CasinoCityTimes.com

Gurus
News
Newsletter
Author Home Author Archives Author Books Search Articles Subscribe
Stay informed with the
NEW Casino City Times newsletter!
author's picture
 

Ask the Slot Expert: Do the machines in New York's racinos really pay back 90%?

22 December 2021

Question: I gamble at local racinos in upstate New York. I have read that the LTR machines are supposed to pay back anywhere from 90 to 94% of credits played. But I have been gambling for many years at these locations and the LTR machines definitely do not pay back anywhere near 90 to 94% of credit played on the LTR machines.

First of all what is time frame all the LTR machines are suppose to pay back 90 to 94% of credits played at the racinos in New York State? Is it a month, 6 months, a year, 5 years, 10 years, 100 years???

I have watched and played the LTR machines for many years now in the racinos in New York and I have spoken to many regulars who play them all the time and I have won small jackpots on them too, but they are definitely not paying back at 90 to 94% of credits played!! More like they are all paying back overall about 33% of credits played. Most people lose and only very few people win or hit jackpots and this has been going on for many years at the racinos in New York State.

That is why I’m asking what is the time frame that all the LTR machines are supposed to pay back at 90 to 94% of credits played at the racinos in New York State??

Also when they say the LTR machines are supposed to payback at 90 to 94% of credits played do they mean each machine or do they mean each racino location as a whole???

John, I wait for your answers because I know firsthand these LTR machines are definitely not paying back at 90 to 94% of credits played and that is from experience over many years of being at the locations, playing the machines, observing people play the machines, and talking to many regulars who play the machines all the time. Overall the machines are way, way, too tight and percentage of payback is way, way too low.

Also another question I have can the main terminal computer that controls all the LTR machines in Schenectady, New York by the New York State Lottery read the amount of each bet on the machines and because of the amount of each bet determine a better payout or a less better payout??

Because I have played and watched many LTR machines that give you a option of betting a dollar all the way up to five dollars and it seems they all pay out better in the bonus rounds on the higher bets of three, four, or five dollars, and they are definitely much more possible to hit the jackpots on the higher bets as well???

John, finally, I know in Las Vegas the gaming commission cracked down on the machines that show almost jackpots all the time because it keeps people playing the machines because they think the machines are about to hit the jackpot, which is not true at all, but the New York State racinos and the LTR machines do that all the time!!

Do not other states have the same gaming laws that stop them from having machines that show almost jackpots all the time to keep people playing the machines???

John, again I will wait for all your answers. Thank You!!

Answer: Let's work our way down through your questions.

First, time on a slot machine is not measured in ticks of the clock but in games played.

Second, in the United States, there is no timeframe associated with long-term payback on an RNG-based machine. Through the magic of Random Sampling With Replacement, the more a machine gets played, the closer its actual payback will tend to get to its long-term payback (the payback calculated from the probability of hitting each winning combination and the amount it pays).

We can calculate the range we expect a machine's actual payback to fall in after a certain number of plays. In the long term, we expect the actual payback to fall in a very narrow range around the long-term payback.

How many plays qualifies as the long term? It depends on how volatile the machine is, the more volatile, the more plays it takes. Nevertheless, there is nothing in the programming of the machine to affect what lands on the payline based on how much the machine has paid out in the past.

That's in the US. Some jurisdictions, like the United Kingdom, allow a machine to alter its probabilities if it has paid out too much or too little.

On an RNG-based machine in the US, though, every winning combination displayed in the paytable is possible on every spin and the probability of hitting it is the same on every spin.

But you're not playing RNG-based machines at New York's racinos. As you wrote, these VLT (Video Lottery Terminal) machines depend on a central server to determine their outcomes. This system works like lottery scratch-off tickets. It is sometimes referred to as Finite Outcome.

The central server maintains many prize pools. All of the machines in a particular prize pool have the same paytable. Each entry in the pool is like a scratch-off ticket. The value can be nothing or one of the payouts on the paytable.

When you press the Spin button on the machine, it sends a request to the central server for an outcome. The central server selects one of the entries in the prize pool for that machine, sends a message to the machine with the value, and removes that entry from the prize pool. The machine gets the message with the amount won on the play and displays a combination that pays that amount.

This is Random Sampling Without Replacement. The probability of getting a winning combination changes with each play. If all of the jackpot entries in the prize pool have been awarded, you have no chance to win the jackpot until the prize pool is exhausted and the central server creates a new prize pool.

Calculating the payback for a prize pool is easy. Add up all of the values of the entries and divide by the product of the number of entries in the pool times the amount bet. Because the entries in the prize pool are not put back after they are selected, the payback of the remaining entries in the prize pool changes with each play.

When the prize pool is exhausted, all of the machines that participated in that pool will have collectively paid back the payback percentage of the complete prize pool.

There's an interesting corollary between looking at results on an RNG-based machines and on a finite-outcome machine. When you look at your results on an RNG-based machine, even if you've played it many times, you're seeing only a very small slice of the total play on the machine. Your results may be good or bad, but if you could see the totality of the play on the machine, you'd see that the payback was very close to the machine's long-term payback.

On a finite-outcome machine, even if you could track the play on the machine from the start of a prize pool to the end, you'd still see only a small slice of all of the play using that prize pool. Your results may be good or bad, but the sum of all of the play in the prize pool across all of the machines that participated would be the payback of the prize pool.

Unlike on an RNG-based machine, which may take hundreds of thousands or even millions of spins for its payback to zero in on its long-term payback, a prize pool achieves its payback the moment all of the entries in it have been used. All of the machines that participate in the prize pool should also pay back the prize pool's payback percentage in the long run.

How long is the long run? I don't know. It depends on the number of entries and the distribution of values in the prize pool and the lottery commission does not publish that information.

To sum up, each time a, say, 90% prize pool is exhausted, all of the machines that participated will have collectively paid back 90% of the money played on them. The actual paybacks on the machines that participated will vary for that instantiation of the prize pool, but over time as they participate in more and more prize pools, their overall actual payback will get closer and closer to 90%.

Moving on, you probably don't have anywhere near enough plays on a machine to make any statistically valid guesses about its payback. If you and the other people you've talked to have not kept track of the amount of action you've given and the amount you've won or lost, you really don't know what payback you've experienced.

If you put $100 in a machine and get $1000 worth of spins before you run out of credits, you've experienced a 90% payback. But what are you going to tell your friends when they ask how you did?

(You bet a total of $1000. Where did it come from? You put only $100 in the machine, so you must have won the other $900. Divide the amount paid by the machine over the action to get payback percentage.)

I've also found players' recollections to be unreliable (selective memory) unless they keep written records of their play.

When you change your bet, you are probably also switching prize pools. Higher bets usually have higher paybacks, so you could very well have a higher long-term payback on the higher bet.

The 90% minimum applies to each machine. After millions of plays, each 90% machine will have paid back very close to 90% of the money played in it.

To be more precise, in the long run each bet on each machine will pay back a percentage very close to the payback percentage for its prize pool.

If the law says that machines are supposed to pay back 90% to 94%, I have no doubt that each prize pool pays back from 90% to 94% and, thus, the long-term payback on each bet on each machine is between 90% and 94%.

Finally, near misses. In the 1980s, the slot machines from Universal Co. Ltd. used a unique way to determine where to stop their reels. The program in these machines would first determine if the outcome of a spin was going to be a win or a loss. If it's going to be a loss, the program would select a combination from a table of losing combinations. This table made an exciting combination like Jackpot-Jackpot-blank appear more frequently than it should have based on the reel layouts. These combinations gave players the impression that the Jackpot symbol was more likely to land on the payline in a winning combination that it actually was.

In 1989 the Nevada Gaming Commission outlawed this "secondary decision" and said that slots had to choose stops at random from a virtual reel layout and that combination had to be displayed without alteration.

That ruling, of course, did not eliminate near misses. Some near misses are always going to occur. It's always going to be possible to have Jackpot-Jackpot-blank or have the Jackpot symbol land above or below the payline.

On-the-payline near misses are actually a good thing. The only way to have the Jackpot symbol land on the payline frequently is to have it likely to land on the payline.

Near-the-payline near misses are a different matter. If a Jackpot symbol frequently lands above the payline, that tells us that stop right under it is likely to land on the payline. It says nothing about how likely the Jackpot symbol itself is to land on the payline. I think the rule in Nevada is that the symbols surrounding a high-paying symbol can be no more than 6 times more likely to land on the payline than the high-paying symbol.

Near-the-payline near misses aren't as much of an issue on today's video slot machines as they were on traditional reel-spinning slots. When there are 40 or 80 paylines, who can tell when a symbol is just above or below a payline?

How about scatter pay symbols? If they appear frequently, that's good. They're likely to appear. There are no paylines for scatter pays. If you need a bonus symbol on reels 2-3-4 and the symbol appears frequently on reels 2 and 3 and less frequently on reel 4, there's nothing deceptive here. The symbol is more likely to appear on reels 2 and 3 and less likely to appear on reel 4.

This is for RNG-based machines. Most (maybe all) states based their RNG-based machine regulations on those in Nevada and New Jersey. I think it's safe to say that all states prohibit a secondary decision, but not all states have a regulation on how frequently the symbols above and below a high-paying symbol can appear relative to the frequency of the high-paying symbol itself.

New York State deals with near misses in its rules for casino gaming [NYCRR Title 9, Subtitle T: Chapter IV (Division of Gaming): Subchapter B (Casino Gaming): 9 NYCRR § 5319.34 Game selection process] but the document for video lottery gaming [NYCRR Title 9, Subtitle T: Chapter IV (Division of Gaming): Subchapter A (Video Lottery Gaming)] doesn't address near misses -- or how the games operate.

I don't know what limitations there may be on your Video Lottery Terminals, but I do know that state lotteries can get away with things that a commercial casino can't. In New Jersey, the lottery slogan was "All it takes is a dollar and a dream" and I doubt the gaming commission would ever have allowed a casino to put the same slogan over a Megabucks machine.

Finally, slots are negative expectation games. It's really not surprising that most of the people leave the casino with lighter wallets than when they entered and that we lose more often than we win on our casino visits.


Here are the latest Covid data. There is a difference in the hospitalization data for US versus NV. The CDC reports total number of Covid hospital admissions. Nevada reports current hospitalizations, not admissions.

All data comes from the CDC (https://www.cdc.gov/covid-data-tracker/#cases_totalcases, https://covid.cdc.gov/covid-data-tracker/#new-hospital-admissions), except for Nevada current hospitalizations (https://nvhealthresponse.nv.gov/).

I added hospitalizations two months ago because the number of cases doesn't tell the whole story now. The vaccinated may get infected, but they are far less likely to require hospitalization or die than the unvaccinated. The deltas in the number of cases shows us how much the virus is spreading, but it's the hospitalizations that strain our healthcare system. To put it bluntly, it really doesn't matter if someone feels like crap for a few days.

Totals Weekly Changes
US NV US NV
Date Cases  Hosp. Adm.  Deaths  Cases  Curr. Hosp.  Deaths  Cases  Hosp. Adm.  Deaths  Cases  Curr. Hosp.  Deaths 
 12/21   51,115,304   3,569,570   805,112   485,401   700   8,311   1,063,296   55,734   9,102   5,639   9   110 
 12/14   50,052,008   3,513,836   796,010   479,762   691   8,201   853,262   57,885   8,946   5,008   17   91 
 12/07   49,198,746   3,455,951   787,064   474,754   674   8,110   821,215   66,191   8,575   5,011   23   125 
 11/30   48,377,531   3,389,760   778,489   469,743   651   7,985   563,746   28,426   6,309   3,343   (50)   55 
 11/23   47,813,785   3,361,334   772,180   466,400   701   7,930   667,924   41,145   10,754   4,784   (24)   86 
 11/16   47,145,861   3,320,189   761,426   461,616   725   7,844   604,748   63,164   7,862   4,577   22   85 
 11/09   46,541,113   3,257,025   753,564   457,039   703   7,759   516,764   33,219   8,290   5,539   34   103 
 11/02   46,024,349   3,223,806   745,274   451,500   669   7,656   555,915   38,028   9,226   4,327   45   109 
 10/26   45,468,434   3,185,778   736,048   447,173   624   7,547   488,829   42,265   9,842   4,753   (87)   96 
 10/19   44,979,605   3,143,513   726,206   442,420   711   7,451   578,396   11,963   4,381   136 
 10/12   44,401,209   714,243   438,039   7,315   795,586   14,067   15,710   149 
 10/05   43,605,623   700,176   422,329   7,166   554,194   10,642   3,852   121 
 09/28   43,051,429   689,534   418,477   7,045   817,218   14,463   6,160   165 
 09/21   42,234,211   675,071   412,317   6,880   971,637   14,691   7,466   152 
 09/14   41,262,574   660,380   404,851   6,728   1,176,763   12,919   10,256   145 
 09/07   40,085,811   647,461   394,595   6,583   975,725   10,076   5,237   104 
 08/31   39,110,086   637,385   389,358   6,479   1,113,414   9,385   7,592   173 
 08/24   37,996,672   628,000   381,649   6,306   1,045,491   7,507   8,117   116 
 08/17   36,951,181   620,493   373,649   6,190   959,978   4,715   7,065   150 
 08/10   35,991,203   615,778   366,584   6,040   819,524   3,987   7,652   122 
 08/03   35,171,679   611,791   358,932   5,918   622,832   2,279   7,489   81 
 07/27   34,548,847   609,012   351,443   5,837   652,251   2,394   8,347   79 
 07/20   33,896,296   606,618   341,096   5,758   169,933   1,478   3,351   28 
 07/13   33,726,363   605,140   339,745   5,730   181,047   1,959   4,982   33 
 07/06   33,545,316   603,181   334,763   5,697   75,104   1,373   2,234   27 
 06/29   33,470,212   601,808   332,529   5,670   87,507   2,057   3,020   24 
 06/22   33,382,705   599,751   329,509   5,646   75,420   2,157   1,930   22 
 06/15   33,303,285   597,594   327,579   5,624   95,797   2,293   1,560   17 
 06/08   33,207,488   595,301   326,019   5,607   114,250   3,762   2,271   21 
 06/01   33,039,238   591,539   323,748   5,586   123,333   3,709   991   27 
 05/25   32,969,905   587,830   322,757   5,559   174,125   4,234   1,676   26 
 05/18   32,795,780   583,596   321,081   5,533   223,966   4,230   2,301   27 
 05/11   32,571,814   579,366   318,780   5,506   303,856   4,687   2,541   33 
 05/04   32,267,958   574,679   316,239   5,473   343,348   4,908   2,559   40 
 04/27   31,924,610   569,771   313,680   5,433   383,163   4,958   2,747   65 
 04/20   31,541,447   564,813   310,933   5,368   464,556   5,072   2,590   36 
 04/13   31,076,891   559,741   308,343   5,332   480,061   5,321   2,986   57 
 04/06   30,596,830   554,420   305,357   5,275   448,935   7,124   3,084   38 
 03/30   30,147,895   547,296   302,273   5,237   439,510   6,793   939   63 
 03/23   29,708,385   540,503   301,334   5,174   388,928   7,446   1,863   53 
 03/16   29,319,457   533,057   299,471   5,121   381,695   8,362   3,078   81 
 03/09   28,937,762   524,695   296,393   5,040   480,902   11,573   2,413   83 
 03/02   28,456,860   513,122   293,980   4,957   463,356   14,129   2,835   75 
 02/23   27,993,504   498,993   291,145   4,882   451,083   13,923   2,406   162 
 02/16   27,542,421   485,070   288,739   4,720   602,906   21,411   4,149   198 
 02/09   26,939,515   463,659   284,590   4,522   779,305   21,828   5,444   244 
 02/02   26,160,210   441,831   279,146   4,278   1,007,777   22,004   7,249   249 
 01/26   25,152,433   419,827   271,897   4,029   1,312,565   23,385   10,324   250 
 01/19   23,839,868   396,442   261,573   3,779   1,317,119   21,318   11,324   279 
 01/12   22,522,749   375,124   250,249   3,500   1,790,345   22,660   17,217   294 
 01/05   20,732,404   352,464   233,032   3,206   1,499,561   18,435   14,655   233 
 12/29   19,232,843   334,029   218,377   2,973   1,258,540   15,460   12,493   186 
 12/22   17,974,303   318,569   205,884   2,787   1,656,411   18,537   16,472   239 
 12/15   16,317,892   300,032   189,412   2,548   1,494,763   17,247   18,825   229 
 12/08   14,823,129   282,785   170,587   2,319   1,375,502   15,483   18,418   175 
 12/01   13,447,627   267,302   152,169   2,144   1,114,175   10,286   15,942   121 
 11/24   12,333,452   257,016   136,227   2,023   1,197,199   10,784   14,130   106 
 11/17   11,136,253   246,232   122,097   1,917   1,099,790   8,501   11,115   65 
 11/10   10,036,463   237,731   110,982   1,852   767,645   6,838   8,868   68 
 11/03   9,268,818   230,893   102,114   1,784   588,207   5,809   5,936   35 
 10/27   8,680,611   225,084   96,178   1,749   492,026   5,585   5,238   (10) 
 10/20   8,188,585   219,499   90,940   1,759   401,037   5,053   4,501   48 
 10/13   7,787,548   214,446   86,439   1,711   351,270   4,886   3,910   48 
John Robison

John Robison is an expert on slot machines and how to play them. John is a slot and video poker columnist and has written for many of gaming’s leading publications. He holds a master's degree in computer science from the prestigious Stevens Institute of Technology.

You may hear John give his slot and video poker tips live on The Good Times Show, hosted by Rudi Schiffer and Mike Schiffer, which is broadcast from Memphis on KXIQ 1180AM Friday afternoon from from 2PM to 5PM Central Time. John is on the show from 4:30 to 5. You can listen to archives of the show on the web anytime.

Books by John Robison:

The Slot Expert's Guide to Playing Slots
John Robison
John Robison is an expert on slot machines and how to play them. John is a slot and video poker columnist and has written for many of gaming’s leading publications. He holds a master's degree in computer science from the prestigious Stevens Institute of Technology.

You may hear John give his slot and video poker tips live on The Good Times Show, hosted by Rudi Schiffer and Mike Schiffer, which is broadcast from Memphis on KXIQ 1180AM Friday afternoon from from 2PM to 5PM Central Time. John is on the show from 4:30 to 5. You can listen to archives of the show on the web anytime.

Books by John Robison:

The Slot Expert's Guide to Playing Slots