Why does computer say this is a good move?

Here is an excerpt from a game I recently played. I analyzed it and it says that Rb1 is the best move. I don't see how this can be the best move. Here's the game:

Why does the computer say Rb1 is a good move?

Maybe it's because pawn push would be really unnecessary and developing the queen won't be a great idea. If you develop your knight, the pawn on e5 could push to e4, and you wouldn't want to play your knight to the side of the board either. Rook b1 would be a somewhat "great(?)" choice because you won't be needing an open-file on the a-rank as of in this position. These are my opinions but I'm not that experienced in chess. Hope this helps!

No good reason, just an artifact of the free analysis being terrible.  Stockfish actually prefers nf3, and the main "book" moves are nf3 or e4.  A lot of moves are playable, but the notion that an engine would prefer rb1 is quite bizarre.  

That's why some people don't use the analysis just because of this reason. chess.com is a business like some members said in another forum which is a bit similar to this case. 

rb1 supports the b4 break its not bad,its vene devloping a piece its no outrageos

Don't put too much stock in the computer suggestions.

Something I have noticed when analyzing my own games is that sometimes, the computer's suggestions as to the strongest move isn't always practical. It isn't uncommon to see a move that the engine claims is the "strongest" but would never be played by a human.

The reason why is because the engine doesn't take the psychological strength of a move into consideration. Some moves - to a human - look more threatening than others. Obviously, if you move your queen into the neighborhood of the enemy king, your opponent will be looking high and low for threats, even if there aren't any. The queen move might  not be as strong as the best move that an engine can find, but if your move burns 30 seconds off your opponents clock in a 5 minute game, there is something to be said for that.

I tried analysing it but for me it says excellent.

I think its in anticipation of the g7 bishop targeting the rook after the b2 pawn push.

Rb1 is definitely not idiotic, and is one of several addequate moves, but its completely inexplicable that an engine would report it to be the best move.  Stockfish doesn't put it in the top 3, and its like the 7th most "popular" choice in the database I look at.  Respect to all saying it has some logic, but for it to be reported as the best move is just the lousy free analysis.  This is far from a particulary offensive example, but it is an example nonetheless.

I'm pretty sure rb1 is just theory. Other than that, it's just plain old 'out of the blue'. I think that the whole concept is a pretty general pawn push, that white may need in the future to gain space. Maybe the fact that the bishop on g7 is really strong really scares white, and the rook moves out of the following diagonal to prevent following pins. 

" but its completely inexplicable that an engine would report it to be the best move. "

Maybe not if you just let the engine calculate for a few seconds. And I believe that's the essential weakness in the Chess.com free analysis. 

A little off-topic, but isn't Black's setup with e5 followed by a king-side fianchetto a little unusual too?

5.Rb1 may be "excellent," but 5.Nf3 and 5.d4 are "excellenter"

What was the point of the Benko gambit stuff?

Because Humans understand nothing.