Ask Mission

It depends on the specific machine, but is often the probabilities that change. Other things that can change are the following:

• Paytable
• Free Games Reels (which would be the probabilities, essentially, only during Free Games which would cause them to return better or worse)
• Progressive Meters-if applicable. (This is the amount of each bet that increases the progressive(s).)

In theory, a slot could be designed to be capable of more or less specificity, but generally speaking, a given slot title is only going to have a handful of RTP settings.

On some occasions, slot games might show an RTP range if the amount that the player is betting is a factor. A player can sometimes be eligible for certain jackpots or features on greater amounts bet than lower ones, or maximum lines played as opposed to fewer than maximum. Otherwise, it could just be that the probabilities of certain events are different with higher bets.

It's the same thing with Video Keno. You might provide a range for all of the possible bet amounts and spots picked for a particular title---but the specific return of the game would depend on how many spots the player is playing and/or what bet amount the player is playing. Denomination can also be a factor with different types of games.

Please note that this post consists of generalizations and is hardly comprehensive.

Good question. I’ll go over the basics for determining if something is a play and how to analyze it:

1.) Determining if There IS a Play:

The first thing is that you have to determine if there is a vulture or progressive play to be had. For variable state quick turnaround vulture games, you’re basically looking for constants and if/thens that you wouldn’t find on a normal slot machine.

(He knew what I meant, but we’ll use Golden Egypt as an example, Wizard has a write-up here:)

https://wizardofodds.com/games/slots/golden-egypt/

The first thing that you notice about that game is that you collect coins as you play, then when you have two coins on any of the reels, that reel turns Wild for two spins.

I don’t just sit down and play every machine in the casino to do this kind of stuff, because to do so would entail a LOT of negative expectation playing as most machines do not have positive states, and those that do are typically not in a positive state. Simply put: I would lose a lot of money doing that.

But, if I notice a new machine and someone playing it, then I’m going to look for those kinds of constants. In the case of something like Golden Egypt, the Rules screen also basically explains how all of that works.

---By the way, Rules screens are often your best friend. If you see a new machine, you can check the rules screen and see if there’s some obvious variable state gimmick that can be exploited for free.

So, I’ll check rules screens and sometimes watch players play to determine if there even should be an advantage somewhere before I do anything else.

Golden Egypt has the ability for the player to make a bet as little as $0.05, I also assumed that this would be the worst-returning version on that unit, (unless they’re all the same) so $100 coin-in at an assumed expected loss of $15 gets me 2,000 spins of data on the coins. I consider that worth it for something where you’re likely to see a lot of plays over time.

So, I basically played and kept track of how frequent the coins were. The best wins come from those occasions where you can get two of the first three reels WILD at the same time, because it’s a guaranteed win. Since pays start at three-of-a-kind on a line (or two Pharaoh symbols) I was able to deduce that the first and second reels (with the third not far behind) are the best ones to have wild in terms of raw value. The fifth reel being wild very rarely does anything for you and I don’t even consider it in deciding whether or not there’s a positive play.

Anyway, a sample size of 2,000 spins (helped because coins are VERY frequent) gave me a pretty high confidence starting point for determining what coin states were positive.

Of course, you’ll sometimes just have to test things by trial-and-error on what you suspect are positive situations. Ocean Magic was one such game because I had no plans on betting 2,000 spins at $0.50/spin on regular mode and $1.00/spin on Bubble Boost straight up to determine if it was positive.

By the time Wizard decided to do a comprehensive analysis based on the full reels, I had already come to some conclusions on the game just by playing it. I’ll link to a few posts in the relevant Wizard of Vegas thread:

https://wizardofvegas.com/forum/gambling/slots/32613-ocean-magic/#post704373

So, prior to the full analysis, there was a fundamental misunderstanding on the value of the bubbles that I cleared up. I also offered some additional insight on how advantage playing the machine works because there is initial raw value and then added value from what can happen on future spins:

https://wizardofvegas.com/forum/gambling/slots/32613-ocean-magic/#post704411

Anyway, read through the thread and you’ll see that the full analysis basically just confirmed what I knew from trial-and-error...if nothing else, I actually played them a little more conservatively than the analysis suggests is optimal, but then the analysis also makes certain assumptions as to free game returns that might not be precise.

It’s a great thread and you should check it out if you’re even considering machine advantage play---required reading. It should be considered a rarity to get mathematicians and multiple advantage players comparing notes in the open to such an extent.

But, let’s get back to my answer:

Anyway, that’s not hard, but what is hard is figuring out how all of this relates back to the overall return of the game. I think there are two ways to do it, though one is going to get you closer to the exact case than the other:

---Watching and learning, you can just look through and make notes (over thousands of total spins) of the actual returns to player of each game state with someone else playing. You could also keep notes from when you try to play particular states (if you suspect an advantage), but that’ll take a long time to get you to a meaningful sample size.

---The harder way is to actually do a mathematical analysis. On most machines, it’ll be possible to look at it on a, “One-Line,” basis, so you can just gather about a thousand spins of data to determine the frequency of each symbol on each individual reel.

NOTE: YES, IT MUST BE EACH INDIVIDUAL REEL!!!! One of the more popular gimmicks that slot designers seem to enjoy right now is to put a disproportionate amount of high paying symbols on reels such as 1/2/4/5 or 1/2/3/5 and then have substantially fewer of those symbols---rarely appearing as, “Stacks,” on one or two of the reels. The effect (if you’re not aware of this tendency) is that casual players will continue to think that they were, “Close,” to hitting a full stack of five of a kind every so many spins.

Symbol frequencies per reel will quickly and easily dispatch this false notion. If you go back to the Ocean Magic thread, for example, you’ll see the reel analysis that demonstrates that all of the high-paying symbols are totally disproportionate. There are also not very many Ocean Magic symbols on Reel 2, which can generate some false notions as to the bubble expansion frequencies.

---Anyway, you can determine the reel frequencies on a, “Single-Line,” basis, which will pretty reasonably give you an idea of base line pay percentage. You just look at the probability of each line pay result, which is easily solved by the symbol frequencies.

The next thing that you’ll want to do is look at Scatter frequencies or results (such as the coin in Golden Egypt) that would impact all lines. Free Games frequencies should also be looked at in the overall sense, unless they are line or left-to-right dependent, which is actually pretty rare.

Left-to-Right dependent (many must-hits have such dependency) aren’t particularly hard to figure out because it’s just the probability of Reel 1, multiplied by the probability of Reel 2, multiplied by the probability of Reel 3...and so on. Line dependent is a little tougher because some lines overlap on two reels and a few on all three reels, sometimes.

Anyway, you’re not likely to get an exact return for the Free Games---even knowing the probabilities---because you won’t be able to take enough of a sample size to full;y analyze the Free Games. For a worst-case scenario, you can just determine the overall return of the base game on a single-line basis and figure that Free Games returns will at least bring it up to the jurisdictional minimum.

As far as the wild reels are concerned, that’s easy, just do the probabilities for line results assuming one or more reels are wild. Time-consuming, but doable.

Of course, if you want to play it safe, then you can just start out with only extremely obvious plays and then slowly become more aggressive with what you’ll try. I went back and forth a bunch on Golden Egypt with any two of the first three reels with one coin, then finally landed on any three of four OR the first two reels.

2.) Progressive plays are pretty simple. You can just watch someone take a few thousand spins and look at actual return. Maybe throw out the best one or two results if it seems like they are extremely unlikely. While you’re doing that, analyze the frequency of whatever event (or symbols) is required for the progressive. Preferably symbols, because the spinning games (and such) do not have to correspond to the probabilities that they appear to be.

Anyway, estimating the probability of one type of scatter symbol (or even a few) is not particularly difficult as long as you are keeping reasonably good notes.

You can also analyze the entire game if you want to, or if Free Games would make that extremely difficult, just make extremely conservative assumptions about the overall return of the game.

Of course, you can also wait for other folks to have done this, but remember information is a commodity in advantage play. If you’re getting tons of it without giving much, then you’ll find yourself not getting tons of it anymore.

3.) Mistakes WILL Be Made

You can’t will an advantage play into existence, so sometimes you’re going to have a hunch about a machine that will be proven wrong.

For that reason, if you have a hypothesis, then try to prove the null hypothesis. Slot machine games are designed to look beatable when they’re not, in some cases, so you should go into analyzing a potential play wanting to be proven wrong. You also shouldn’t be afraid to have your initial thoughts be proven wrong or have it be a blow to your psyche. It’s better to admit that you’re wrong than continue to feed the game money as if you were right.

4.) Observe Other Players

If you’re a vulture who hits enough casinos, eventually you will run into some competition. They’ll be pretty easy to spot because they’ll be doing basically what you’re doing---going to a bunch of machines, looking at them, either on and off quickly or deciding not to play. If you see them eyeing up a machine that you are not going to, then there’s a reasonable chance that there’s an advantage play to be found on that one.

It's a very good idea and there are already a few games (even in USA) that do this. The two examples that come to mind are Super Bonus Keno in the WV parlors where you must make your base bet for each play, but all wins are either doubled or quadrupled. (Forget which)

Another example is a game on some Spielo machines called Triple Ace Poker. In that game, your video poker hands are multiplied...but on at least some variations, you still pay for each of the individual games.

I also agree with those who have said the W2G threshold should be raised. I'd be fine with any increase, (obviously better than none) but 5k seems reasonable enough. It seems like some higher denomination/higher bet games are structured in such a way as to generally avoid W2G...which somewhat defeats the (I assume) purpose of playing such a game.

There may also be a few slot games that do this, but I can't think of any off of the top of my head.

I’m no longer working off of the top of my head and can think of a few slot games. Many single-line slot machines (old school ones) in denominations ranging from quarters to $5 would be set up such that the top jackpot, or second-top jackpot, would pay exactly $1,199. If that’s not trying to circumvent the mandatory tax reporting requirements, then I don’t know what does.

Other games, often with high bets, are often designed such that these sorts of jackpots will be infrequent and would rely on the triggering of additional bonus games, or hitting a screen of all WILD symbols, or equivalent. In short, they happen, but not very often.

My Answer:

First of all, this question was asked in person when Jake pointed out that he thought the Royal Flush progressive on a Mississippi Stud game seemed high, which it was, but only because the casino required a $5 minimum to be bet on it rather than the usual $1 for that particular side bet. My response after finding this out was, “Yeah, it’s high, but only if you like a 72% return to player.”

The answer to Jake’s question is simply that minimum jurisdictional payouts generally only apply to electronic gaming devices such as slots, video poker and video keno. While it’s not true in all jurisdictions, in most, table games side bets can return basically whatever the casino wants them to return. The same is also true for live keno and live bingo barring any laws that exist to the contrary.

Generally speaking, the higher the top possible jackpot relative to the bet amount on any particular situation, the lower the overall return to player will be to compensate for the variance and ability to win huge. State lotteries are a great example of this and is why the return of an instant ticket is generally going to be a higher percentage (but still awful) as opposed to something like PowerBall.

My Answer:

Aside from Free Games Progressives and other types of Progressives (which I don’t even really consider, ‘Vulturing,’ though it is a loosely defined term) the key to discovering machines that can be vultured is just to keep your eyes open and watch people play any new machines that you see.

One component that all vulturing machines share (as I would use the term) is that there is going to be a set of conditions under which a certain thing either ALWAYS happens, or becomes more likely to happen. Wizard has already written about Golden Egypt over on WizardofOdds and, on that game, a reel becomes wild for two spins after two coins have been collected on that reel.

The most obvious positive situation on this game is one in which two of the first three reels (or even more) are already going to be wild for the next spin because such a situation is a guaranteed profitable result. There are a few other machines that share this exact dynamic.

Having identified the situations that are always the case, now we have to determine whether ONLY those situations have value, or whether something close to those situations have value. While Wizard did a full analysis based on the reels, I preferred a more direct approach, which was just to throw a $20 in and play at $0.05 per spin (and use other positive plays) to determine what I believed to be the advantageous times to play. On the Wizard of Odds page.

Anyway, some of Wizard’s configurations are more aggressive than anything I would play on the game. My rule of thumb is pretty simple, “An already wild reel anywhere or any three coins on the first four reels.” I used to play any two of the first three, which Wizard’s chart actually suggests, but my admittedly limited sample size of real world results has led me to conclude that I don’t like any two of three well-enough and instead prefer any three of four. I will, however, sometimes play reels one and two only or the 1-1-0-0-1 configuration.

Another difference is that my method of choosing when to play it disregards the fifth reel completely, which I also do with Ocean Magic. There is a game called Icy Wilds where I do consider the fifth reel, but that’s only because two adjacent wild reels always pay a minimum of 5-FOR-1 on the bet regardless of where they appear, and the same goes for a wild reel adjacent to a full stack of the lady symbol.

Other vulturing type games have certain situations in which a select number of reel stops will automatically become wild. One example of such a machine is the relatively well-known Scarab and there are a couple of others...but those are pretty new sao you’re just going to have to discover them for yourself for maybe two years, or so.

Anyway, it all begins and ends with identifying a slot machine in which certain situations automatically cause a result to come to pass and some are more obvious than others. Plants v. Zombies was an immensely popular machine for advantage players and is now gone from most casinos, but the thrust of that game was that a meter would go up to 500 (which would unlock free games or any three zombie symbols) and people would have different, “Play points,” depending on where the meter was. My, “Play points,” and, “Stop points,” were more conservative than I heard of most other people using, but I guess I’m just a cautious guy. The rub to the machine was that any spins that did not have Brains symbols appear would result in a certain number of points being removed from the meter. The goal to the game was figuring out the frequency of symbols that would add to the meter and the average result of free games.

Most of Plants v. Zombies was intuition and guesswork as many of these vulturing types of plays are. Honestly, as long as an advantage player is not ridiculously aggressive, it really doesn’t make enough of a difference to sit down and fully analyze 1,000-2,000 spins, or what have you, and the return of bonus games can be extremely tough to analyze. Assumptions need to be made sometimes and my conservative assumptions must have been correct enough; because I made an absolute killing on that machine overall.

My Answer:

Percentage-wise, generally you would go with higher denomination slots. I wouldn't jump back and forth on variable state (vulturing) slots probably because there aren't going to be that many playable at a Vegas casino, and also, loading small amounts of free play each into fifty machines doesn't look great.

Penny slots have a higher house edge, but the ones with a ton of lines also tend to be lower variance than are high dollar machines. You could play a dollar denomination at five lines and run 40% on a small amount of free play, easy. The penny machines that have a ton of small wins (over 50% of the return can come from wins of 1x thge bet amount, or less) are going to basically guarantee some return if you have a high preference of getting a reliable actual return.

If you do the pennies, I'd go with a non-themed game (have to pay for licensing for certain themed games, which gets reflected in House Edge) that's not too flashy (great sound and visual costs money, which gets reflected in house edge). I would bet as low as you can tolerate, so hopefully you get as many spins in as possible which is going to stabilize the range of your actual return a bit.

There might be games that are not slots AND are not video poker that you can do. PM me if you want.

My Answer:

I didn’t say much in the thread about this one because this is something of a lengthy and multi-faceted answer.

The first thing that should be mentioned is that the VLT’s (Video Lottery Terminals) found in racinos of New York have a minimum jurisdictional return to player of 90%, which doesn’t seem so hot if you’re used to online casinos, but is one of the highest in the U.S.A. when it comes to land casinos. In most other jurisdictions, slot machines can (and often do) pay less than 90%, particularly penny machines.

One correction that should be made to the question above is that the pool does not need to be fully exhausted in order to be replenished. Otherwise, many of the machines would lock up and refuse to play when it was down to the last result. These pools, often called, “Seeds,” can be added at anytime. Therefore, any thought of waiting for a jackpot to not be paid out over a long enough course of time and therefore making the jackpot more likely is moot. Granted, if we knew how many results were in the pool and had access to every single result that went out it would be possible, but it’s not the case.

In terms of game play, most players wouldn’t know the difference between one set of machines and the other. Most of the machines just look like slot machines regardless of the type of unit. With Class II machines, what is actually happening is the game chooses a result from the results pool (similar to pull tabs) and the slot machine like graphics are used to display the win (or loss) and result...but it’s all just for show. In other cases, it amounts to an electronic bingo game and occasionally there is a little bingo board on the screen.

From a practical standpoint, the difference between the two should only really matter for video poker players because video poker is a game where the player’s decisions affect the expected return of a hand. When it comes to Class II machines, the decisions made by a player do not affect that return as the result from the pool has already decided whether or not the player wins.

The short answer is, it doesn’t make any real difference for slot players, but 90% is a very good minimum guaranteed expected return.

Jennifer Vatter, your cheapest possible drinks would generally come from playing something like Video Poker at the penny denomination (single-handed, $0.05/hand) or video keno at the nickel denomination for a nickel per hand. Most properties that I am aware of you can get free drinks anywhere on the floor, except there are a few higher end properties that you get drink vouchers that print out at the bar based on bet amounts. Most Downtown properties it usually doesn't even matter at the bar, though you will occasionally be asked to make a certain minimum bet amount. This will also sometimes happen on machines on the regular floor.

My advice is Video Poker and the cheapest bets you can get away with.

While this looks great and congratulations to Richard L., I would like to encourage people to remember a few things.

1. With six decks, this event will occur 1 in 62500 attempts.
2. Given a $5 minimum bet, the house edge of this bet, using the paytable I believe I've seen:

Is roughly 97%.

Granted, that's actually a pretty decent return (around 3% house edge) for a side bet, but that is still a much higher house edge than the main game and nearly half of the return comes from this jackpot event. In other words, most players are simply going to lose more than $2.50 for every $5 bet.

I don't know what this resets to, but just know that it is going to be a lot worse of a return at its reset point than the point at which it was hit.

(Specific circumstances will vary, this is just based on one paytable. There are a few paytables which could theoretically be positive with this jackpot amount, though I'd be surprised if they were in use in this case. I'll try to remember to look next time I'm there and will report back.)

Recommendation: DO NOT PLAY IN GENERAL

EDIT: I called and confirmed the paytable and the bet amount. The jackpot would need to be at $112,312.07 at a $5 bet for this to be a breakeven proposition with well over 50% coming from the 100% and 10% jackpot hits combined. In the state when hit, the expected return was just under 97%, which is actually very good for a side bet (in general) and outstanding for one with such a huge potential jackpot. Still not a good bet, though.

ADDED: Just one man’s opinion, but I think it would be nice if casinos were to put the odds of hitting a particular jackpot if they are going to advertise on Facebook that a jackpot has been hit. Never happen.

Answer: I can't speak for UK casinos, but on some games in the states, I would say that information is accurate enough to result in people going broke in a serious hurry. It's all about the probability of the jackpot event v. the value of the jackpot v. the percentage return of all of the other results. (Base Return)

So, it's true for some games, but not others. When it comes to uncapped progressives, I would guess that the, "General Rule," is false more often than it is true.

ADDED: Basically, unless the machine is a must-hit by jackpot, there can be tremendous differences in the probabilities of hitting a jackpot on one machine vs. the probabilities on other machines. Unless you have access to the PAR sheets, or the manufacturer chooses to make the probabilities known, the only way you can get close on the probabilities is through Empirical analysis to determine the frequency of jackpot symbols or events.

Obviously, doing this requires a significant time investment as well as some fairly rudimentary algebra.

It’s also not always worth it. In many cases, all you will discover is that the probability of a jackpot is so remote that the machine almost never reaches positive territory. Alternatively, certain jackpots start at such a low amount, but have a frequent enough probability that the jackpot will generally not miss enough, “Cycles,” to ever go positive.

I would say that, if a person analyzes ten machines, that person may find one with the right combination of meter movement, (how much the jackpot goes up based on the bet) probability of the jackpot hitting and jackpot starting point to find something that can be played with an expected profit, on occasion.