In the introduction of Magriel's classic book he writes Backgammon is emphatically not a game of mathematical calculations. In view of all the computer analysis today, I find this hard to believe.

mathematics is about archetypal exactitude. backgammon rests a lot on educated guesswork. the computer is probably also doing guesswork, based on its thousands of rules it has constructed by trial and error over billions of games.

Magriel was essentially correct, although there are a couple of parts of the game where knowing a formula can help (the bearoff, doubling in tournament play).

The backgammon programs help you understand which move is correct in certain situations, although grasping the underlying theory behind the bot's moves is quite difficult.

When the bots do rollouts, they produce reams of numbers, which can look intimidating and don't convey much information. Some players like to cut-and-paste these into their posts, which usually only serves to intimidate the average player and give a misleading impression as to what the game is all about. If I refer to a rollout result in one of my problem explanation, I try to boil it down to the essentials, which (I think) makes problems easier to grasp.

Magriel's book was years ahead of it's time. It inspired me to be a better player. I heard John Clarke say that 90% of the book is still correct so I rolled out chapters 12 to 29 (the beyond basics). I excluded chapter 15 Opening rolls. I used xg2. I found that 72% of the actual positions had the correct answer not 90%. Most of the concepts were correct. It was a worth while use of time because it challenged me to think about why the incorrect ones were incorrect and to think using general principles but to also look beyond those principles.