An upgrader looks like a trade-up but works like a bet. You pick a target item worth more than what you have. The site shows a win chance, usually the value of your stake divided by the value of the target, reduced by a house edge. A provably fair roll decides. Win, and the target item lands in your site balance. Lose, and your stake is gone. There is no middle result.
How does an upgrader calculate the win chance?
If you stake 20 dollars toward a 100 dollar item, a fair chance would be 20 percent. Most sites shave that number down. With a 10 percent edge, the displayed chance becomes 18 percent. Some sites let you slide the chance up and down, which changes the target value rather than the edge. Higher chance, cheaper target. Lower chance, dearer target. The edge stays the same at every point on the slider.
The item prices the site uses matter a lot here. The target's value is set by the site. If a knife is listed at 100 dollars on the site but sells for 85 dollars on a marketplace, your real chance of winning 85 dollars of value is priced as if it were 100. The effective edge is larger than the one on screen.
Is an upgrader better than opening cases?
It is a different shape of the same bet. A case spreads outcomes across many items, most of them cheap. An upgrader is binary: one item or nothing. The expected value in both cases is the stake times one minus the edge, so a 10 dollar stake at a 10 percent edge is worth 9 dollars on average whether it goes into a case or an upgrader. The upgrader just concentrates the loss into a single moment.
People chain upgrades, staking each win toward a bigger target. Each step applies the edge again, so a five-step chain at a 10 percent edge keeps only about 59 percent of the starting value in expectation, before the low odds of actually surviving all five. Provably fair rolls, which you can check with the provably fair verifier, prove the roll was not moved after your bet. They do not improve the odds. See the house edge guide and responsible gaming before playing.