If you don't mind deferring the fight for the initiative, go ahead and even play the Three Knights Opening or the Ponziani - each is a perfectly fine opening; there are other ways to fight for the center than a concerted attack on Black's e-pawn, and there is much more to chess than Bf1-b5-a4-b3-c2.

For the record, the first known mention of both the Italian and the Spanish occurred in the same source, and the implication then was that the Italian was better for White than the Spanish. The Spanish wasn't widely accepted among masters until the late 19th century, and the theory of "the Spanish torture" was in nacent form well into the 20th century.

As a matter of fact, it wasn't until around the mid 19th century that anyone outside of the Italian School accepted 2.Nf3 as the equal of 2.f4 or 2.Bc4.

First, I'm desaagre with the conclusions about openings statistics. 

1) The size of the sample change with the fashion, this is true. However a sample of hundreds (+1000 in the Italian) of recent master games is enough for obtain a statistical significance. The strength of sample increases a lot if it pass from 30 to 60 or 100 to 200 samples, but when you have a lot of samples increase the number more and more only improve a little this strenght. 

2) I'm In agree, but I talk about a database filter with players of relative similar strenght and the average Elo of the players of to sides are similar. 

3) But Practice include this inferior moves, and the opponents play this lines in the tournaments. This lines have interest because are played and influence in the overall.

4) This is valid, and for this extract conclusions with few games is risky.  I only expose statistics  of the first moves for don't use a Mass of numbers and data. However the branches refuted (completly o partial) arise in very few master games and for example the Giuco Piano (the c3+ direct d4) only arises in 48 games. 

5) Likely this is True.

Second, I think your post is a very instructive discurse about Spanish Opening and Italian Opening. Thanks for this.

I did not intend to talk about the objective strenght of the Spanish on Italian my intention was talk about the practical viability or not of the Italian.  

Is the "viability" of an opening not related to its "strength"?

I think you misunderstood my point in 3 - it was more a potential than an actual problem.

Say you are trying to decide between two opening moves in a line for White... One has the following stats: 65-20-15, and the other has 37-35-28. The raw stats seem to suggest that White should prefer the first, but upon further investigation, it turns out that on the next move the stats are 32-38-30 for one move, and 95-5-0 for another, and more games were played in the second line, because the former line is a new way for Black to steer the game to avoid those dismal results, and has been played few times. Thus the original 65-20-15 is revealed as in no way representing the true stats of the opening, which, based on what we know so far should actually be 32-38-30, and so the second of the two original choices (the one with stats 37-35-28) - assuming its variations don't have a similar defect - should actually be preferred.

My point was that opening stats are not all that helpful. All that matters are the results of best play by both sides.

JM3000, you are wrong.

One line that is great for white is played 1000 games with great results. Then somebody finds a refutation of that line and starts playing the countermove for black, now black wins. Strong players (the one in the database) are aware of these novelties and really soon everybody gets the news, nobody plays that line for white anymore. But its great scores that has been accumulated over time is not "updated", it remains in the database. People in the field (i.e. titled players) know that it's bad, then a guy like you comes with his database nonsense and gets smashed.

Even without this, the fact that GMs achieve a certain results in a particular line is totally irrelevant to any non-titled player.

Just play what you like among the things that are considered popular and sound. Italian is perfect if you like it. I prefer the Ruy, but that's just my personal preference. Looking at database results is a waste of time.