BBO Discussion Forums: How Did Folks Who Play a Lot of Robot Hands Get a Good Score on This One? - BBO Discussion Forums

Jump to content

Page 1 of 1
  • You cannot start a new topic
  • You cannot reply to this topic

How Did Folks Who Play a Lot of Robot Hands Get a Good Score on This One?

#1 User is offline   msheald 

  • PipPipPipPip
  • Group: Full Members
  • Posts: 106
  • Joined: 2021-March-17

Posted 2024-May-27, 07:25

Hello! Just curious. On this pre-bid, pre-dealt hand, against me, the robots took their three aces off the top, so that I made 4 spades.

I received a 22%, but I do not see how it is possible to get a better score than 4 spades making 4. However, those who play a lot of robot hands must know how to do so since my "obvious" score only netted a 22%.

For my own education on how to play robot games, how did folks play the hand against the robots with this hand such that they made 5 spades instead of 4? Thank you and best regards.

Mike

https://tinyurl.com/23xpnvw6
0

#2 User is offline   Cyberyeti 

  • PipPipPipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 14,306
  • Joined: 2009-July-13
  • Location:England

Posted 2024-May-27, 12:37

I don't play robot, but looks like 3N may well make 10
0

#3 User is offline   smerriman 

  • PipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 4,105
  • Joined: 2014-March-15
  • Gender:Male

Posted 2024-May-27, 14:33

View PostCyberyeti, on 2024-May-27, 12:37, said:

I don't play robot, but looks like 3N may well make 10

Nope:

View Postmsheald, on 2024-May-27, 07:25, said:

On this pre-bid, pre-dealt hand...

OP - I suspect you're still misunderstanding the scoring like in a similar thread recently, given the way you implied the other players "must know" how to do better.

Suppose you had a two-way finesse and no clue which was better. Half the field finesses one way, and half the other way. You will score 25% half the time, and 75% half the time. Every time you scored 25%, it of course doesn't mean everyone else obviously knew better!

You followed to the diamond lead with the Q. Some players would have followed with a low diamond. Suppose these are completely irrelevant to the play, and 50% of the field do each.

The robots are programmed to generate the same random sample of deals for players who have made the same plays. So everyone who plays high will get the same continuation, and everyone who plays low will get the same continuation. But since these two groups have played different cards, the robot for each group will have generated a different random sample.

Most of the time, both samples lead to the same conclusion, and everyone will score 50%.

Occasionally, one of the samples will cause that robot to take a different line - like continuing with a diamond after the first two aces. In this case, half of the field will score 75%, and half of the field will score 25%. The fact you scored 25% doesn't imply "everyone else took a better line" - it means you were the unlucky recipient of a 50/50 coin flip. Next time, you might score 75% for the exactly the same play.

Putting that aside, it is possible that playing one of low / high is more likely to cause the robot to continue in diamonds, thus truely meaning a better play. Perhaps when it thinks East might hold the queen, it's more likely to want to lead through the king. Or perhaps since you're limited in HCP, having the queen make East marginally more likely to have the ace of clubs. But I did a quick test and the robots seemed to cash their aces roughly the same proportion of times against both defenses.

So if there was indeed a better choice, you'd get 50% most of the time, 25% a small amount of the time, and 75% a small but tiny bit greater amount of the time, for perhaps a 51% long term average - but that long term average will still include some 25%s.

In this case, I expect more people would have blindly followed low to the diamond, so since your choice would be in the smaller group, that could make the numbers 20% and 80% - but still averaging to near 50% in the long run.

But in general, this was just a rare event multiplied by a very close to 50% coin flip that you were on the wrong side of.
0

#4 User is offline   harikannan 

  • PipPipPip
  • Group: Full Members
  • Posts: 76
  • Joined: 2021-June-19

Posted 2024-May-28, 04:15

Smerriman's posts help a lot in understanding gib's quirks. Many thanks to him.
0

#5 User is offline   msheald 

  • PipPipPipPip
  • Group: Full Members
  • Posts: 106
  • Joined: 2021-March-17

Posted 2024-May-28, 04:38

Thank you! I appreciate your guidance. Best regards.

Mike
0

#6 User is offline   notamike 

  • Pip
  • Group: Members
  • Posts: 1
  • Joined: 2024-May-28

Posted 2024-May-28, 14:45

With W bidding 2h and N dbl, and all those low cards in the other suits, 3NT seems reasonable and should make 10 tricks. I get a sense that robots prefer 3NT over 4 of a major quite often.
0

#7 User is offline   LollyPopp 

  • Pip
  • Group: Members
  • Posts: 2
  • Joined: 2010-February-20

Posted 2024-May-28, 15:30

Hello,
I was just reading this cause I'm also very curious how the Robots play
but even more about how they bid
I'm still trying to find out the Robots home adress to send some bad guys to pay them a visit, cause they really drive me crazy with
the way they bid, but it must be my low level of skill
Anyway I hope you all enjoy playing on BBO, even I do altough
sometimes I would like to shoot the PC, but that will probably make
the Robots even laugh more
0

#8 User is offline   gbgb_gbgb 

  • Pip
  • Group: Members
  • Posts: 7
  • Joined: 2011-June-12

Posted 2024-May-28, 15:53

Interesting Comments Simmeriman appreaciated
0

Page 1 of 1
  • You cannot start a new topic
  • You cannot reply to this topic

1 User(s) are reading this topic
0 members, 1 guests, 0 anonymous users