Tweet about a $200 MH and lost money

I prefer posts / tweets / articles where people lost money rather than screenshots or photos of big wins. I learn more from these posts than some random person getting lucky and won big.

Here is the tweet: https://mobile.twitter.com/davecinc/...63257819443201

There are two photos: (before) $200 MH at $197.05 and (what appears to be the after) $200 MH gets taken down, drops, cracks, etc at $199.49 meaning the meter moved $2.44.

Person writes (paraphrasing here) lost a couple hundred due to crappy bonus rounds.

When you approach these MH’s (also called MHB’s), you need to know at least two things: RTP & meter movement. Generally, the larger of the two MH is harder to solve due to the Minor jackpot, the meter movement for the Minor Jackpot, etc.

However, we can simplify the analysis by assuming the Minor won’t get hit so we now have one equation with two unknowns: meter movement for Major jackpot & RTP (or 1 less house edge).

When I get a chance, I’ll show the the three main math approaches to these MH’s: (a) WoO’s simple formula, (b) the B/E method that uses the most likely jackpot, and (c) the village idiot approach of using mid-point.

You know (c) is a stupid approach because this specific MH is decidedly NON-RANDOM, and we know this due to survivorship bias.

So we will make the following assumptions:
(a) RTP is 90% (this means house edge is 10%)
(b) meter movement for Major jackpot is 10 basis points
(c) a method to discern most likely jackpot amount when it breaks

... math analysis to follow.

WoO formula can be found here: https://wizardofodds.com/games/slots...pot-ainsworth/

“Math formula is
The general formula for the Target Point can be given as:

t = m × (h + r) / (h + 2r), where:

t = Target Point
m = Max jackpot
r = Rise of meter rate
h = House edge of game (taking into consideration the average value of the progressive)”

So based on my limited understanding of WoO formula (which btw ignores the Minor jackpot), we get the following:
$200 * (0.101 / 0.102) = roughly $198.04 so the person played TOO EARLY based on this formula.

Had the RTP being 88% (which is more likely based on my sources for this type of game), then house edge is 12% such that the Target would be 200 * (.121 / .122) or roughly $198.36 (which makes sense since the house edge is higher so you need a higher target).

There are huge methodological problems with WoO’s approach that is not the case with option math. But these formulae are fast and easy to use.

The village idiot approach is using a mid-point of $197.04 & $200 or assuming a target of a $198.52 or ($197.04 + $200) * 0.5. You can read more about this approach here:
https://www.advantageslots.com/must-...u-should-play/

You might as well read animal entrails or tea leaves if you are thinking of this stupid, moronic mid-point method.

Folks, we have survivorship bias at work here. The Major jackpot is at $197 for a reason: it didn’t drop at $196, $195, $194, $193, etc. SURVIVORSHIP BIAS ... this is a non-random data point.

Next post is the B/E method.

In summary: all these methods uses assumptions but the key assumption is the jackpot amount that is most likely to break. WoO & mid-point are two simplistic but fast methods.

If you want to read about survivorship bias: https://en.m.wikipedia.org/wiki/Survivorship_bias

Nice article about how survivorship bias help the US during WWII: https://www.technewsiit.com/hidden-h...vivorship-bias

Abraham Wald is a genius & I still use his Run’s Test in casino’s today.

In the MH’s, you DO NOT SEE THE JACKPOTS THAT GOT HIT (EARLY), you only see the ones that have not gotten hit.

B/E analysis or math is simply the Break Even jackpot where you neither make money nor lose money (on average).

We know from statistics, e.g. a normal curve, that there is a most likely event and there is a distribution amongst the most likely event. Under a normal curve, the most likely event is the mean (also the mode & the median). When we see an unhit jackpot, there is a most likely jackpot and a distribution amongst that most likely jackpot.

So B/E is the most likely jackpot that results in a break even outcome. You only need algebra to solve this based on the aforementioned assumptions listed in this thread.

Supposed the most likely jackpot was $199.49 and we knew this ahead of time. So the question is: what must be the unhit jackpot (assumes the Minor jackpot is not hit) to result in a B/E situation?

199.49 - [(199.49 / 0.10) * 0.001] should be the B/E answer.

So B/E is roughly 197.50.

$199.49 / 10% house edge results in $1,994.90 coin in (on average).

$1,994.90 of coin in times 10 basis point meter movement moves the meter $1.9949 or $1.99 rounded or floored.

$199.49 less $1.99 in meter movement gets us to $197.50

So the person played at $197.05 or $0.45 below the B/E point assuming RTP of 90% & 10 basis point meter movement & Minor jackpot not being awarded.

That $0.45 shortfall implies a $45 expected loss vs a couple hundred dollars in losses. So the larger loss vs expectations could be a larger house edge, variance, slow meter movement, etc.

$0.45 meter movements implies $450 of coin in and 10% house edge results in expected loss if $45.

This post will quickly explain why the mid-point approach is MORONIC and should not be used (btw, WoO formula is easier & faster to use WITH LESS CALCULATIONS!!).

Case #1 assume the RTP changes from 90% to 88%
WoO formula will make adjustments for this.
B/E math also makes adjustments for this.
Mid-point approach as I understand it does not make any adjustments.

Case #2 assume the Meter movement changes from 10 bp to 5 bp
WoO formula will make adjustments for this.
B/E math also makes adjustments for this.
Mid-point approach as I understand it does not make any adjustments.

Neither a change in the RTP nor meter movement causes the user to use a different value than mid-point per the mid-point approach; the underlying economics will change but the mid-point target does not change.

Suppose there is no Minor jackpot & only Major jackpot
WoO formula (as I understand it because it is a general formula) will give the same answer regardless of the current unhit jackpot amount
The B/E’s approach is a bit more robust in that you need to input the most likely jackpot to be hit GIVEN the current unhit jackpot

In the B/E approach, you need enough data points to build a function that will approximate the most likely jackpot for a given unhit jackpot amount. Now, you see why people are drawn to option math in these MH’s because for each unhit jackpot, what you really care about is the most likely jackpot to be hit.

This post is about attribution analysis or whether the person lost due to variance or other factors.

Here, I will use WoO formula.

Target is 200 * (0.101/0.102) = $198.04 floored.

$198.04 - $197.05 = $0.99 in meter movement (“mm”) too earlier
$0.99 mm = $990 of coin in @10 bp mm
$990 coin in @ 10% house edge = $99 expected loss

In summary, the person should have lost an expected $99 since he played too early.

Suppose the RTP was 88% or 12% house edge.

Target = 200 * (0.121 / 0.122) or $198.36 floored
$198.36 - $197.05 = $1.31 mm too early
$1.31 mm is $1,310 coin in based on 10 bp mm
$1,310 coin in @ 12% house edge is $157.20 in expected losses by playing too early.

One possible explanation to the couple hundred dollars in losses could be a RTP or 88% or 12% house edge. $157.20 in expected loss is pretty close to couple hundred in losses.

I wasn’t there, but these posts make me run my numbers & do my analysis as a sanity check.

Good luck with those MH’s.

Quote:

---

Originally Posted by Ex-AP

WoO formula can be found here: https://wizardofodds.com/games/slots...pot-ainsworth/

“Math formula is
The general formula for the Target Point can be given as:

t = m × (h + r) / (h + 2r), where:

t = Target Point
m = Max jackpot
r = Rise of meter rate
h = House edge of game (taking into consideration the average value of the progressive)”

So based on my limited understanding of WoO formula (which btw ignores the Minor jackpot), we get the following:
$200 * (0.101 / 0.102) = roughly $198.04 so the person played TOO EARLY based on this formula.

Had the RTP being 88% (which is more likely based on my sources for this type of game), then house edge is 12% such that the Target would be 200 * (.121 / .122) or roughly $198.36 (which makes sense since the house edge is higher so you need a higher target).

There are huge methodological problems with WoO’s approach that is not the case with option math. But these formulae are fast and easy to use.

The village idiot approach is using a mid-point of $197.04 & $200 or assuming a target of a $198.52 or ($197.04 + $200) * 0.5. You can read more about this approach here:
https://www.advantageslots.com/must-...u-should-play/

You might as well read animal entrails or tea leaves if you are thinking of this stupid, moronic mid-point method.

Folks, we have survivorship bias at work here. The Major jackpot is at $197 for a reason: it didn’t drop at $196, $195, $194, $193, etc. SURVIVORSHIP BIAS ... this is a non-random data point.

Next post is the B/E method.

In summary: all these methods uses assumptions but the key assumption is the jackpot amount that is most likely to break. WoO & mid-point are two simplistic but fast methods.

---

This is pure hogwash. Never-an-ex-AP, you are more full of shit than a Christmas Turkey. This kind of shit has to be called out because it is simply wrong.

Now, go ahead and fire me up, never-an-ex-ap, but it won't do you any good.

Quote:

---

Originally Posted by Ex-AP

This post is about attribution analysis or whether the person lost due to variance or other factors.

Here, I will use WoO formula.

Target is 200 * (0.101/0.102) = $198.04 floored.

$198.04 - $197.05 = $0.99 in meter movement (“mm”) too earlier
$0.99 mm = $990 of coin in @10 bp mm
$990 coin in @ 10% house edge = $99 expected loss

In summary, the person should have lost an expected $99 since he played too early.

Suppose the RTP was 88% or 12% house edge.

Target = 200 * (0.121 / 0.122) or $198.36 floored
$198.36 - $197.05 = $1.31 mm too early
$1.31 mm is $1,310 coin in based on 10 bp mm
$1,310 coin in @ 12% house edge is $157.20 in expected losses by playing too early.

One possible explanation to the couple hundred dollars in losses could be a RTP or 88% or 12% house edge. $157.20 in expected loss is pretty close to couple hundred in losses.

I wasn’t there, but these posts make me run my numbers & do my analysis as a sanity check.

Good luck with those MH’s.

---

Anyone with any real experience on Ainsworth MHB"s knows the real reason for the two hundred dollar loss is the huge variance of these games. A 1 tenth meter, 197 starting point, puts the expected hit at 198.5. At $1 a spin that would be 1500 spins. It's commonplace to get lower than a 50% return on Ainsworth's in 1500 spins or less because of how top heavy the payscale is. A huge chunk of the payback is in the long shot line pays.

What sort of midpoint approach are you using where you don't factor in meter move or RTP? That wouldn't be a calculation, that's just a guess.

Quote:

---

Originally Posted by Ex-AP

When you approach these MH’s (also called MHB’s), you need to know at least two things: RTP & meter movement.

---

You've written quite a bit about the MHB's. Nowhere do you mention this: 1) payback represented by the Major meter 2) payback represented by the Minor meter. For, besides RTP and meter movement, these other two factors also come into play.

Do you know how to determine the payback the meters add to the game?

Quote:

---

Originally Posted by Ex-AP

This post will quickly explain why the mid-point approach is MORONIC and should not be used (btw, WoO formula is easier & faster to use WITH LESS CALCULATIONS!!).

Case #1 assume the RTP changes from 90% to 88%
WoO formula will make adjustments for this.
B/E math also makes adjustments for this.
Mid-point approach as I understand it does not make any adjustments.

Case #2 assume the Meter movement changes from 10 bp to 5 bp
WoO formula will make adjustments for this.
B/E math also makes adjustments for this.
Mid-point approach as I understand it does not make any adjustments.

Neither a change in the RTP nor meter movement causes the user to use a different value than mid-point per the mid-point approach; the underlying economics will change but the mid-point target does not change.

Suppose there is no Minor jackpot & only Major jackpot
WoO formula (as I understand it because it is a general formula) will give the same answer regardless of the current unhit jackpot amount
The B/E’s approach is a bit more robust in that you need to input the most likely jackpot to be hit GIVEN the current unhit jackpot

In the B/E approach, you need enough data points to build a function that will approximate the most likely jackpot for a given unhit jackpot amount. Now, you see why people are drawn to option math in these MH’s because for each unhit jackpot, what you really care about is the most likely jackpot to be hit.

---

Mid-point most assuredly makes adjustments according to payback and meter movement so both of your cases are wrong.

[QUOTE=mickeycrimm;96719]

Quote:

---

Originally Posted by Ex-AP

This post will quickly explain why the mid-point approach is MORONIC and should not be used (btw, WoO formula is easier & faster to use WITH LESS CALCULATIONS!!).

Case #1 assume the RTP changes from 90% to 88%
WoO formula will make adjustments for this.
B/E math also makes adjustments for this.
Mid-point approach as I understand it does not make any adjustments.

Case #2 assume the Meter movement changes from 10 bp to 5 bp
WoO formula will make adjustments for this.
B/E math also makes adjustments for this.
Mid-point approach as I understand it does not make any adjustments.

Neither a change in the RTP nor meter movement causes the user to use a different value than mid-point per the mid-point approach; the underlying economics will change but the mid-point target does not change.

Suppose there is no Minor jackpot & only Major jackpot
WoO formula (as I understand it because it is a general formula) will give the same answer regardless of the current unhit jackpot amount
The B/E’s approach is a bit more robust in that you need to input the most likely jackpot to be hit GIVEN the current unhit jackpot

In the B/E approach, you need enough data points to build a function that will approximate the most likely jackpot for a given unhit jackpot amount. Now, you see why people are drawn to option math in these MH’s because for each unhit jackpot, what you really care about is the most likely jackpot to be hit.

---

Mid-point most assuredly makes adjustments according to payback and meter movement so both of your cases are wrong.

Quote:

---

Originally Posted by Ex-AP

Suppose there is no Minor jackpot & only Major jackpot. WoO formula (as I understand it because it is a general formula) will give the same answer regardless of the current unhit jackpot amount. The B/E’s approach is a bit more robust in that you need to input the most likely jackpot to be hit GIVEN the current unhit jackpot

In the B/E approach, you need enough data points to build a function that will approximate the most likely jackpot for a given unhit jackpot amount. Now, you see why people are drawn to option math in these MH’s because for each unhit jackpot, what you really care about is the most likely jackpot to be hit.

---

Here is your chance to give a clinic on finding the most likely jackpot amount.

Ainworth Mystery Progressive
93% RTP
Single Meter....starts at 250, maxes at 500.
Meter runs at .2%
The meter currently is at $480.

1. What is the most likely jackpot number and please show your math for determining it.

2. What is the theoretical win/loss on the play? Please show the math.

It's been 3 days since my last post challenging ex-AP to answer the questions I listed. Apparently he is not interested in any back and forth on analyzing these games. Whats up with that, ex-AP? Can't stand the heat or what? You are good at dishing it out but a little bitch punk at taking it.

Quote:

---

Originally Posted by Ex-AP

WoO formula can be found here: https://wizardofodds.com/games/slots...pot-ainsworth/

“Math formula is
The general formula for the Target Point can be given as:

t = m × (h + r) / (h + 2r), where:

t = Target Point
m = Max jackpot
r = Rise of meter rate
h = House edge of game (taking into consideration the average value of the progressive)”

So based on my limited understanding of WoO formula (which btw ignores the Minor jackpot), we get the following:
$200 * (0.101 / 0.102) = roughly $198.04 so the person played TOO EARLY based on this formula.

Had the RTP being 88% (which is more likely based on my sources for this type of game), then house edge is 12% such that the Target would be 200 * (.121 / .122) or roughly $198.36 (which makes sense since the house edge is higher so you need a higher target).

There are huge methodological problems with WoO’s approach that is not the case with option math. But these formulae are fast and easy to use.

The village idiot approach is using a mid-point of $197.04 & $200 or assuming a target of a $198.52 or ($197.04 + $200) * 0.5. You can read more about this approach here:
https://www.advantageslots.com/must-...u-should-play/

You might as well read animal entrails or tea leaves if you are thinking of this stupid, moronic mid-point method.

Folks, we have survivorship bias at work here. The Major jackpot is at $197 for a reason: it didn’t drop at $196, $195, $194, $193, etc. SURVIVORSHIP BIAS ... this is a non-random data point.

Next post is the B/E method.

In summary: all these methods uses assumptions but the key assumption is the jackpot amount that is most likely to break. WoO & mid-point are two simplistic but fast methods.

---

The gist of the above post is if you believe it's random you are the village idiot. But only a village idiot would believe it's not random. ex-Ap is a quack.

Quote:

---

Originally Posted by Ex-AP

B/E analysis or math is simply the Break Even jackpot where you neither make money nor lose money (on average).

We know from statistics, e.g. a normal curve, that there is a most likely event and there is a distribution amongst the most likely event. Under a normal curve, the most likely event is the mean (also the mode & the median). When we see an unhit jackpot, there is a most likely jackpot and a distribution amongst that most likely jackpot.

So B/E is the most likely jackpot that results in a break even outcome. You only need algebra to solve this based on the aforementioned assumptions listed in this thread.

Supposed the most likely jackpot was $199.49 and we knew this ahead of time. So the question is: what must be the unhit jackpot (assumes the Minor jackpot is not hit) to result in a B/E situation?

199.49 - [(199.49 / 0.10) * 0.001] should be the B/E answer.

So B/E is roughly 197.50.

$199.49 / 10% house edge results in $1,994.90 coin in (on average).

$1,994.90 of coin in times 10 basis point meter movement moves the meter $1.9949 or $1.99 rounded or floored.

$199.49 less $1.99 in meter movement gets us to $197.50

So the person played at $197.05 or $0.45 below the B/E point assuming RTP of 90% & 10 basis point meter movement & Minor jackpot not being awarded.

That $0.45 shortfall implies a $45 expected loss vs a couple hundred dollars in losses. So the larger loss vs expectations could be a larger house edge, variance, slow meter movement, etc.

$0.45 meter movements implies $450 of coin in and 10% house edge results in expected loss if $45.

---

In this post the village idiot uses the example of the twitter poster playing a MH at 197.05 when it maxes at 200. That the meter runs at one tenth of a percent, and the prog hit at 199.49. Then the village idiot uses some psycho math to tell us the player should not have played until the prog reached 197.5....because that would be the breakeven number.

What the village idiot hasn't done is show us how to determine that the prog will hit at 199.49. You would need that information to determine what the B/E number (breakeven number) is. Here he has determined that it is 197.5 but he did it after the fact of the prog hitting at 199.49....and using that number to find the B/E number. What good would this do anyone as you can't do the math until after the prog has hit?

Quote:

---

Originally Posted by Ex-AP

This post is about attribution analysis or whether the person lost due to variance or other factors.

Here, I will use WoO formula.

Target is 200 * (0.101/0.102) = $198.04 floored.

$198.04 - $197.05 = $0.99 in meter movement (“mm”) too earlier
$0.99 mm = $990 of coin in @10 bp mm
$990 coin in @ 10% house edge = $99 expected loss

In summary, the person should have lost an expected $99 since he played too early.

Suppose the RTP was 88% or 12% house edge.

Target = 200 * (0.121 / 0.122) or $198.36 floored
$198.36 - $197.05 = $1.31 mm too early
$1.31 mm is $1,310 coin in based on 10 bp mm
$1,310 coin in @ 12% house edge is $157.20 in expected losses by playing too early.

One possible explanation to the couple hundred dollars in losses could be a RTP or 88% or 12% house edge. $157.20 in expected loss is pretty close to couple hundred in losses.

I wasn’t there, but these posts make me run my numbers & do my analysis as a sanity check.

Good luck with those MH’s.

---

I don't know if ex-AP is trying to baffle us with bullshit or is baffling himself with his own bullshit. Who gives a fuck what attribution analysis is. Here, ex-AP is trying to determine if the player lost due to variance. Well, Well, Well.

The player started at 197.05 and it went off at 199.49. It's got a one tenth meter. The meter ran $2.44. Multiplying 2.44 by 1000 tells us the player made a $2444 wager. He lost $400 spinning off the pay but was paid $200 on the end so lost $200. Let's divide the $400 loss by 2444.

400/2444 = 16.4%

ex-AP cited a 12% house edge on this game. The prog meters represent about 5% of the payback so we are looking at about a 17% theoretical drop until you hit the prog. The player got dropped for just 16.4%.

So did the player lose $200 due to variance? No. Let's repeat that. No, he did not lose $200 because of variance.

He lost $200 on this play simply because the prog ran well past the mid-point number, 198.5.
200 + 197.05 = 397.05
397.05/2 means the midpoint number is 198.5

How often will this happen? At 197.05 it was 2.95 from the top. The player put another 2.44 in it.

2.44/2.95 then subtract 1 means when you start at 197.05 it will run to 199.49 or higher 17.3% of the time. Thats about once every six plays. Not very long odds on it happening.

Quote:

---

Originally Posted by mickeycrimm

Quote:

---

Originally Posted by Ex-AP

This post is about attribution analysis or whether the person lost due to variance or other factors.

Here, I will use WoO formula.

Target is 200 * (0.101/0.102) = $198.04 floored.

$198.04 - $197.05 = $0.99 in meter movement (“mm”) too earlier
$0.99 mm = $990 of coin in @10 bp mm
$990 coin in @ 10% house edge = $99 expected loss

In summary, the person should have lost an expected $99 since he played too early.

Suppose the RTP was 88% or 12% house edge.

Target = 200 * (0.121 / 0.122) or $198.36 floored
$198.36 - $197.05 = $1.31 mm too early
$1.31 mm is $1,310 coin in based on 10 bp mm
$1,310 coin in @ 12% house edge is $157.20 in expected losses by playing too early.

One possible explanation to the couple hundred dollars in losses could be a RTP or 88% or 12% house edge. $157.20 in expected loss is pretty close to couple hundred in losses.

I wasn’t there, but these posts make me run my numbers & do my analysis as a sanity check.

Good luck with those MH’s.

---

I don't know if ex-AP is trying to baffle us with bullshit or is baffling himself with his own bullshit. Who gives a fuck what attribution analysis is. Here, ex-AP is trying to determine if the player lost due to variance. Well, Well, Well.

The player started at 197.05 and it went off at 199.49. It's got a one tenth meter. The meter ran $2.44. Multiplying 2.44 by 1000 tells us the player made a $2444 wager. He lost $400 spinning off the pay but was paid $200 on the end so lost $200. Let's divide the $400 loss by 2444.

400/2444 = 16.4%

ex-AP cited a 12% house edge on this game. The prog meters represent about 5% of the payback so we are looking at about a 17% theoretical drop until you hit the prog. The player got dropped for just 16.4%.

So did the player lose $200 due to variance? No. Let's repeat that. No, he did not lose $200 because of variance.

He lost $200 on this play simply because the prog ran well past the mid-point number, 198.5.
200 + 197.05 = 397.05
397.05/2 means the midpoint number is 198.5

How often will this happen? At 197.05 it was 2.95 from the top. The player put another 2.44 in it.

2.44/2.95 then subtract 1 means when you start at 197.05 it will run to 199.49 or higher 17.3% of the time. Thats about once every six plays. Not very long odds on it happening.

---

Ex-Crement also states that it is a quandary (obviously it is not) if someone who is using the mid-point method to determine if a must hit is a play leaves the machine only to then find the progressive meter has been pushed up (but not yet hit) by another player such that the original mid-point calculation would no longer be valid (everyone except Ex-Crement would be thanking the ploppy internally for pushing up the meter, leaving the chair vacant and raising the EV, rather than obsessing that this is some sort of quandary - what an absolute idiot Ex-Crement is). Well then, by his own cretinous logic, it would still be a quandary regardless of whether the mid-point method is used or his own blow-hard option method is used, since, regardless of method, you would still have to do a recalculation of the most likely hit point using the new initial condition.