Disclosure: This is a paid guest post. The author or an affiliated third party paid for its publication, and the article may contain promotional links. The views and opinions expressed are those of the author and do not necessarily reflect those of GameTyrant or its editorial staff.
Disclaimer: The following content contains references to gambling, casinos, and related topics.
Published odds are the worst number in gaming for predicting what happens next. They were never built for that job. A drop rate describes what an entire player base gets out of millions of rolls, and anyone who has opened a pile of loot boxes without a single legendary has already met the gap between that crowd and their own short run of results.
That gap is the cleanest way into the two figures that come up in almost any discussion of online slots: return to player, nearly always shortened to RTP, and volatility. Both sound like casino jargon. Both describe things gamers handle all the time: a published rate measured over a very long sample and the difference between steady common loot and the rare drop people actually chase.
The gaming side comes first here, since a drop rate and an RTP rest on the same long-run averaging, and the examples are familiar. After that, the focus moves to what the two numbers mean on a licensed slot in Canada, who checks them before a game goes live, and where they're shown.
Blizzard's 2017 Odds Described A Crowd, Not A Player
In May 2017 a Chinese rule took effect that made online game publishers in the country disclose the reward probabilities in their games. Blizzard was among the first to comply. Its Overwatch figures for China, reported by Game Informer on May 5 that year, said each loot box held four items with at least one rare or better, that epics opened on average every 5.5 boxes, and legendaries on average every 13.5.
The words doing the work there are "on average". Nothing in that disclosure promised a legendary in box 13 or box 14. If every box carried the same independent chance of about 1 in 13.5, roughly one player in five would open 20 boxes and get no legendary at all, while someone else pulled two back to back. Both outcomes fit the published rate. It only settles down across the whole population of boxes.
Disclosure spread quickly, and well beyond China. Game Informer had pointed out that Blizzard hadn't said whether its Chinese numbers applied in North America. In December 2017, Apple wrote a new line into its App Store Review Guidelines: any app selling loot boxes, or anything else that hands out randomized virtual items for money, now had to show customers the chance of getting each kind of item before they paid.
Consoles followed. In August 2019 the Entertainment Software Association used a Federal Trade Commission workshop on loot boxes to announce that Sony, Microsoft and Nintendo would require publishers of upcoming games to reveal the chances of earning rare items, with a target of the following year. PC wasn't part of that commitment.
RTP Is A Drop Rate Paid Out In Money
RTP is the same kind of population figure, applied to cash. It's the fraction of total wagers on a slot that the game is set up to return as prizes, counted across every spin it will ever take from every player. A game set at 95% that took C$1,000,000 in total stakes would be designed to pay out about C$950,000 and keep about C$50,000. That retained 5% is the house edge.
There's one difference from a loot table, and it matters for the next section. A drop rate counts items, so one legendary is one legendary. RTP weighs every prize by its size. A single large payout and a thousand small ones can add exactly the same amount to the figure, which means an RTP tells you how much a game returns overall and nothing about how often it pays.
The Overwatch caveat carries straight over, since a few hundred spins are a sliver of a game's lifetime and can end well ahead of or well short of the stated percentage with nothing broken. The figure isn't a rebate either. No individual player is owed any part of it.
In Ontario, that margin is a documented part of every game. The Registrar's Standards for Internet Gaming, which came into force with the province's online market on April 4, 2022, require a game's specifications to document the "odds of winning for each prize available to players" and "the advantage of the operator in relation to each wager."
Volatility Is How Lumpy The Loot Table Is
Two slots can carry the same RTP and behave nothing alike. Volatility describes how that return is spread out: plenty of small payouts, or a handful of large ones with long dry stretches between them.
Gamers know both shapes. Genshin Impact's wish system, as Siliconera laid it out in October 2020, set a 0.6% starting chance for a five-star character against 5.1% for a four-star, promised at least one four-star in every 10-pull and capped the wait for a five-star at 90 pulls. Most pulls are filler. Four-stars arrive at a steady clip, and the item people actually want is rare.
Strip that down to a toy model with a single prize each. Picture one slot built like a common-loot table and another built like a legendary hunt, both set to 94%. The first returns 1.88 times the stake on half of all spins. The second returns 470 times the stake on one spin in 500. Multiply either pair and the answer is 0.94.
Played out, they couldn't look more different. On the common-loot slot, ten straight spins without a win happens about once in a thousand tries. On the legendary hunt, roughly 55% of 300-spin stretches pay nothing at all, and about 13% of stretches still come up empty after 1,000 spins. The long-run cost is identical. The experience on the way there isn't remotely similar.
Actual slots spread their payouts across many prize sizes, from small line wins up to tiered jackpots, so few are as stark as the legendary hunt. The principle holds anyway. A low-volatility game behaves like a dungeon that drops small, common loot on a good share of its clears. A high-volatility one behaves like farming a boss for a mount with a tiny drop chance, where most runs give nothing and a rare result carries the average.
A Licensed Slot Has No Pity Timer
The comparison breaks in one useful place. Genshin's 90-pull cap is a pity system: the game tracks how long a player has gone without a five-star and eventually forces one. Ontario's standards rule that out for slots. Standard 4.24 says that where an outcome is intended to be random, as in dice games or slot games, "the outcome must not be dependent or based upon any history or other factors." Alberta's internet gaming rules repeat that sentence word for word.
On a slot licensed in either province, then, a player's losing streak doesn't make the next win any more likely. Every spin starts fresh, which is why the long-run figures from the previous two sections are the right way to describe the game.
Ontario's Standard 4.08 adds a check before launch: no igaming game or random number generator can go onto a gaming site until it holds either the Registrar's own approval or a certificate from a test lab on the Registrar's register. The Alcohol and Gaming Commission of Ontario registered four such labs in July 2021, ahead of the market's launch: BMM North America, eCOGRA, Gaming Associates Europe and Gaming Laboratories International. It said at the time that it wouldn't test igaming technology in its own lab.
Alberta's equivalent, section 4.12 of the standards issued by AGLC, the province's gaming regulator, lists slot games by name among the products a lab has to certify, and the certificate has to come from an Accredited Testing Facility that AGLC has registered, before the game is deployed anywhere in the province. Ontario's rules keep going after launch, too. Standard 4.09 tells operators to monitor the payback of their live games for behaviour that could point to faulty performance.
Ontario's Standard 4.06, mirrored in the AGLC text, also says that before any bet goes in, a game's rules, help or how-to-play pages must give accurate information on winning odds, payout odds or the return to players. Volatility doesn't appear on that list. It doesn't appear anywhere in either document, so when a volatility rating does show up on an information screen, it's the studio's own description rather than a figure either regulator spells out.
The Figures Sit Behind The Help Screen Of A Licensed Site
For a game licensed in Ontario or Alberta, the help screen is therefore the first place to look, and it's where studios usually print their own RTP as well.
Which sites count as licensed depends on the province. Ontario's regulated market, open since April 4, 2022, has registered private operators running sites through iGaming Ontario. Alberta's opened on July 13, 2026, regulated by AGLC. Elsewhere, provincially run sites fill the role, among them Loto-Québec's online casino and PlayNow, which serves British Columbia, Saskatchewan and Manitoba. All of it is real-money play restricted to adults. Quebec, Manitoba and Alberta set the minimum age at 18, and most of the rest of Canada sets it at 19.
That patchwork is the awkward part, since the Ontario and Alberta rules above bind only the sites registered in those markets. Matching a game to a licensed site in a given province calls for a reference that deals with the provincial split, and the explainer on online slots in Canada from PlayCanada covers how access differs between Ontario and the rest of the country, alongside sections on slot types, the studios behind the games, progressive jackpots and RTP.
Publishing Overwatch's odds in 2017 changed what players in China knew about their loot boxes, not what came out of them. A slot's published RTP and a studio's volatility label work the same way. They describe how a game is built, and no figure on a help screen changes what any single session will pay.
