What a review score does not tell you
Good and worth the money are two different questions.
A Steam rating is a good answer to one question: do the people who bought this game like it. That is worth a lot. It is also the beginning of the decision, not the end of it. A 96 percent game can still be a bad buy, and a 78 percent game can be the best twenty dollars you spend this year. The percentage leaves out most of what makes a purchase good or bad.
Price is not in the rating
People review the game they played, not the money they paid. A rating does not move when a game is full price and does not move when it is ninety percent off. Two games at the same 92 percent can be a fair deal and a bad one on the same afternoon, depending entirely on price and length. The rating cannot see that. You have to bring it.
Length is not in the rating
A rating treats a four-hour game and a hundred-hour game the same way. Both can be beloved. They are not the same purchase. What you actually want is price against real length, with the length taken from what people report finishing, not from the store page. A short game can be worth full price. A long game can be a chore you abandon at hour eight. The percentage says nothing about which one you are looking at.
The rating hides how many finish
A high rating with a low finish rate is a specific and common pattern. People buy it, respect it, bounce off it, and still leave a positive review for the part they saw. That is useful to know before you spend, because those unplayed hours are hours you are paying for and may never touch. Steam publishes global achievement rates, which is a rough but real read on how far owners actually get.
Recent versus lifetime
A single lifetime percentage flattens time. A game can sit at "very positive" for its life and "mixed" for the last month after a bad patch, a server shutdown, or a new store full of paid unlocks. The recent line is where a game tells on itself. When lifetime and recent disagree, the recent number is usually the one you care about.
The rating is the first thing I look at and never the last. The BANG Score keeps the review quality as the base, then adds the price, the length, the finish rate, and the monetization that the raw percentage cannot carry. Good is one question. Worth the money is another, and it is the one you are actually asking at checkout.