Predicting a Haywire - sort of

[poppo]

This has probably been mentioned already, but I just noticed something about Haywire based games. Under certain conditions you can predict when a Haywire is going to happen and even how many times it will.

I have my S+ game set for a credit limit of 300 (and wins going to credit). Any wins that will put it over that limit will pay from the hopper. So I had 205 credits. I got triple single bars which is 30. But the game started paying out of the hopper. This told me that it was getting ready to Haywire, and it would do it at least 3 times since a double Haywire would have been 90 total and still under the hopper limit.

I know - useless info.   But I guess it does prove that the number of Haywires has already been determined when the reels first stop.

[Buzz]

I wouldn't bet the farm on this statement, but I think a S 2000 doesn't pay out anything untill all the Haywire is finished.

[stayouttadabunker]

Actually, that's what's kinda cool about S+ Haywires...
esp. when the the credits are going to the hopper.
It pays out then Haywires, then pays out, then Haywires, etc.etc.

It's a lotta of fun to have the payouts go to the tray instead of the meter display...

[knagl]

Very interesting -- I would have never thought of that.

But I guess it does prove that the number of Haywires has already been determined when the reels first stop.

Yep.  I'm pretty sure it shows that on the PAR sheet, too.

[StatFreak]

Very interesting -- I would have never thought of that.

But I guess it does prove that the number of Haywires has already been determined when the reels first stop.

Yep.  I'm pretty sure it shows that on the PAR sheet, too.

Poppo, if you're talking about the S2000 implementation, then I defer, as I don't know how it is set up, but if you're talking about the S+. that is not exactly true.

If you mean that the RNG has picked the outcome before the spin, then of course that's correct, but if you mean that the first reel determines the Haywire, then it's not correct. If you mean that the RNG picks some mysterious extra number that determines the Haywire aside from the three virtual stops, then that's also incorrect.

<ALERT> If you want to preserve the mystery and fun of the Haywire and Haywire Deluxe, DON'T READ THIS... 
.
.
.
.
.
.
The Haywire payouts are standard combinations of three symbols -- just like any other payout on any machine. The only difference is that they are not advertised. For example, the S+ Haywire Deluxe has two types of blanks, two types of single bars, two types of double bars, and two types of triple bars. The difference between them is not physically displayed. The PAR sheet lists one type of each as "normal" and the other type of each as "special". Digitally, they are different symbols. They could just as well have had 1B, 2B, 3B, 4B, 5B 6B, white blank, and blue blank printed on the strips.

It's not as simple as saying that you will get a 2x haywire if you get three "special" double bars, but very specific combinations of "special" and/or "regular" symbols comprise each Haywire, and if you get the exact same symbols in the same order, you will ALWAYS get the same exact Haywire payout. Think of Haywire pays as payouts with dynamic, undisplayed multiplier symbols. Since they are not displayed, specific symbols don't have to always multiply every payout by the same amount, or at all, but the same symbol outcome always pays the same, each and every time.

For example, on the Haywire Deluxe SS6413, if you get a Haywire symbol on reel 1, a special blank on reel 2, and anything on reel three (except another Haywire symbol), you will get a 2x payout of "Any One Haywire" (It will Haywire once).

If you get a Haywire symbol on reel 1, a "special" blank on reel 2, and a "special" double bar on reel three, you will get the elusive 10x pay for a "Any One Haywire" (It will Haywire 9 times). You will also get this 10x single Haywire pay if you hit a "special" triple bar on reel 1, the Haywire on reel 2, and a red seven on reel 3. Those are the only two combinations that pay the 2 coin 10x payout. Remember that there are "regular" versions of the blanks, double and triple bars, so you won't get the 10x Haywire every time you see "blank / HW / 2B" in the window -- only when the blank and 2B are the "special" versions.


One thing to note on the S+ is that they didn't assign both "special" and "normal" virtual stops to any one physical stop. In other words, the blank on line 9 in the PAR sheet (below the 1B and above the 2B) is always a "special" blank, and never a "normal" blank. So, if you were to mark each of the physical positions on the strips as "normal" and "special" (or memorize their positions), and you had a copy of the PAR in front of you to look up which specific combinations Haywired and how many times, there would be no surprises.

[poppo]

[Poppo, if you're talking about the S2000 implementation, then I defer, as I don't know how it is set up, but if you're talking about the S+. that is not exactly true.

If you mean that the RNG has picked the outcome before the spin, then of course that's correct, but if you mean that the first reel determines the Haywire, then it's not correct. If you mean that the RNG picks some mysterious extra number that determines the Haywire aside from the three virtual stops, then that's also incorrect.

I was talking about S+

What I meant was that after the reels stop, the number of Haywires is already determined at that point. I concluded this because it started paying out of the hopper when it would take at least a win of 96 coins to make it do that. Since I landed on a 30 coin win, it would have gone to credit had it been a normal win. If it was only going to haywire once, it would have been a 60 win and also gone to credit. Two haywires would have been 90 coins and still gone to credit. But since it started paying out of the hopper immediately, I knew it was going to haywire at least 3 times or more.

I guess what I thought before I noticed this was that when it haywires, that another random number may be generated to determine if it will haywire again when it stops each time.

Per your explanation of the PAR, that pretty much verifies this. I just found it a little interesting to know ahead of time that it was going to do it without knowing that I had hit the special 3 haywire combination.

[stayouttadabunker]

Bad Bad Stat 
I never knew that - now you've ruined my favorite game!   
I will do my best to forget this entire thread forever....lol

[Buzz]

David when you put the alert up I quit reading, guess I want to stay dumb on some things.

The other day I installed a SB001000 base chip and a SS4079 ( Haywire ) in a S 2000. The machine preformed exactly like the Double Diamond/Haywire (SB100156) machine I just got. The only thing I can't do is change the pay %. I think the machine with the SS chip would Haywire more often. With the SS chip you still get the reels shaking and the sounds with the backward reel spins.

My next experiment might be to place the two machines side by side and play 100 games on each and log the number of Haywires on each.

[stayouttadabunker]

Although I have absolutely no facts on this - I'd have to agree that
it seemed like my S+ used to Haywire way more than the S2000.

Of course, with no evidence to back it up - take it with a grain of salt...

[StatFreak]

David when you put the alert up I quit reading, guess I want to stay dumb on some things.

The other day I installed a SB001000 base chip and a SS4079 ( Haywire ) in a S 2000. The machine preformed exactly like the Double Diamond/Haywire (SB100156) machine I just got. The only thing I can't do is change the pay %. I think the machine with the SS chip would Haywire more often. With the SS chip you still get the reels shaking and the sounds with the backward reel spins.

My next experiment might be to place the two machines side by side and play 100 games on each and log the number of Haywires on each.

What that tells us is that the code to shake the reels and spin them backwards, and the sound files, are either located on the SG chip, or are included in the generic SB001000 base chip as part of its overall program to drive the older SS chips, but I don't know enough about the S2000 to say which. It's possible that different parts of the code are in each.

Regarding your proposed experiment:

*  First, 100 spins isn't a large enough sample to prove anything; it's essentially useless.

*  Second, a Double Diamond Haywire won't haywire as much as a regular Haywire because there have to be fewer payouts to
    compensate for the 2x multiplier of the jackpot symbol, regardless of whether it's an S2000 or an S+. Comparing a regular
    Haywire to a Double Jackpot/Diamond Haywire is comparing apples to oranges.

*  Third, the best way to determine whether the S2000 or S+ version of the same game Haywires more frequently (pick either Double
    Diamond Haywire or Haywire) is to look at the PAR sheets for an S+ version and an S2000 version that have payout percentages
    as close as possible to one another. This will be much more accurate that any small sample test, even if the test involves several
    thousand spins.


Before I joined NLG, I walked through the virtual stops of all of my kits and made my own PAR sheets. I was able to do it with every game I had, including nudge games the respin game "4th of July". Every game, that is, except my Haywire Deluxe. Since I couldn't figure out exactly how many Haywires there were for each payout and since they weren't displayed in the virtual stop test or the recent spin data, I printed up a spreadsheet, taped it to the award glass, and ticked off every single haywire that occurred (by specific win and number of Haywires).

After several thousand spins, I was finally able to say with some degree of certainty how often the more common Haywires would occur, but even after 10,000 spins I still had too small a sample to be sure about the more rare Haywires. I still have the tick sheets in my slot stand.


I was not able to figure out how the Haywires were determined and their exact odds until I joined NLG and got a PAR sheet from CactusJack.    Actually, my first request on NLG was for PAR sheets for the kits I owned, although it sat unanswered at first. I slowly got most of them as the members got to know me, but the HD was one of the last ones I got, and it was the one that I REALLY wanted more than any other.

[StatFreak]

Okay, the results are in for the Double Jackpot/Diamond Haywire!

...and the winner is... ... ... envelope please  ... ...


The S-PLUS !!


And it's not even close.
Apparently, when they ported the game to the S2000, they increased the hit frequency and reduced the number of multiplier (haywire) payouts.


Here are my "spot" checks of two payout percentages, based on the SB100156 Double Jackpot Haywire for the S2000 and the stated SS chips for the S-Plus:
(all of these are 72 stop paytables, so the total number of combinations is the same for both platforms: 373,248)

Comparing 98% payback pay tables

S-Plus
SS7684 has a payback of 98.024%, a hit frequency of 0.12343, 46,070 winning combinations, and 72 stops.

It has 8,633 haywire pays.

1 in 8.10 pulls is a winner.
1 in 43.24 pulls will haywire.
1 in 5.34 winning combinations are haywires.

S2000
SB100156, LS014295 has a payback of 98.022%, a hit frequency of 0.14185, 52,944 winning combinations, and 72 stops.

It has 6,520 haywire pays.

1 in 7.05 pulls is a winner.
1 in 57.25 pulls will haywire.
1 in 8.12 winning combinations are haywires.


Comparing 95% payback pay tables

S-Plus
SS4264 has a payback of 95.021%, a hit frequency of 0.12079, 45,084 winning combinations, and 72 stops.

It has 9,234 haywire pays.

1 in 8.28 pulls is a winner.
1 in 40.42 pulls will haywire.
1 in 4.88 winning combinations are haywires.


S2000
SB100156, LS014298 has a payback of 94.998%, a hit frequency of 0.13968, 52,134 winning combinations, and 72 stops.

It has 5,844 haywire pays.

1 in 7.16 pulls is a winner.
1 in 63.87 pulls will haywire.
1 in 8.92 winning combinations are haywires.

[StatFreak]

If someone can tell me the SB number for the regular Haywire (and if I have the PAR sheets), I will compare the S+ and the S2000 versions of those as well.

[brichter]

Hey Stat, shouldn't you be sleeping right about now? 

[StatFreak]

Hey Stat, shouldn't you be sleeping right about now? 

That's funny, I was going to ask you the same question on the other thread.

[Ron (r273)]

If someone can tell me the SB number for the regular Haywire (and if I have the PAR sheets), I will compare the S+ and the S2000 versions of those as well.

SB100873 Haywire. I don't have a PAR sheet.

Ron (r273)

[StatFreak]

If someone can tell me the SB number for the regular Haywire (and if I have the PAR sheets), I will compare the S+ and the S2000 versions of those as well.

SB100873 Haywire. I don't have a PAR sheet.

Ron (r273)

That appears to be a Wild Star RWB. It may have a Haywire feature, but it doesn't specifically say, and it doesn't look like something one would find on the S+.

<ADD> I looked in one sheet that did not describe the type 76 feature, but I then found another PAR that described it. This is a 3-reel game with a 4^th reel haywire bonus feature. It is not a standard 3-reel haywire and cannot really be compared to an S+, particularly because the HW feature is a Buy-A-Pay (with the second coin). Since the player is buying the feature, the odds will be based only on the value of the second coin purchase.

Interestingly, not only does it use a fourth reel, it uses both a fourth and fifth random number to determine if there will be a haywire and how many times it will happen. The one PAR sheet that I looked at mapped the fourth reel with a 40% probability of getting a Haywire on any win, and then picked another number from 1-100 to determine the number of Haywire pays, ranging from one to five.

<ADD 2> The haywire feature uses the same virtual tables for every pay table, from 85% to 98%, so the odds of hitting a haywire remain at 40% for every game, and the odds of getting any specific number of haywires also remains the same. The overall value of the haywire feature only changes because the number of winning combinations on the first three reels changes.

The haywire feature adds an overall 1.08 times each original winning combination, and since the player is paying 1 extra credit for the feature, one can see that it just slightly increases the overall payback. The average overall value of a haywire is 2.7 times the base pay of any win mitigated by the 0.4 probability of getting the haywire in the first place.

It's really a simple, elegant implementation that adds entertainment value to the game. Kudos to the IGT designers.

[Ron (r273)]

Thanks Stat! I will change my list.

IGT SB100858 DOUBLE DIAMOND ST SG000251 U21 27C4002 2928
IGT SB100859 9/8/04 5X GLD-TRIPL E STR SG000251 U21 27C4002 1HAC
IGT SB100860 10/8/4 5X GLD-TRIPL E STR SG000251 U21 27C4002 FP30
IGT SB100865 THREE TIMES PAY SG251 U21 27040 7135
IGT SB100866 THREE TIMES PAY SG251 U21 27040 CF4H
IGT SB100867 TRIPLE TEXAS STARS SG000312 U21 27C4002 SG0 A218
IGT SB100868 TRIPLE TEXAS STARS SG000312 U21 27C4002 SG0 37HP
IGT SB100869 SILVER STRIKE DBD SG254 U21 27040 C448
IGT SB100871 DIAMOND WHITE ICE SG000251 U21 27C4002 UFH5
IGT SB100872 DIAMOND WHITE ICE SG000251 U21 27C4002 F36P
IGT SB100873 HAYWIRE SG251 U21 27040 UU3A
IGT SB100875 FIFTY TIMES PAY SG000363 U21 27C4002 PHUF
IGT SB100876 FIFTY TIMES PAY SG000363 U21 27C4002 82U6
IGT SB100877 DOUBLE CLASSIC 7S SG000251 U21` 27C4002 H210
IGT SB100878 DOUBLE CLASSIC 7 SG251 U21 27040 8F8P
IGT SB100879 BIG TIMES R,W&B SG251 U21 27040 1H34
IGT SB100883 BITZ & PIZZAS SG000261 U21 27C4002 274P
IGT SB100886 RUN FOR YOUR MONEY SG000269 U21 27C4002 F15C
IGT SB100887 RUN FOR YOUR MONEY SG000269 U21 27C4002 HUH2
IGT SB100888 LIZ TAYLOR DAZZLI SG000254 U21 27C4002 77F4
IGT SB100889 TRPL DBL WHITE ICE SG000363 U21 27C4002 12UU
IGT SB100890 TRPL DBL WHITE ICE SG000363 U21 27C4002 UAH9

Ron (r273)

(edit) Now I know why my Triple Diamond Haywire only hits small amounts.

[StatFreak]

Ron, it is a Haywire, but it's a 4-reel bonus Wild Star RWB with a Haywire bonus on the fourth reel.


Stat

P.S. They could have put 2x, 3x, 4x, 5x and 6x multipliers on the fourth reel and weighted them accordingly (similar to Big Times Pay) and it would come out the same, but the added animation and suspense of the haywire function makes the game more entertaining.

[PLUNGER BOY]

So what your saying is BIG TIME PAY is just HAYWIRE without the drama   (same game)

[StatFreak]

So what your saying is BIG TIME PAY is just HAYWIRE without the drama   (same game)

No, because Big Times Pay has multipliers that range from 2x to 100x, and they only apply when one or two jackpot symbols are part of the payout. The Haywire bonus reel spins on every win and is not associated with the jackpot symbol (although it will spin when you hit the jackpot). <ADD> Besides, this is a Red, White, and Blue variant.

What I was saying was that they could have replaced the haywire symbols with the specific multipliers of 2x, 3x, 4x, 5x, and 6x and weighted them with the same probability as that of each of the haywires and the game would have paid out exactly the same. But instead of having a bonus reel with those multipliers clearly listed, they made all of the symbols generic HW symbols and hid the selected multiplier from the player. That way, when the random outcome selected maps to a 3x pay, instead of displaying a 3x and paying triple, the reel displays a generic HW symbol and then the winning combination haywires two times.

In essence, this is a very different game than the S+ 3-reel Haywire (S2000 still TBD). The S+ Haywire works completely differently by having very specific combinations of hidden symbols on the main reels that pay specific amounts, some of which are then displayed to the player as haywires.

[Ron (r273)]

Talk about confusion.        My Bonus Spin Wild Star RWB, Vision (SB000022) sounds
like the same. It has the 4th reel.


IGT SB000019 MYSTERY MUTIBLI * 27040 HH00
IGT SB000020 MYSTERY MUTIBLI * 27040 H6H3
IGT SB000021 B SPIN W STAR RWB * U21 27C4402 8P4F
IGT SB000022 B SPIN W STAR RWB * U21 27C4002 4U84
IGT SB000025 BALLOON BARS * U21 274002 2430
IGT SB000026 BALLOON BARS * U21 274002 8F92
IGT SB000029 PURE PLEASURE B W * U21 27040 4H19
IGT SB000030 PURE PLEASURE B W * U21 27040 94F6


Ron (r273)

[StatFreak]

SB000022 only initiates the bonus on natural bar pays (1B, 2B, and 3B) with or without the DD and TD multipliers.
It doesn't spin on mixed bars, cherries, 7B or red 7, with or without multipliers.

While it appears (to the player) to have a range of pays for each winning combination, the weighting of the pays is such that the long-term payout for each is exactly what would normally be paid for the wins without the bonus feature.

For example, 1B 1B 1B has an average pay of 10 coins (which is typical for many slot games), but is paid out the wheel bonus through the following range of seven payouts: 4, 6, 8, 10, 15, 20, or 30 coins. The pay values of any natural 1B payout with wild multipliers is increased proportionately.

The 2B 2B 2B has an average pay of 20 coins and uses the bonus payouts of 10, 12, 16, 20, 30, 40, and 60 coins.

The 3B 3B 3B averages 30 coins per pay with payouts of 15, 20, 25, 30, 40, 60, and 90 coins.

So you've got me curious as to how it looks to the player.   What is on the bonus reel?

If I've missed something obvious forgive me; I'm very tired, and just off to bed... 

Stat