GTOWizard Acquires Ruse - Poker News - News, Views and Gossip

I'm a big fan of GTO Wizard, but I'm honestly not exactly sure what Ruse is. This announcement is the first I have heard of it, and I can't seem to find a clear explanation of specifically what it is anywhere.

It appears to be some kind of HU AI?

Quote

06-01-2023 , 11:30 PM

MicroDonkYT

adept

Join Date: Jan 2020 Posts: 818

Quote:

Originally Posted by TRUSTtheDRAWCESS

It is a neural net solver, so it runs much faster. You can node lock and re-solve really fast. It is like Deep Solver. This won't kill online poker, but it will kill Pio.

Quote

06-01-2023 , 11:35 PM

Koshko

adept

Join Date: Mar 2021 Posts: 901

Quote:

Originally Posted by MicroDonkYT

It is a neural net solver, so it runs much faster. You can node lock and re-solve really fast. It is like Deep Solver. This won't kill online poker, but it will kill Pio.

We built a neural network solver 3 years ago. Essentially all it is , is Pio/je solver connected to an api that sends requests and solves on the fly.

It’s not a Pio killer at all since effectively Pio is as good as the computer running it.

The whole concept around AI this and that is pretty bs marketing angle. It’s simply a normal I/O solver connected to a fast computer producing the recalculations on the fly.

Quote

06-02-2023 , 01:54 AM

Elizabeth5579

grinder

Join Date: Feb 2021 Posts: 529

prefer live poker where you dont have to worry about any of this crap

Quote

06-02-2023 , 06:02 AM

Duncsumn7

journeyman

Join Date: Apr 2022 Posts: 224

they buying out their competition

Quote

06-02-2023 , 06:53 AM

newguyhere

Carpal \'Tunnel

Join Date: Jul 2016 Posts: 7,180

As long as recs don't have to jump through hoops to make a deposit, online poker is far from dead.

Even regs that are utilizing these tools still make plenty of mistakes.

Quote

06-02-2023 , 07:38 AM

Xenoblade

Pooh-Bah

Join Date: Sep 2017 Posts: 4,090

Quote:

Originally Posted by Koshko

it’s not just recalcs though, it solves really fast with ranges from scratch

Quote

06-02-2023 , 03:55 PM

#10

MicroDonkYT

adept

Join Date: Jan 2020 Posts: 818

Quote:

Originally Posted by Koshko

Using GPU and neural nets are orders of magnitude faster than brute forcing through CPU.

This is for Deepsolver, which I assume Ruse is in the same ballpark:

Not to mention, the Pio interface looks like it is from 1996

Quote

06-02-2023 , 06:27 PM

#11

Koshko

adept

Join Date: Mar 2021 Posts: 901

Quote:

Originally Posted by MicroDonkYT

Using GPU and neural nets are orders of magnitude faster than brute forcing through CPU.

This is for Deepsolver, which I assume Ruse is in the same ballpark:

Not to mention, the Pio interface looks like it is from 1996

Yeah the sad part about this is what eventuates is ranges being built from HUD stats on the fly. RTA will be specific to ranges ie HUD stats, bots won’t be Input output but instead be specific to your ranges.

We are looking at ways to combat this, essentially would be randomising your ranges. Sht getting real fast

Quote

06-02-2023 , 06:57 PM

#12

jayme87

veteran

Join Date: Apr 2010 Posts: 3,496

as they say, if you cant beat em, join em

Quote

06-02-2023 , 07:12 PM

#13

Joe-exotic69

old hand

Join Date: Apr 2020 Posts: 1,302

Quote:

Originally Posted by Elizabeth5579

prefer live poker where you dont have to worry about any of this crap

Some kid was using a solver live on his phone.

Honestly in 20 years everyone will have a RTA in their glasses/contact lenses, the people who dont use them will be the whales.

Quote

06-02-2023 , 07:54 PM

#14

Dejavudu666

grinder

Join Date: Mar 2020 Posts: 428

Or cfr in their brain

Last edited by Dejavudu666; 06-02-2023 at 08:14 PM.

Quote

06-02-2023 , 07:56 PM

#15

borg23

Carpal \'Tunnel

Join Date: Nov 2012 Posts: 15,049

Quote:

Originally Posted by Joe-exotic69

Some kid was using a solver live on his phone.

Honestly in 20 years everyone will have a RTA in their glasses/contact lenses, the people who dont use them will be the whales.

Probably true and the guys you actually want playing won't be playing

Last edited by borg23; 06-02-2023 at 08:05 PM.

Quote

06-03-2023 , 05:12 AM

#16

Koshko

adept

Join Date: Mar 2021 Posts: 901

Quote:

Originally Posted by MicroDonkYT

Using GPU and neural nets are orders of magnitude faster than brute forcing through CPU.

This is for Deepsolver, which I assume Ruse is in the same ballpark:

Not to mention, the Pio interface looks like it is from 1996

This is interesting, if they can pull off instant recalcs from any stack size and vs a specific stat, ie 13% btn vs co 3b this would render an insane advantage.

It leaves me trying to think of ways to counter the strategy and the only thing that comes to mind would be randomising ranges.

Quote

06-03-2023 , 06:02 AM

#17

ioanguradeaur

grinder

Join Date: Jul 2022 Posts: 507

Tin foil hats ITT

Quote

06-03-2023 , 06:37 AM

#18

Dejavudu666

grinder

Join Date: Mar 2020 Posts: 428

Quote:

Originally Posted by Koshko

Wizard said they had sufficient "Anti-RTA" measures in place. I think the backlash they'd face from giving players carte blanche to use in game would disincentivize, that being said, as you said, there is players/stables utilising Wizards current offerings to cheat.

Hopefully it's not feasible to take the browser based client and utilise it offline though and manipulate the code.

Quote

06-03-2023 , 06:42 AM

#19

Koshko

adept

Join Date: Mar 2021 Posts: 901

Quote:

Originally Posted by Dejavudu666

Yeah it would be the Non Anti-RTA dealings I would be worried about where they simply just sell the API and then anyone could weaponise it.

Quote

06-03-2023 , 11:09 AM

#20

MicroDonkYT

adept

Join Date: Jan 2020 Posts: 818

Quote:

Originally Posted by Koshko

Probably no way to game it, tbh. The best thing they can do it have a cumulative time bank that restricts you the more you use it in a short period of time.

Quote

06-03-2023 , 11:32 AM

#21

MastaAces

Carpal \'Tunnel

Join Date: May 2007 Posts: 10,010

Quote:

Originally Posted by jayme87

as they say, if you cant beat em, join em

adapting Polk's motto of if you can't join 'em, beat 'em will prove difficult.

Quote

06-04-2023 , 07:33 PM

#22

dude45

veteran

Join Date: Sep 2019 Posts: 2,807

dont know a lot about Using NNs in poker but a saw a YT vid comparing the 2 a few months ago and the NN solutions were markedly different than the solver solutions in a lot of spots. So are NNs close to the accuracy of a solver ?

Quote

06-04-2023 , 09:20 PM

#23

Koshko

adept

Join Date: Mar 2021 Posts: 901

Quote:

Originally Posted by dude45

Without going into detail without too much context that may seem a little bit more complicated(over complicated).

Here is a link where the two are somewhat explained.

https://www.scitepress.org/papers/2014/51489/51489.pdf

Reinforcement learning in regards to Neural networks as opposed to simple Input Output solved sims.

Quote

06-04-2023 , 10:48 PM

#24

Ceres

veteran

Join Date: Mar 2019 Posts: 3,058

I thought gtow already had a gto trainer

Quote

06-04-2023 , 11:38 PM

#25

dude45

veteran

Join Date: Sep 2019 Posts: 2,807

Quote:

Originally Posted by Koshko

Ok NN versus solver over 10 million hands who wins ? I assume the solver. Also that link is weird when I try to scroll it just puts a blue line

Quote

Page 1 of 2

First

1 2

Last

Post Reply Subscribe

...

Page 1 of 2

First

1 2

Last