One Billion Online Hand Database

magicpigmagicpig Red Chipper Posts: 97 ✭✭
In a recent podcast Doug mentioned that there is a one billion hand database of online hands out there.

Does anyone have links to observations made from that database?

Or observations of their own?

Comments

  • Doug HullDoug Hull RCP Coach Posts: 1,857 -
    I think the site is dead. This is the best hit Google had.

    http://deadspin.com/what-a-billion-poker-hands-can-tell-us-about-luck-1463281627

    Co-founder Red Chip Poker,
    Author Poker Plays You Can Use
    Author Poker Workbook for Math Geeks
  • magicpigmagicpig Red Chipper Posts: 97 ✭✭
    Doug Hull wrote: »
    I think the site is dead. This is the best hit Google had.

    http://deadspin.com/what-a-billion-poker-hands-can-tell-us-about-luck-1463281627

    Thank you
  • magicpigmagicpig Red Chipper Posts: 97 ✭✭
    Doug Hull wrote: »
    I think the site is dead. This is the best hit Google had.

    http://deadspin.com/what-a-billion-poker-hands-can-tell-us-about-luck-1463281627

    Hi Doug,

    Do you remember any interesting observations from the database or your discussion with collector of the data?
  • Doug HullDoug Hull RCP Coach Posts: 1,857 -
    Not really, I wish I did.
    Co-founder Red Chip Poker,
    Author Poker Plays You Can Use
    Author Poker Workbook for Math Geeks
  • persuadeopersuadeo Red Chipper Posts: 4,296 ✭✭✭✭✭
    I would posit that the Billion Hand Database much less valuable than it sounds like it is and that treating the game like an EV puzzle for hole cards by position is part of many players' weakness.
  • The MuleThe Mule Red Chipper Posts: 790 ✭✭✭
    persuadeo wrote: »
    I would posit that the Billion Hand Database much less valuable than it sounds like it is and that treating the game like an EV puzzle for hole cards by position is part of many players' weakness.

    I'm going to assume this is a reference to @Riverboat Bill 's thread on playing small suited connecters (maybe I'm being sensitive now ?). Apologies in advance for hijacking your thread @magicpig - doesn't sound like there was much insight from the billion hand database, hopefully we can get some good discussion from a tangent.

    Knowing what hands you can profitably open is difficult. Even with tracking software online it will take a long time to build up data on the profitability of a particular hand type, and even then for the most marginal hands the estimation error will be large relative to the estimated EV. The population won't be i.i.d. because your game will have changed over time etc. Estimating profitability of a hand from live play will be even tougher, with all sorts of cognitive biases to contend with as well.

    So how does a player know what hands they can play profitably ? A database aggregating thousands of other players hands across a wide variety of games probably isn't a good indicator of what hands I can play profitably myself. The opening ranges of high stakes pros aren't particularly relevant to what hands I can play profitably in a very different game using a very different and obviously inferior strategy. And even what I consider the best place to start, hand charts produced by content providers specifically for the games we are playing in, are rubbished by the content providers.

    The discussion on suited connectors focussed on the EV/potential EV of opening these hands, and was cut short before fully addressing @kagey's very good questions of why would we open these type of hands. Convincing ourselves these can be played profitably is just a start - just because someone playing a perfect GTO game could profit playing small suited connectors in position doesn't mean the rest of us can, due to errors in our implementation for a start.

    But if GTO provides the constant framework from which good exploitative winning poker can be measured, then I think an insight into the GTO solution provides a useful reference point, and a goal that can be worked towards. If my goal is to build a game where GTO is the backbone, then I will need to be able to play hands that I believe are part of a GTO opening range. To be even better, I will need to understand whether or not it's the best play in a given situation (the rest of the answer to @kagey 's question...).
  • Doug HullDoug Hull RCP Coach Posts: 1,857 -
    They were using the DB for some really interesting stuff, beyond the obvious. Too bad that money seems to have dried up.

    I am a pretty big believer in Big Data. If empirically certain hands were just losers UTG over that many hands, lets say by 1BB, I don't think any skill differential is going to turn them into long term winners for any of us. If they are marginal winners or losers -0.1BB to +0.1BB then I am more open to that argument that these hands have
    • Deception value to help the other hands.
    • Skill gap over the herd can turn them into winners.
    • If they are marginal, I am not winning or losing much, and I like to play hands.

    There is absolutely a set of hands that are unprofitable in certain positions. There are absolutely hands that are profitable in certain positions. There is a ton in the grey area between. I think Big Data like this can tell us where the major, unquestionable lines are drawn and tell us a little bit about how profitable those lands are as a baseline.

    There is a ton of "maybe" in between.
    Co-founder Red Chip Poker,
    Author Poker Plays You Can Use
    Author Poker Workbook for Math Geeks
  • persuadeopersuadeo Red Chipper Posts: 4,296 ✭✭✭✭✭
    If it were that compelling it never would have disappeared.
    colldav wrote: »
    persuadeo wrote: »
    I would posit that the Billion Hand Database much less valuable than it sounds like it is and that treating the game like an EV puzzle for hole cards by position is part of many players' weakness.

    I'm going to assume this is a reference to @Riverboat Bill 's thread on playing small suited connecters (maybe I'm being sensitive now ?). Apologies in advance for hijacking your thread @magicpig - doesn't sound like there was much insight from the billion hand database, hopefully we can get some good discussion from a tangent.

    No, it's not a cross-thread reference but I see that there is some overlap in discussion topic.

    I would shocked out of my socks if the 1BHD told us anything we couldn't reason out. Preflop ranges are fungible because they are not the drivers of profit.

    By the time the researchers tell you margarine is bad you are already fat.



Leave a Comment

BoldItalicStrikethroughOrdered listUnordered list
Emoji
Image
Align leftAlign centerAlign rightToggle HTML viewToggle full pageToggle lights
Drop image/file