how to improve

Well, the reason is that you are one of them too.

Lol

Sort of yes, but I think what the op is trying to say is that even they can tell how "badly" their opponents are playing, but they still frustratingly lose. This is perhaps a lot more common than most think it is. However, improving out of that "rut" isn't always so simple. 

First of all @Legend_pr0, that is great if you recognize many opponent mistakes. When you realize where your opponent is playing badly, then you are a portion of the way there to exploiting that. The problem is probably that you aren't capitalizing on these mistakes - or at least not as much as you possibly could be. 

The good news is that you likely already understand why their play is "bad" (all relational anyhow on what is "good" or "bad" chess ratings/ability) whether that be them blundering their pieces, not castling early enough, delaying piece development too much and so on. 

The bad news is that you must be humble enough to realize that sub-700 isn't so bad for you - it is a starting point - not your maximum potential. If you try not to focus too much on rating numbers, then use that energy into learning how to take advantage of common "mistakes." For example, say the opponent doesn't castle early enough. Great for us! We can understand the opponent isn't playing optimally. Typically, we exploit an uncastled King by trapping the enemy King in the center and ripping open lines of attack. 

This isn't necessarily intuitive, or simple to always execute; no worries, no one said chess was easy, it is a process. 

What if the opponent blunders their Queen "badly" a lot? Then we are usually best to exploit this by trading pieces of equal value and simplifying the position. Our material advantage in the endgame (especially as much as +9 from taking a Queen) might be more significant than the same material advantage in the middlegame or opening.

It is up to you to create plans to exploit opponent plans (or even single moves) you suspect might be dubious. It might be upsetting to lose to play you recognize as "bad", but you must look past ego and begin to explore what you can do about it. Until then, there is no reason to believe anything higher than 700 rating is fair. Your chess.com rating probably should be around 700 right now, since it is a process and improving takes time and effort. As metnioned, we aren't saying that you can't improve past 700 (I'm sure you can and maybe even 1000+ fairly quickly!), but merely stating that the current rating might be more accurate than our ego wants us to believe xD Good luck working on that journey though. 

I find that focusing more on learning than increasing rating actually helps increase rating more. If you learn a lot and implement it, then the rating eventually follows your ability. 

Looking at your last loss...you set up a lovely little fork of the rook and king with your knight, but missed the fact that the square you jumped into was defended by opponents knight and was undefended by you. So you just handed your opponent a knight for no reason. Seems like you have enough knowledge about what to do (you talk like you do anyway) but your style of play is very chaotic Try just playing daily games for a while and practice studying the board for threats with a bit of time on your side. I started playing rapid games on here and still have a rating in the 500s as I was struggling with it switched to focus on daily insteaf. Slowly making my way back to 10 min games now and finding it much easier to win games (though of course I've got lots of work to do still on applying what I know against the clock).