viewsource
Newbie member
- Joined
- Sep 16, 2024
- Location
- United Kingdom
Introduction
I was curious to understand more about the game pop-o-gold. The game itself is a ways type game with an RTP of 96% and a max win of 20000x. The game has a bonus buy mechanic for 100X which also provides a wheel gamble which can be used to gamble the free spins from a starting point of five to eight (gamble 1) or twelve (gamble 2). Each gamble phase also has a 'gold' slice which awards 10 spins from gamble 1 and 14 spins on gamble 2.
Things I was curious to understand: -
1.) What are the odds of the gamble wheel?
2.) What are the odds of winning the 'gold' outcome during either gamble?
3.) How does the wheel gamble affect the volatility?
3.) How difficult would it be to max win?
Approach
In order to get an understanding of odds on a slot machine it is necessary to have a reasonable sample size of data, therefore, to gain an insight I was going to need to automate the execution of the game process.
Armed with the high level process flow of the game I set about writing a tool using the python programming language to automate the playing of the game programmatically.
I consider it good fortune that I just so happened to dive into a game which had a fairly basic implementation in terms of how the client communicates with the game server. In particular, the results of the bonus are sent back as a single json document so the process of capturing the data was much more straight forward.
Results
Below is an output of the results which are based on 20,000 bonus buys per gamble strategy i.e. no gamble, gamble once or gamble twice. All results were generated against a demo instance of the game based on 100 eur bonus buy.
The graph below shows how the RTP performed over 1K intervals. As you'd expect the number all head towards the tRTP but it's worth noting that the no-gamble strategy was forever chasing the RTP.
I also note the variance between the various strategies. Gambling for maximum spins gives a lot more volatility which is what you would expect. But in the real world that could be pretty brutal. Equally, losing 36 consecutive gambles on the bonus gamble can and will happen, most likely leaving you questioning your sanity when it does.
The questions
1.) What are the odds of the gamble wheel?
I assumed going into the game that the wheel was roughly 50/50. Looking at those results you can see it's a smidgen off. This could well be standard variance but overall I would say I was happy with the results. Seems fair enough.
2.) What are the odds of winning the 'gold' outcome during either gamble?
Looking at the game code you can see a wheel is made up of 20 slices. 10 dedicated to win, 9 dedicated to lose and 1 dedicated to gold. This would 'suggest' that your chances of hitting a golden slice are 5% ish. You can counter this and say the gold slice is indeed smaller but I didn't bother going down the route of measuring. But looking at the stats - the gold slice odds are pretty rubbish. Additionally, the average return from the gold slices is poor. I would say that overall the gold slices are a red herring and hitting them will not have much of an impact on your overall slot session. It's certainly not the case that hitting 14 = payday.
3.) How does the wheel gamble affect the volatility of the bonus game?
Overall you can say that the results are inline with what you might expect. One anomaly I thought stuck out was the %age of wins above 100x in the basic bonus buy i.e. 5 spins.
4.) How difficult would it be to max win?
I managed to get the first max win on bonus game 3861. At this point I had not implemented the saving of the game result data so I wasn't sure how the max win was triggered. On game 13687 I managed to hit the second max win. This time I had the game data saved which meant I'd need to figure out how to replay it.
But on the no gamble game results and the gamble 1 game results I did not see any max wins.
The interesting part here is if you look at how the max win was achieved it's seemingly achievable on any of the bonus games.
My expectation is that the max win will show up eventually in all game results but more data required for the no-gamble and 1 gamble results.
Replaying the game
This was a bit fiddly but I managed to get there in the end. Below is an animated gif of the game outcome.
As you can see - it really is a one hit wonder. Possibly an anomaly it was seen twice on max gamble but not once on the others. Maybe I'll put some more spins through it to see.
Final thoughts
It was a fun exercise and if I were to play this game again in real play then I'd probably lean towards the single gamble. I didn't poke into the specifics on the individual plays but my gut feel is that the overall results are scripted and/or tightly controlled based on interactive play. Things like never hitting 4oak when 3oak are jam packed and the multi is high etc. The max win also looks suspect to me and dare I say ever so slightly lazy. If I were to win 20000x on a slot I want to go on a bit of journey but you wouldn't exactly complain either way.
Hopefully you enjoyed the read.
I was curious to understand more about the game pop-o-gold. The game itself is a ways type game with an RTP of 96% and a max win of 20000x. The game has a bonus buy mechanic for 100X which also provides a wheel gamble which can be used to gamble the free spins from a starting point of five to eight (gamble 1) or twelve (gamble 2). Each gamble phase also has a 'gold' slice which awards 10 spins from gamble 1 and 14 spins on gamble 2.
Things I was curious to understand: -
1.) What are the odds of the gamble wheel?
2.) What are the odds of winning the 'gold' outcome during either gamble?
3.) How does the wheel gamble affect the volatility?
3.) How difficult would it be to max win?
Approach
In order to get an understanding of odds on a slot machine it is necessary to have a reasonable sample size of data, therefore, to gain an insight I was going to need to automate the execution of the game process.
Armed with the high level process flow of the game I set about writing a tool using the python programming language to automate the playing of the game programmatically.
I consider it good fortune that I just so happened to dive into a game which had a fairly basic implementation in terms of how the client communicates with the game server. In particular, the results of the bonus are sent back as a single json document so the process of capturing the data was much more straight forward.
Results
Below is an output of the results which are based on 20,000 bonus buys per gamble strategy i.e. no gamble, gamble once or gamble twice. All results were generated against a demo instance of the game based on 100 eur bonus buy.
The graph below shows how the RTP performed over 1K intervals. As you'd expect the number all head towards the tRTP but it's worth noting that the no-gamble strategy was forever chasing the RTP.
I also note the variance between the various strategies. Gambling for maximum spins gives a lot more volatility which is what you would expect. But in the real world that could be pretty brutal. Equally, losing 36 consecutive gambles on the bonus gamble can and will happen, most likely leaving you questioning your sanity when it does.
The questions
1.) What are the odds of the gamble wheel?
I assumed going into the game that the wheel was roughly 50/50. Looking at those results you can see it's a smidgen off. This could well be standard variance but overall I would say I was happy with the results. Seems fair enough.
2.) What are the odds of winning the 'gold' outcome during either gamble?
Looking at the game code you can see a wheel is made up of 20 slices. 10 dedicated to win, 9 dedicated to lose and 1 dedicated to gold. This would 'suggest' that your chances of hitting a golden slice are 5% ish. You can counter this and say the gold slice is indeed smaller but I didn't bother going down the route of measuring. But looking at the stats - the gold slice odds are pretty rubbish. Additionally, the average return from the gold slices is poor. I would say that overall the gold slices are a red herring and hitting them will not have much of an impact on your overall slot session. It's certainly not the case that hitting 14 = payday.
3.) How does the wheel gamble affect the volatility of the bonus game?
Overall you can say that the results are inline with what you might expect. One anomaly I thought stuck out was the %age of wins above 100x in the basic bonus buy i.e. 5 spins.
4.) How difficult would it be to max win?
I managed to get the first max win on bonus game 3861. At this point I had not implemented the saving of the game result data so I wasn't sure how the max win was triggered. On game 13687 I managed to hit the second max win. This time I had the game data saved which meant I'd need to figure out how to replay it.
But on the no gamble game results and the gamble 1 game results I did not see any max wins.
The interesting part here is if you look at how the max win was achieved it's seemingly achievable on any of the bonus games.
My expectation is that the max win will show up eventually in all game results but more data required for the no-gamble and 1 gamble results.
Replaying the game
This was a bit fiddly but I managed to get there in the end. Below is an animated gif of the game outcome.
As you can see - it really is a one hit wonder. Possibly an anomaly it was seen twice on max gamble but not once on the others. Maybe I'll put some more spins through it to see.
Final thoughts
It was a fun exercise and if I were to play this game again in real play then I'd probably lean towards the single gamble. I didn't poke into the specifics on the individual plays but my gut feel is that the overall results are scripted and/or tightly controlled based on interactive play. Things like never hitting 4oak when 3oak are jam packed and the multi is high etc. The max win also looks suspect to me and dare I say ever so slightly lazy. If I were to win 20000x on a slot I want to go on a bit of journey but you wouldn't exactly complain either way.
Hopefully you enjoyed the read.
