Black Just Kept Pushing Pawns... And It Worked

Someone recently suggested that I take a look at this game from Round 6 of the 32nd Abu Dhabi International Chess Festival, which ended a few days ago.
I'm very glad they did.
For the first 18 moves, Black barely developed anything. The queenside knight appeared on c6, but was immediately traded off. The queen moved twice, just to return to its starting square. Almost everything else Black did involved pushing pawns.
It sounds like the sort of strategy you would expect from a beginner. But everything Black played was somehow perfectly legit.
By the time Black finally began developing pieces, White was the one struggling to find a good plan.

The Game

What We Can Learn

The final 29...Bb7! is a beautiful way for the game to end, but I think the most remarkable part came much earlier.
Look again at Black's first 18 moves.
Black developed the knight to c6, but that knight was immediately exchanged. The queen went to c7 and then straight back to d8. Apart from that, Black essentially spent the entire opening moving pawns:
...e6, ...bxc6, ...e5, ...g6, ...h5, ...h4, ...g5, ...d5 and ...f6.
On general principles, this is highly unusual and just feels wrong.
We are normally taught to develop pieces, castle quickly and avoid making unnecessary pawn moves. Black seemingly ignored all three rules.
But chess principles only matter when they fit the position.
The centre remained relatively closed, so Black's lack of development was not immediately punishable. Meanwhile, every pawn move served a concrete purpose. ...e5 controlled the centre, ...h5-h4 chased the knight, ...g5 prevented f4, ...d5 struck before White could play d4, and ...f6 reinforced the structure.
There were definitely weak squares that Black created along the way, but White's pieces were simply not well placed to exploit them.
White actually handled the complications quite resourcefully. 20.Na3!, sacrificing the exchange rather than passively retreating the rook, gave excellent practical counterplay. After Qxc6+, Black's king suddenly looked exposed and the position became really chaotic.
The bigger problem was the clock.
White spent enormous amounts of time during the opening and had only a few minutes left once the position became extremely sharp. In a position with such a high price of a mistake, that time disadvantage ended up being very costly.
First came the unnecessary 24.Qd5+?, which helped Black's king reach g6. Then 27.Qf3?? just flat out left the knight on c4 undefended. And in the end, with barely a minute remaining, 29.e6?? allowed the spectacular 29...Bb7!.
It is slightly unfortunate that we never got to see this completely mad position develop for another ten or twenty moves.
But maybe ...Bb7! is a fitting ending to a game like this.
Black spent the opening breaking almost every development rule I can think of, somehow emerged with a perfectly fine position, and then finished the game with a bishop that had been sitting on its original square for the previous 28 moves.