I liked this article by Saffron Olive because it points out something that I don't think we acknowledge enough: tiny percentages over time add up.
This is why in the recent discussion at CFB about loot or don't loot was so contentious. Because Magic is such a complicated game, it's pretty easy to brush off the math for how it feels or insist that the math is the only thing that matters.
But: in the situation described, there is a nonzero chance of having to keep your best card but lose your second best card in a situation where you are already winning.
And the opponent is going to have another threat. This is something LSV doesn't address in the objections and I feel like that is key to understanding things.
Because the situation described plays out like this:
How many turns do you have left? If you are able to take the angel, then the game lasts 4 more turns assuming nothing else happens. If anything else happens, then the game continues and you still have the best answer in your deck, along with the opportunity to draw into win conditions.
It's a tiny, tiny percentage but tiny percentages add up. And since I'm already winning, I go conservative.
That doesn't mean it's right. Most of the time, the looting will have a positive impact and if I was in a losing situation then looting is always right. But if I'm winning, then I look for the small percentage that causes me to lose and avoid it.
Because the small percentages add up over time. And as this article by Neil Oliver points out, cumulative growth is the goal.
With this in mind, I need to remember when I make good plays, too. If all I can do is catalog my mistakes, that's going to take a lot of the fun out of this game-hell out of life-for me.