When I attributed the preference for QGD over QID to fashion it was an opinion, or at best an educated guess. When LeeRoth questions whether fashion even exists today, it's a valid question, but I think the facts he presented are far from conclusive. We should proceed methodically if we want to draw firm conclusions.

• I have no problem with only considering 2700+ games, as long as we take it as just an arbitrary number, rather than as a judgment that GMs below that number are unable to have correct and possibly even complete knowledge of theory. I must say though, I thought the recent chesspub gold standard was correspondence games.
• I think it would be better not to mix rapid games with classical games. It's well known that risky openings are considered more playable in rapid. That may not have anything to do with QID vs QGD, but still I think it best not to muddy the waters. Anyway at Tata Steel, after 1.d4 Nf6 2.c4 e6 3.Nf3, my informal count shows 3...d5 eight times and 3...b6 zero.
• When we notice the 100% preference for QGD at the top, this *still* could be due to fashion. This is not stubbornness on my part, it's just the way fashion works. More on this below.
• Even saying that the top players are careful and concrete does not in any way prove there is no fashion. A logical inference that is consistent with (a) 100% preference for the QGD, (b) careful and concrete preparation, and (c) the existence of fashion, would be: "the QGD is no worse than the QID".

So the question becomes, what would be decisive evidence that the QID is worse than the QGD, or indeed that there is no such thing as fashion at the top today? These are not at all the same question. Certainly the QID could be worse than the QGD and yet maybe say the Rossolimo Sicilian is popular due to fashion. And as I noted above, even positing there is no fashion at all wouldn't prove the QID is worse than the QGD, at least not on current evidence. I won't pretend to be able to answer these questions all by myself, simply because different people will have different standards of proof.

A little digression on fashion in chess openings....

I hope we can all agree that in the past there was such a thing as fashion in chess openings. When any of the world champions, or to a more limited extent any other top player, took up an opening, the whole chess world followed. But maybe my next hypothesis is a little radical: A top player didn't take up an opening because it was theoretically good; it was the other way around. An opening became theoretically good simply because a top player took it up. Opening theory was essentially a tug-of-war, and whichever side had the bigger heavyweights pulling for it would "win" the theoretical battles, at least in the short run. Well, you may not agree with my hypothesis, but if you did, then you would necessarily have quite a high burden of proof for concluding that opening A is absolutely better than opening B!

And how is theory chosen today? Everybody uses engines, but let's not be glib about their impact. We should go slowly.

1. Preparation time is limited. At some point a human is deciding to use an engine on this opening and not on that opening. So even in an engine-driven world there is still room for fashion. It's at least plausible that fashion could still exist.
2. Despite their enormous strength (even compared to our 2700 GMs they are +700 or so), there are still some openings where engines do not do well, and this is not always due to insufficient depth.
3. Engines are causing  a kind of schism in chess openings. It used to be that openings were on a continuum from top-tier, to second-rate, to slightly dodgy, to dubious, and finally unsound. Because of their ability to analyze to great depth, engines are driving a wedge, forcing openings into basically two camps: either they lead to equalilty, or they lead to a huge advantage. (For crying out loud, even the Modern Benoni is being analyzed to equality!) How huge doesn't matter to the top players. If it doesn't lead to equality, it's in the bin. In olden days white used to expect += and try for +/-, but today that borderline +/- seems to have disappeared and it's more about getting something "interesting" and seeing how the opponent handles it.

And here I will state an opinion and just leave it out there for others to discuss. Given a shortage of preparation time, and given the engine's relative weakness in certain types of positions, it stands to reason that it might be very much harder to make some openings "work" compared to other openings, even if there is nothing fundamentally wrong with the opening. In my humble opinion, the QID belongs in the eventually equal bucket, but when trying to demonstrate that equality the engine doesn't help you much. So the top players choose to spend their time either on something more engine-suitable (e.g. the QGD), or on something where the engine's unsuitability carries more risk for the opponent (e.g. the KID).

After 1.d4 Nf6 2.c4 e6 3.Nf3, the move 3..d5 is the overwhelming choice of elite players.  Looking at the Tata Steel A group and the recently-completed Opera preliminaries, 3..d5 was played 25 times, 3..b6 was played twice, and 3..a6 was played once.  If you look back at high level tournaments from 2020, I think you will see roughly the same.   

I don't think this is simply fashion in the sense that 3..d5 and 3..b6 are equally as good and they are picking 3..d5 just because.  3..d5 appears to be regarded as more solid and more flexible than 3..b6.  Indeed, I wonder if "fashion" is even still a thing given how carefully and concretely today's top players prepare their openings.

JFugre wrote on 02/01/21 at 13:00:05:


I am surprised by this comment, and respectfully skeptical. 

May I ask - what makes you assert this? Have you discussed this with many of the, say, top 20, top 100, or top 500? 

My impression is quite the opposite - that stronger GM players are very tuned in to the strengths and weaknesses of differing softwares, and are very up-to-date. 

Noting Shanklnd's suggestion that QID is not trusted so much nowadays, we can helpfully restrict this to players likely to be somewhat respected in that orbit - say 2650 plus. Of course there are so many random GMs of 2400-2500 level who are clueless with tech, and those are of little concern.

I will critic my own posts that mention statistics. A one-year look is often too short a duration. Sometimes a year differs greatly from the prior year.

I've done some more searches in the QID, specifically in lines where white fianchetos and answers ...Ba6 with b3. A few systems have been recommended for black.

A general result is that black does very well (around 50%) in an unrestricted search. Black may run into trouble when the games are restricted to players rated over 2500. It might not be the best opening to play against Sam Shankland.

I'm no Sam Shankland, but that assessment can't be correct. I mean, white can get "pressure" in the QID, but not more than in the Catalan or the QGD. Isn't it just fashion and move orders?

• Against the Catalan black needs something else.
• Against the English the Hedgehog has lost popularity.
• Nimzo-Indian players have way too many choices against 3.Nf3: Benoni, Ragozin, Semi-Tarrasch, etc.

At club level, it is strictly a matter of preference.

At the top level, against 3.Nf3, the elites prefer some form of the QGD (including defenses such as the Ragozin and Semi-Slav).

The QID is a bit out of fashion at the top.  Per Sam Shankland:

“Not so long ago, the Queen's Indian Defence was a mainstay in elite-level chess, and it would seem unfathomable to think that just 15 years later, it would be mostly gone. In general, I think it is the defence against 1. d4 that suffered the greatest hit to its reputation at the hands of computers in recent times. I wouldn't be so bold as to say I have conclusively proved an advantage, but I think I can put Black under a lot of pressure.“

FreeRepublic wrote on 01/29/21 at 17:22:20:

It's an important question, but not necessarily a good question. There's a difference. If the OP had been willing to return here and follow up on the one response they did get, perhaps it could have turned into a useful discussion.

I think the QID requires black to be able to play many different structures, compared to the QGD which might need knowledge of only three or four (Carlsbad, IQP, perhaps QGA, and perhaps hanging pawns). And as MNb ironically points out there are still some move order issues to be resolved.

I think this is a very good question. It was asked almost a year ago, and only received one response. I think the reason for the question is not readily apparent. It relates to a variety of issues/questions.

What if you play the Nimzo-Indian defense. What do you pair it with, if white plays 3Nf3? There is the QID, QGD, Bogo-Indian, modern Benoni. Blumenfeld. Anything else?

What if you tentatively decide to play the Nimzo with the QGD? Well why bother with the Nimzo in the first place? Why not simply play the QGD all the time? As an added bonus, the QGD pretty much provides a ready answer against the English opening.

Personally I like the forthright character of the queen's gambit declined. However, it involves learning the exchange variation, Bg5 variation, Bf4 variation, and the Catalan. If you play the Nimzo and the QGD, the exchange variation is largely neutralized. Also, Ne2 vs the Tartakover is gone. It may be worth it.

What if you play the Nimzo and QID instead? Well, no Catalan to worry about. The Nimzo and QID seem to be very good openings in both theory and practice. Still, there are many variations in the QID to learn. Also, you need an answer to the English opening. I think 1c4 e5 is good for black, but it certainly benefits from preparation and experience.

In theory, the Nimzo, QID, and 1c4 e5 provide a rock solid, with chances, repertoire and is the way to go. In practice, there are many variations to learn. Do they suit your style?

The Nimzo, QGD, and 1c4 e5, is a lot of work also. While this might suit my style better, statistics on the QGD are far from stellar from black's perspective.

Simply playing the QGD is an option. It's just one opening, despite several variations. Black plays directly and logically. Unfortunately, black provides no impediment for white, who can likewise play directly and logically. I think theory and practice are more difficult here.