# [Computer-go] algorithm quality assessment

Nick Wedd nick at maproom.co.uk
Thu Jun 13 01:02:36 PDT 2013

```On 13/06/2013 01:16, Jason House wrote:
> The algorithm I put was based on the first link which did not have an
> open face variant.
>
> My interpretation was as follows:
> After splitting your 13 cards into 3 hands, the top hands are compared
> to each other. The best top hand gets a point and the lowest loses a
> point. The same goes for the middle and bottom hands. That yields a
> score between -3 and +3.
>
> If you dump everything into one best hand, you're easily defeated by
> opponents who try to keep their hands at comparable strength... You
> would earn one point for your best hand and lose one point from each of
> the other two hands (net result = -1 point). There is some kind of

You might even achieve the best back hand and still lose 3-0, if you end

Nick

>
> Sent from my iPhone
>
> On Jun 12, 2013, at 6:08 PM, Don Dailey <dailey.don at gmail.com
> <mailto:dailey.don at gmail.com>> wrote:
>
>> I don't quite see the point.   The goal is to find the best possible
>> hand YOU can make with your 13 cards and there is no betting,   so I
>> see no point in using Monte Carlo here.
>>
>> What am I missing?
>>
>> Is it whether to sacrifice one of the 3 hands to strengthen the other
>> 2?  Or in the case of a really bad hand to at least make 1 really
>> strong hand?
>>
>> Don
>>
>>
>> On Wed, Jun 12, 2013 at 6:03 PM, Jason House
>> <jason.james.house at gmail.com <mailto:jason.james.house at gmail.com>> wrote:
>>
>>     For a particular breakdown into 3 hands, it should be possible to
>>     do a monte carlo simulation by randomly distribute the remaining
>>     cards to the other players and then randomly separating each
>>     player's cards into 3 hands. A node in the search tree would be
>>     scored as the average result of many simulations.
>>
>>     I can think of a few ways to build a search tree. If you have
>>     experience in the game and know a few general strategies, they
>>     will likely be very handy for achieving enough strength to
>>     evaluate the approach. The search tree should be able to give
>>     feedback on which strategy is best. The same strategies may also
>>     help improve the random opponents, but that might require more
>>     care. It's easy to introduce bias.
>>
>>     Sent from my iPhone
>>
>>     On Jun 12, 2013, at 4:06 PM, Oleg Barmin <j2ee_designer at mail.ru
>>     <mailto:j2ee_designer at mail.ru>> wrote:
>>
>>>     Sure. It's open chinese poker:
>>>     http://www.pokerlistings.com/poker-rules-chinese-poker
>>>
>>>
>>>     Среда, 12 июня 2013, 20:57 +01:00 от Nick Wedd
>>>     <nick at maproom.co.uk <mailto:nick at maproom.co.uk>>:
>>>
>>>         On 12/06/2013 20:33, Oleg Barmin wrote:
>>>         > > For quality assessment, play many games against one or
>>>         more reference
>>>         > opponents.
>>>         > It's difficult to assament algorithm with a game against
>>>         humans. The
>>>         > game is young and there are no recognized masters at the
>>>         moment. So it's
>>>         > very hard to find human-opponent with a really good game
>>>         skills.
>>>         >
>>>         > > With card games you can get some serious intransitivity,
>>>         rocks,
>>>         > paper, scissors type of stuff.
>>>         > The aim of this game is to max your scores. Each turn you
>>>         need to select
>>>         > one of three choices. Each choice has an expectation value
>>>         of your
>>>         > scores. Optimal strategy here is to select a choice with
>>>         max expectation
>>>         > value. But it will take a years to calculate an expectation
>>>         value at the
>>>         > start of the game. So the game has no such intransitivity
>>>         as rocks,
>>>         > paper, scissors.
>>>         > At the last turns we can make a complete choice enumeration and
>>>         > calculate an exact scores expectation value ( does go
>>>         algorithms use the
>>>         > same technique? ) . It's not the way for the first half of
>>>         the game. But
>>>         > the first half is more important.
>>>
>>>         Can you give a link to the rules of this game? Or even just
>>>         tell us its
>>>         name?
>>>
>>>         Nick
>>>
>>>         >
>>>         > Oleg
>>>         >
>>>         >
>>>         > Среда, 12 июня 2013, 14:24 -04:00 от Don Dailey
>>>         <dailey.don at gmail.com
>>>         <https://e.mail.ru/sentmsg?compose&To=dailey.don@gmail.com>>:
>>>         >
>>>         >
>>>         >
>>>         > On Wed, Jun 12, 2013 at 11:30 AM, David Fotland
>>>         > <fotland at smart-games.com
>>>         <https://e.mail.ru/sentmsg?compose&To=fotland@smart%2dgames.com>
>>>         > <sentmsg?mailto=mailto%3afotland at smart%2dgames.com
>>>         <http://2dgames.com>>> wrote:
>>>         >
>>>         > For quality assessment, play many games against one or more
>>>         > reference opponents.
>>>         >
>>>         >
>>>         > Especially for a game that is not a game of perfect
>>>         information such
>>>         > as go or chess. With card games you can get some serious
>>>         > intransitivity, rocks, paper, scissors type of stuff.
>>>         >
>>>         > Don
>>>         >
>>>         >
>>>         > ____
>>>         >
>>>         > __ __
>>>         >
>>>         > David____
>>>         >
>>>         > __ __
>>>         >
>>>         > *From:*computer-go-bounces at dvandva.org
>>>         <https://e.mail.ru/sentmsg?compose&To=%2acomputer%2dgo%2dbounces@dvandva.org>
>>>         >
>>>         <sentmsg?mailto=mailto%3acomputer%2dgo%2dbounces at dvandva.org
>>>         <https://e.mail.ru/sentmsg?compose&To=sentmsg%3fmailto%3dmailto%253acomputer%252dgo%252dbounces@dvandva.org>>
>>>         > [mailto:computer-go-bounces at dvandva.org
>>>         <https://e.mail.ru/sentmsg?compose&To=computer%2dgo%2dbounces@dvandva.org>
>>>         >
>>>         <sentmsg?mailto=mailto%3acomputer%2dgo%2dbounces at dvandva.org
>>>         <https://e.mail.ru/sentmsg?compose&To=sentmsg%3fmailto%3dmailto%253acomputer%252dgo%252dbounces@dvandva.org>>]
>>>         > *On Behalf Of *Oleg Barmin
>>>         > *Sent:* Wednesday, June 12, 2013 8:02 AM
>>>         > *To:* computer-go at dvandva.org
>>>         <https://e.mail.ru/sentmsg?compose&To=computer%2dgo@dvandva.org>
>>>         > <sentmsg?mailto=mailto%3acomputer%2dgo at dvandva.org
>>>         <https://e.mail.ru/sentmsg?compose&To=sentmsg%3fmailto%3dmailto%253acomputer%252dgo@dvandva.org>>
>>>         > *Subject:* [Computer-go] algorithm quality assessment____
>>>         >
>>>         > __ __
>>>         >
>>>         > Hi, everybody,____
>>>         >
>>>         > I am working at the development of a cards game algorithm using
>>>         > MCTS. Technically, the game model is expect minmax tree search,
>>>         > where direct search takes up too much time, that is why I
>>>         > decided to use MCTS.____
>>>         >
>>>         > The issue of using MCST, like any other approximation algorithm
>>>         > is its quality assessment. I am developing an algorithm for a
>>>         > game where no recognized masters exist. How do you think, guys,
>>>         > if for instance Go (or Amazons) provided no way to assess an
>>>         > algorithm playing with professional gamers (or other programs),
>>>         > how would you assets its quality?____
>>>         >
>>>         > My second question: I have not yet learned Go in and out,
>>>         > however in my opinion, any search of a next step should
>>>         identify
>>>         > a number of options with similar or even the same assessment.
>>>         > How do you resolve this issue?____
>>>         >
>>>         >
>>>         > Best regards,
>>>         > Oleg Barmin.____
>>>         >
>>>         >
>>>         > _______________________________________________
>>>         > Computer-go mailing list
>>>         > Computer-go at dvandva.org
>>>         <https://e.mail.ru/sentmsg?compose&To=Computer%2dgo@dvandva.org>
>>>         > <sentmsg?mailto=mailto%3aComputer%2dgo at dvandva.org
>>>         <https://e.mail.ru/sentmsg?compose&To=sentmsg%3fmailto%3dmailto%253aComputer%252dgo@dvandva.org>>
>>>         > http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
>>>         >
>>>         >
>>>         >
>>>         >
>>>         > Best regards,
>>>         > Oleg Barmin.
>>>         >
>>>         >
>>>         > _______________________________________________
>>>         > Computer-go mailing list
>>>         > Computer-go at dvandva.org
>>>         <https://e.mail.ru/sentmsg?compose&To=Computer%2dgo@dvandva.org>
>>>         > http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
>>>         >
>>>
>>>
>>>         --
>>>         Nick Wedd
>>>         nick at maproom.co.uk
>>>         <https://e.mail.ru/sentmsg?compose&To=nick@maproom.co.uk>
>>>         _______________________________________________
>>>         Computer-go mailing list
>>>         Computer-go at dvandva.org
>>>         <https://e.mail.ru/sentmsg?compose&To=Computer%2dgo@dvandva.org>
>>>         http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
>>>
>>>
>>>
>>>     Best regards,
>>>     Oleg Barmin.
>>>     _______________________________________________
>>>     Computer-go mailing list
>>>     Computer-go at dvandva.org <mailto:Computer-go at dvandva.org>
>>>     http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
>>
>>     _______________________________________________
>>     Computer-go mailing list
>>     Computer-go at dvandva.org <mailto:Computer-go at dvandva.org>
>>     http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
>>
>>
>> _______________________________________________
>> Computer-go mailing list
>> Computer-go at dvandva.org <mailto:Computer-go at dvandva.org>
>> http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
>
>
> _______________________________________________
> Computer-go mailing list
> Computer-go at dvandva.org
> http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
>

--
Nick Wedd
nick at maproom.co.uk

```