BarSoapSoup
Yeah, it seems to me like you're just being impatient. You are picking a lot of aggressive options in neutral and approaching while you have the lead. Then when you are behind you are panicking a lot and compulsively dodging or shielding.
Just looking at the start of the first match, you run in with your shield up and get grabbed for it, there wasn't really much reason to do a run-up into shield there. You were only behind by 8 percent and the match just started, so you could have taken your time and waited for a better moment to do something. Then when you did manage to get the lead, you still played aggro, so all your friend has to do is wait and play his punish game, which is exactly what he did and is largely why he is winning.
You also want to be more aware of the situation, for instance when you're on your last stock and Dorf is offstage, that's not a good time to charge upsmash at the ledge, your friend almost ganoncided you for that one. However, I did like the earlier dunk attempt on his side b. Which brings up my next point.
It looks like, due to frustration perhaps, that your punish game was a little off, and you were also either making a lot of mis-inputs which you did mention was a problem, or just going for super hard reads. In the case of inputs, those clean themselves up over time as long as you continue to practice, as far as reads go, the rule of thumb for a read is that if you do it and it works, you need to know exactly why and how it worked. If you can do that, then you can duplicate the results in subsequent matches, otherwise you most likely just got lucky and should focus your energy more on other strategies. For instance, (
And just for fair warning this is going to be the start of a massive tangent.) let's look at the classic hard read, reverse forward smash. How does it work? Your opponent rolls behind you and you throw a hitbox where they will end up when their i-frames are over. Why does it work? Because you put yourself in a situation where you know your opponent wants to use that roll to end up behind you, and you know that's a situation they will roll in because you noticed them roll once in that situation, then intentionally found ways to get back into that situation throughout the match.
For instance, maybe when you jab your opponent's shield, you happen to go for the full jab combo but they manage to roll between jab 1 and 2 to get behind you and grab you for it, so you might think "Alright, so my opponent chose to roll when I hit their shield." and you save that information until you are back in a situation where your opponent is shielding again. Then you would remember this information and think "Alright, last time I jabbed shield my opponent rolled, let's see if they are willing to do it again." so this time you might just do a jab 1 then wait and see what happens and if they roll again then maybe it's a habit and you can save that information for later. Now later in the match, you're both at high percent last stock, your opponent decides to shield, so you go for jab 1 to reverse fsmash and they roll into it.
This is actually a very very simplified explanation but is about the gist of how hard reads work, you gather information on your opponent, test that information, then later use the information if it's valid or perhaps drop it if it's not. There are also certain "universal" situations where you can make reads without having to gather much information, or any at all. In Smash one such situation is your opponent landing on a platform, most players will shield when they are stuck on a platform with you nearby underneath them, it's just basic player psychology. So as Bowser, you can easily land a side b or get a down b shield break in this situation with little to no information needed on that specific player. Of course, your opponent is also aware of how that works and might choose other options instead but that's when you'll just start gathering information again and figuring out what you can do in that situation to catch that player if they happen to be avoiding the side-b or down b. Though even then, side b is still a pretty safe option regardless, if you land on the platform you will just auto-cancel and if they chose to roll or spotdodge you can likely still punish them for it depending on how it was timed, if they jumped then they are now above you and have to land again and you're probably still safe. Also, even when a player knows better, it can be difficult for them not to fall victim to deeply ingrained habits (like shielding on the platform) especially since it'll probably work at least a few times since you're not going just going to go for down b or side b each and every time. (
End of tangent.)
Anyway, lag could have also been a factor, so I don't want to call you out too hard on any flubs.
Now towards the end of the matches, it looked like you just psyched yourself out super hard, your friend positioned himself pretty well in these situations and was able to exert a lot of pressure just by standing near you, this would trigger panic reactions, then he would punish them to close out the game. It helps to be wary of when you are compulsively picking options so you can remain lucid in situations like this and pick better options. Also, if you find yourself getting flustered or frustrated, take a breath, take time to hang out on the respawn platform for a bit and calm yourself down. Take a moment on the character select screen between games, but do not simply rush back in while you're still on tilt, you won't be doing yourself any favors this way.
Now as a little extra, while this didn't exactly cause you to lose or anything, it's something I think will help immensely. I notice you are not using Bowser's run skid, it's a useful option that is relevant in a lot of match-ups. Especially on the stronger members of the cast. It has pretty much all the same uses as extended dash dancing except it's easier to do and you have more options out of it. It's definitely worth practicing. The inputs are simply run, stop, run, stop, make sure you go into a full run first. At first, you can use the audible and visual ques from Bowser's footstep to time your stop, this will cause the screen to shake as you run back and forth, but eventually as you get the timing down you'll be able to run back and forth without shaking the screen at all. I recommend practicing this on battlefield since you can use the floating platforms to practice by running back and forth on them without falling off.