How do you convert this +3.5 position? I ended up in a losing position.

I thought Nc3 was main line

Hmph. I've never really payed attention to checking whether my opponent has something or nothing to do. I am able to recognize it in really obvious positions though.

That’s one of the main things I do before moving

It's very useful, and you're at a level where it should start making sense.

Where it comes from is being aware of where each player will seek play (queenside, center, or kingside) and how they will initiate it (piece play or pawn break). In general you seek play in an area where you have a superiority in space, a superiority in force, or ideally both.

I learned about it from the book Pawn Structure Chess by Soltis. Others recommend Pawn Power in chess by Kmoch.

lmao I don't have either of those pawn structure books, I have chess structures by Mauricio Flores Rios

Oh yeah, I think you mentioned that before. It sounded good.

I'm old though

Keep in mind that I am rated roughly 1800 rapid, 1700 blitz, so we are of equal strength. I thought about the position for 3-4 minutes and one idea that I came up with is to play for the a4 pawn break. However, I can't make it work. In some lines which I calculated it would work but only if black decides to take the a4 pawn which he is not obligated to do. Therefore, I would move to another plan of activating the last piece which is kind of doing nothing. I would play Bc3 followed by Be1. I would try to transfer some pieces (queen and bishops) to the kingside. Meanwhile the opponent will not be able to push c5 for a very long time so we don't have to worry about that. He has a more or less permanent weakness. Also, I have to add that it is weird to me that black played h6 on move 10 instead of c5. Had he played c5 on move 10 he would not have the c pawn weakness. Am I missing something about the move 10.. c5 is there a reason why 10..c5 is not a good move?

Doesn't e4 blunder Bxf4?
Yes, white's knight is a lot more active then black's light square bishop, but that bishop is doing a specific purpose of defending the c pawn. He is removing the defender and making black constantly have to worry about the possibility of white winning that pawn. That trade ties down black's rooks to the defense of the pawn. Although, placing a knight on e5 or a5 would apply more pressure to the pawn. It is generally better that your pieces attack instead of defend. If your pieces are attacking and your opponent's pieces are defending, almost always that means that your pieces are better. That's why keeping your knight and his bishop on the board would have probably been a better decision because your knight is attacking a weakness and his bishop is defending it, therefore your knight is better then his bishop.

It doesn't help to use engine analysis before doing your own. It prevents you from learning for yourself how to evaluate and assess a position. As a rule I never read the posts of others before assessing and commenting, which accounts for some clangers but which allows me to look at the position with fresh eyes and to think for myself. Here, I think BlueEmu and I both reached the plan of Rac1, Nc5, g3 and e4, along with any intervening moves necessary. As I mentioned, I wouldn't recapture on c5 with the d pawn if takes on c5 because it helps black free his game and g3 and e4 is possible for white. All black really has is g5 with an attempted K-side attack. White has to be careful of that, of course.

It was 5. ...c6 that was black's appalling move. That plan is more suitable a little later in the opening when white is already mainly developed but in the QGA, it's terrible unless black uses it to transpose to a book line Slav, the nuances of which are already worked out and the lines of which have to be followed by black. In the QGA black plays c5 or possibly Nc6 if easy development via e5 isn't possible. Not c6, as a rule, which loses a move since black should play c5 at some stage to pressurise white's centre.

Incidentally, white's fifth, 5. Nc3, isn't a great move at all. It isn't a blunder so much as rather weak because it allows black the plan of c5, a6 and b5, where a4, as usually played to attack b5, isn't much good because b4 by black hits the knight. It's more normal for white to play a4 against that plan, followed by Nbd2 if ... b4 by black. It's one reason why the QGA can be so effective at lower levels for either side when that side really knows their stuff.

In this game, black confused the QGA with the Slav and so did white!

I haven't even noticed that black played c6 on move 5, I flied through the opening moves and I assumed it was a slav since black had a pawn on c6. In many Slav lines you have to play c5 to liberate the bishop on b7 that's why it seemed natural to me for black to use the opportunity to play c5 on move 10. He instead played h6 and white immediately punished him by playing b4 stopping c5 basically for the rest of the game. 
In my second comment I was talking about the position which occured after Rac1. 
As you said both black and white confused QGA with the Slav and so did I while looking through their game, I guess that's why I have roughly the same rating as them.
 

It was the first one that came up in Amazon search

This is what I was trying to get at in a few posts.

Can also come up via Reti and QGD (yeah decline/then accept).

But you basically covered it all at one time.

Both messed up, but it's Black that should suffer.

The game I posted, White managed to lose. Check those engines (maybe +3.0).

Something I forgot to mention...

Both players (below 1800) played many really nice moves in this game. That was impressive for the ratings imo

And since the discussion has gone to earlier moves and older books...

Nimzowitch's My System, early in the open files part, shows using exchanges to move a pawn from a half open file. In the game here, a knight was traded on e4, maybe d5 would be ideal if the goal is to open the c-file (a point for playing e4 first).

iirc, about the time white played Rfc1, was the reason due to possible opening of the a or b files? I didn't go back and look...hard to do when typing a post.

Yes, of course. If that line had been a Slav then white, playing an early e3, signalled a Meran, and in the Meran, black has three main methods of continuing development. One is Bb7, one is a6 and one is an immediate b4. But black needs to concentrate on developing the queenside, since white is already commited to e3 for the time being and is just developing carefully. In this game, black played the Meran move and then abandoned it and thought he was in a QGA.

Exactly

What I like about this thread is that most replies are actually to the point and trying to be helpful. Kudos to everyone!

What I find highly interesting though is that almost everyone tries and tackles this position with moves and lines and only a few stronger players answered with ideas. But is that the answer to OP's question? Playing a position is not just finding the right moves, it's understanding it too. For that reason, I just want to add my 2 cents. 

When I look at the position after move 20 and I want to get a good feel for the position, I will try and look at all the different aspects of the position. We have a little bit of space advantage in the center but for the rest that's pretty equal, we have a half-open c-file, while black has a half-open d-file, we both have a bishop pair, while Bb2 and Bb7 are bad bishops and the other two are our good bishops. Black's knight has a nice outpost on d5 or potentially at c4, while our knight has nice outposts at c5 or e5. So far, there's not that much difference in the position. But when we start looking at the weak pawns, things get worse for black. c6 is the weakest pawn ever, as it's on white's half-open c-file. a6, a3, e6 and e3 also are kinda weak, but nowhere near as weak as that pawn on c6. Also, white is much more active. That Bd6 is really being hemmed in by the white pawns, while Bf3 looks through that d5-knight to c6. That Nd5 really isn't doing all that much, as only the e3 pawn is in need of protection, but can't be attacked more and is easily defended by the white queen. The white rooks clearly are more active than black's counterpart. We should also notice that this is a closed position and that typical play in closed positions will be playing for pawn breaks and usually playing on the flanks. 

If black's this passive, you should look at his plan even more than you would otherwise do. In closed position that usually means looking for pawn breaks. Clearly black's only chance is on the kingside, but it's really hard to see anything happening there soon. If he's ever going for ...g5-...g4 white can respond with g3 and Bg2 and then respond thematically in the center with e4.

If we look at our own plans, we can easily see that putting pressure on c6 is going to tie black down to the defense of that pawn. And with black not being able to do anything (that kingside plan is a bad idea, we would welcome black trying that), we can take a bit of time to prepare everything we have. So 21. Rac1 should be a good candidate, maybe play g3 at some point to get ready for a possible e4 and Bg2 (profylaxis, black can't do anything, so let's protect our bishop). After that we can look again. 

What I'd be worried about as white is Bb2. There's no real way of making it into a useful piece right now though. I could play Bc3-e1 and put it on the kingside, but there's no real diagonals there either. 

Of course the question is also what to do after we've put all of our pieces on their ideal squares. How are we going to force the action? Well, e4 is a good candidate. If black does take, we can really take aim at that weak e6-pawn and could also take aim at the weak light squares on black's kingside (Bd3 and Qe4 for example). If black doesn't take, he needs to be aware of the e4-e5 push that will unleash Bf3, and that could win the c6 pawn. 

The other plan with Ne5 looks much more logical to me however. If black takes that knight, Bb2 will be unleashed with dxe5! Note that we still have that e4 break in the bag (to use when we have introduced our Bb2 to the action for example). I see that llama has shown a computer tactic for black, but I feel that this is only possible, because white has failed to prepare for the action first (and even here there are other moves than Bd4 and the position just looks real good for white). 

I just wanted to share my thinking in this position. Hope you found it useful, OP!

Yes, it was very helpful. I was able to understand the ideas of every move better. I didn't play e4 in the game because I didn't see a clear way to continue and attack the c6 pawn, but I should have seen that the e6 pawn was very weak. I wasn't really considering Ne5 because I thought that the knight was very strong and it wasn't worth trading for the bishop. I guess I would get an active dark squared bishop from that though. I would rather go for the e4 idea without the Ne5.