Thank you for the kind words, and as for your interest in how the data are collected, we are using the Websockets from those casinos and saving every rounds in our databases, hence we are using our own databases which allow use to parse past the whole seeds ( millions of rounds ), the datas are then highly accurate.
While we can't really make it open source, (as it's just a small part of our larger project), the crash part will always remain free to use.
We’re also expanding into slots analysis, where players can view detailed stats about their favorite slots. For instance, did you know "the Dog House" slot has a max win hit frequency of 1:8,333,333,333? It’s surprising that this slot is so popular given these odds, especially when better options exist with higher max multipliers and much better hit frequencies.
As for adding more casinos, we’ll only include those offering crash games with a 1% house edge for now since it takes alot of work to add one. Casinos with higher house edges (2–5%) aren’t really aimed toward serious players but rather those who don’t mind "donating" their money to the casino, since it exists better options to play crash game than playing on a 5% house edge one ... Thanks too for your feedback !
Thanks for your answer edgeGPT, sockets sound as a good way to collect the data, if i think in the process i would say you connect to the same soket as the casino front end and that way you get the information like with zero delay, that's a cool move.
About the implementation of the slot, that would be cool, more than for chasing a big multiplier it would be nice in an educational mode, the fack that a slot takes 8.3B spins to show a max win is a good data and makes me avoid playing that slot.
We actually use the casino Websocket to collect real-time data and then redirect it to our frontend using SSE. This setup allows us to provide live data to users while avoiding the risk of our system getting overloaded with API calls. It’s efficient and ensures a smooth user experience, we had done alot of versions before this one without using the Websocket which weren't good enough.
As for slots, we already offer a database of slot stats here:
https://edgegpt.bot/slots. Players can find detailed and accurate data for most popular slots, which I highly recommend checking out before playing.
While slots are “random” in execution, they’re built on mathematical models, meaning some slots offer better chances to profit or max win than others.
For example, if you’re depositing on a casino and took a Wager x40 (which means you have to wager 40 times your deposit), I’d suggest opting for slots like Sweet Bonanza, which has a high base game return percentage after hitting some big wins (around 50% base game hit, so you'd hit something 1/2 spin while being able to hit a bonus). It’s an excellent choice for wagering, even though max wins are less likely. At the end of the day, it’s all about balancing your preferences—whether it’s chasing a big win or maximizing returns while meeting wager requirements. With this, you'll be able to first play on slots where the 1000x happen more often, and once you hit some big hits, be able to wager "safely".
No wonder my colleague seoincorporation is kind of enthusiastic about the project as I was reading it and saw that it was typical of what he likes.
First of all thank you for exposing this project here and taking so much time in the detailed explanations, we rarely see that here.
In my personal opinion, if you had access to a tool that could improve your odds—even by less than 10%—wouldn’t you use it? For example, with a 2x multiplier, the house edge gives a 49.50% chance to win versus 50.50% to lose. If we could shift this to 50.50% in the player’s favor, wouldn’t it be worth a try? I believe we'll give more than "just" 50,50% in the player's favor, but we'll see !
We deliberately avoid delving into complex statistics or probabilities on the site because that would require explaining concepts like the Markov property, which might confuse users. While crash games may seem random, the median over a large number of games stabilizes around 1.98, meaning historical results do have some bearing on future outcomes since it's how the game actually works. People will probably say that the crash game is like flipping a coin, always 50% chance, that the previous results doesn't matter in the future rounds outcome, but at the end of the day, the median of the game has to remain at around 1,98x over a large number of rounds.
I believe the crash are modulated in frequencies and amplitudes, and that historical results have actually bearing on future results, for example, if a 100x multiplier is supposed to occur once every 100 rounds but hasn’t appeared in 1000 rounds, the game will eventually adjust to bring the average back to normal. This could mean more frequent 100x multipliers in the near future; or for example if you have only like 3-4 100x in 2000 rounds, it means you're short 17-16 of them correct? then the game will eventually "even" itself, and you'll most likely see much more 20 in the next 2000 rounds to have something like 40 time 100x over 4000 rounds.
Very interesting, it is logical that if we take a series of events and it deviates greatly from the average, there comes a time when it will return to the average. If we also rely on the results of a series of upcoming bets (and not on an individual bet) to make a profit it seems we have the odds in our favour. The problem I see is that if this really works, casinos will realise at some point and will modify the algorithm so that the player cannot take advantage of it.
I don't know if you know this story:
How the Pelayos Broke the Roulette Wheel: The Story of Human Endurance and Mathematics.The clan that ruled the roulette wheelIn this context, I have a question: why sell this system instead of using it to make money? Are you afraid that if the casinos realise that you are making money with the system from the same account, they will notice and react?
Hi !
I am the one who should be saying thanks since you're giving your feedback without anything in return, this is kinda like doing a QA contribution work !
To answer your question, we’re not worried about casinos noticing or reacting to this system for a few reasons:
- What we provide isn’t about exploiting the system—it’s about giving players better information to make decisions. It doesn’t hack or manipulate the game in any way, also the crash game is provably fair, they can't just change it either, all they could do at most is simply remove it.
- Even with trends hot & cold nnumbers, outcomes are still more or less random, and a player will most not likely follow exactly the site.The real challenge in gambling isn't just beating the house edge— it's controlling emotions, managing bankrolls, and knowing when to stop! Casinos understand that and know their edge holds in the long run, they’re unlikely to see this as a threat since even thought some players will manage to profit from it, some players will want to bet even more using this. Our site may encourage more activity across casinos, as players will likely gravitate toward games or sites where it’s "hot". If Stake, for instance, is "cold", some players may choose another platform—but for others, it’s the opposite, as emotions and habits play a significant role in gambling behavior.
- Our site is more educational than anything else—it helps players taking decisions, also as i said before, the crash game part is just a small part of our project, kinda like showcasing something bigger.
If casinos ever chose to remove crash games because of this, it would validate our site and build our reputation within this space. For players, this might be profitable if used properly. For us, we’re focused on long-term vision rather than short-term gains
Let me know if there is anything else you'd like to know, i'll try to stay around here and come at least once or twice a day !