Thread Rating:

100xOdds
100xOdds
  • Threads: 638
  • Posts: 4256
Joined: Feb 5, 2012
March 14th, 2016 at 11:32:03 AM permalink
So my local casino killed 9/6 JoB and 9/6 ddb (all single line).

best is now $1 8/6 JoB (98.4%) triple play.
after 10 sessions ($1000 coin-in each), I have yet to have a winning session.
yes, I know, variance.

How much more variance is Triple play over single line?
Craps is paradise (Pair of dice). Lets hear it for the SpeedCount Mathletes :)
Romes
Romes
  • Threads: 29
  • Posts: 5600
Joined: Jul 22, 2014
March 14th, 2016 at 11:42:39 AM permalink
Quote: 100xOdds

...How much more variance is Triple play over single line?

My immediate gut response would be, well, the same or even less?

You're playing 8/6 JoB, just 3 times as fast. So, technically, you're getting a lot more hands a lot faster which should lessen the effects of variance I would think? Short term (like a 2 hour session) is essentially a 6 hour session. It shouldn't change the actual math variance behind JoB as each hand is still drawing from independent decks with an RNG and playing according to the JoB rule set.

The more hands you get, the larger your EV will be (positive at a +EV game and negative at a -EV game) but your cone of variance and standard deviations should get smaller and smaller in relation. I would think like blackjack, when you get to X hands, variance has effectively been negated and you'll be "around" EV within a standard deviation or so.
Playing it correctly means you've already won.
100xOdds
100xOdds
  • Threads: 638
  • Posts: 4256
Joined: Feb 5, 2012
March 14th, 2016 at 11:58:38 AM permalink
Quote: Romes

My immediate gut response would be, well, the same or even less?

You're playing 8/6 JoB, just 3 times as fast. So, technically, you're getting a lot more hands a lot faster which should lessen the effects of variance I would think? Short term (like a 2 hour session) is essentially a 6 hour session. It shouldn't change the actual math variance behind JoB as each hand is still drawing from independent decks with an RNG and playing according to the JoB rule set.

The more hands you get, the larger your EV will be (positive at a +EV game and negative at a -EV game) but your cone of variance and standard deviations should get smaller and smaller in relation. I would think like blackjack, when you get to X hands, variance has effectively been negated and you'll be "around" EV within a standard deviation or so.


less variance?! wow..
$10k total coin-in and I'm waaaaaaaaaaaaayyyyyyyyyyy down compared to the expected -EV.
(10k x 1.4% should be -$140. I'm down 4 figures.)

there are 5 of these 8/6 JoB triple play machines and I've played each twice.
stupid variance.

almost makes me want to give up my VP addiction. Almost...
Craps is paradise (Pair of dice). Lets hear it for the SpeedCount Mathletes :)
ukaserex
ukaserex
  • Threads: 21
  • Posts: 262
Joined: Jul 12, 2015
March 14th, 2016 at 12:16:02 PM permalink
Romes gut instinct is correct.

The 8/6 JoB, whether one line or multiple is a losing play. The longer you play, the more you lose, theoretically.

As to variance - the three lines is normalized a bit. Should be less variance. As to how much less, that would involve a calculator. (At least, that's what I got out of reading Dancer's book this past weekend.)
"Those who have no idea what they are doing, genuinely have no idea that they don't know what they are doing." - John Cleese
MathExtremist
MathExtremist
  • Threads: 88
  • Posts: 6526
Joined: Aug 31, 2010
March 14th, 2016 at 12:31:32 PM permalink
https://wizardofodds.com/games/video-poker/appendix/3/

The new WoO site is harder to navigate but search works fine. I got there by searching on "triple play variance".
"In my own case, when it seemed to me after a long illness that death was close at hand, I found no little solace in playing constantly at dice." -- Girolamo Cardano, 1563
Romes
Romes
  • Threads: 29
  • Posts: 5600
Joined: Jul 22, 2014
March 14th, 2016 at 12:37:50 PM permalink
Quote: 100xOdds

less variance?! wow..
$10k total coin-in and I'm waaaaaaaaaaaaayyyyyyyyyyy down compared to the expected -EV.
(10k x 1.4% should be -$140. I'm down 4 figures.)

there are 5 of these 8/6 JoB triple play machines and I've played each twice.
stupid variance.

almost makes me want to give up my VP addiction. Almost...

The total variance for 9/6 JoB is 19.514676 (from the wiz). Variance is the Standard Deviation squared... so SD = SQRT(19.514676) = 4.4175.

So let's see...

Average Bet (assuming you're playing $1 level, 5 coin) = $5
Avg Advantage = -1.4% (-.014)
OriginalSD = 4.4175 * AvgBet = 4.4175 * 5 = 22.09

SD(x hands) = SQRT(x) * OriginalSD

So while your EV is -140 on 2000 hands, your SD(2000 hands) = SQRT(2000) * 22.09 = 987.90

This is just ONE standard deviation, which comes with 68% confidence. To say with 95% confidence where your results "should" fall, that would be 2 SD, which would be 1975.79.

Thus, after your 2,000 hands and $10k coin in, you would Expect to lose $140, PLUS or MINUS $1,975.79.... Oh, don't we all love variance =P.

Hands Expected Value Standard Deviation
1
-$0.07
+-22.09
100
-$7.00
+-220.90
1,000
-$70.00
+-698.55
2,000
-$140.00
+-987.90
10,000
-$700
+-2,209.00
100,000
-$7,000
+-6,958.47
1,000,000
-$70,000
+-22,090.00


I hope you can see here that with 99% confidence when you reach 1 million hands I can guarantee you'll be negative. You might be only down $5-$10k instead of EV which is -$70k, or you might be down $130k, but you will be down.. As you go on to more hands, even with 3SD and 99% confidence, you can't win =(.


...become the AP you were born to be!
Playing it correctly means you've already won.
RS
RS
  • Threads: 62
  • Posts: 8626
Joined: Feb 11, 2014
March 14th, 2016 at 1:33:26 PM permalink
If it's same total bet (i.e: $5 single play verse $1 5-play), then single play has higher variance. If it's the same denom, then multi-play will have higher variance (i.e.: $1 single play [$5/pull] vs $1 triple-play [$15/pull]).

So yeah, you should be experiencing more variance on 3-play than 1-play.
100xOdds
100xOdds
  • Threads: 638
  • Posts: 4256
Joined: Feb 5, 2012
March 14th, 2016 at 1:44:45 PM permalink
Quote: MathExtremist

https://wizardofodds.com/games/video-poker/appendix/3/

The new WoO site is harder to navigate but search works fine. I got there by searching on "triple play variance".



ahh.. so 3 play is MORE variance. thx!

9/6 Jacks or Better
# lines/ Variance/ Standard Dev.

1play 19.514676 4.417542
3play 23.447459 4.842258
Craps is paradise (Pair of dice). Lets hear it for the SpeedCount Mathletes :)
Ibeatyouraces
Ibeatyouraces
  • Threads: 68
  • Posts: 11933
Joined: Jan 12, 2010
March 14th, 2016 at 2:11:35 PM permalink
Saturday, my first 25¢ 9/6 JoB session, I lost $200 with only one 4oak. Today in less than 75 hands, I got 3 of them. This game can be very swingy at times.
DUHHIIIIIIIII HEARD THAT!
Hunterhill
Hunterhill
  • Threads: 53
  • Posts: 2145
Joined: Aug 1, 2011
March 14th, 2016 at 2:48:38 PM permalink
Quote: 100xOdds

ahh.. so 3 play is MORE variance. thx!

9/6 Jacks or Better
# lines/ Variance/ Standard Dev.

1play 19.514676 4.417542
3play 23.447459 4.842258


But if I understand correctly you're betting 3x as much per spin, so of course you're going to have bigger swings .
The mountain is tall but grass grows on top of the mountain.
DRich
DRich
  • Threads: 86
  • Posts: 11597
Joined: Jul 6, 2012
March 14th, 2016 at 3:26:18 PM permalink
I like this page on Jazbo's site for multi-hand poker.

http://jazbo.com/videopoker/nplay.html
At my age, a "Life In Prison" sentence is not much of a deterrent.
mustangsally
mustangsally
  • Threads: 25
  • Posts: 2463
Joined: Mar 29, 2011
March 16th, 2016 at 3:42:26 PM permalink
Quote: 100xOdds

after 10 sessions ($1000 coin-in each), I have yet to have a winning session.

$1000 coin in each session only?
$15 per round?
is 67 rounds at $1005 coin in
that is it?

VPforWinners shows 70% is the chance for one session of that length to show a net loss
so 1/0.70^10=35.4
and you are that 1 - lucky you!
to show a loss after 670 rounds is abouts 66% (witout comps)
I Heart Vi Hart
  • Jump to: