A starting pool and a current pool are different
Hypothetical example: one chase among 100 equally likely cards gives a 1% initial chance. If ten non-chase cards leave without replacement, one chase among the remaining 90 gives about 1.11% for the next draw.
If the chase leaves and isn't replaced, its next-draw probability is zero. Those calculations require uniform selection, complete counts and no replenishment. They aren't verified odds for a named platform.
Inventory counts may not be selection weights
A product might choose a rarity tier first and then choose a card within that tier. In that model, the number of cards shown in each tier does not by itself establish the chance of receiving that tier.
A buyback may re-enter the pool, or new stock may replenish it. Ask how those events affect selection. Don't infer a fixed-pool model just because a site shows a list of remaining cards.
Questions for the displayed odds
Check whether probabilities describe launch inventory or the remaining pool, when they update, whether sold-back cards return, and how cards are selected within each tier.
Save a timestamped odds table before opening. If the seller doesn't publish enough information, describe current odds as unknown instead of repairing the gaps with assumptions.
Don't confuse past misses with a promise
In an independent model with unchanged odds, previous misses don't improve the next attempt. In a finite pool, only actual changes to the eligible contents and selection rules justify a new probability.
Knowing more about the model can help you read an offer. It isn't a reason to increase your spending limit or chase a depleted pool.