A discussion about statistics behind BSS starting hand chart

    • faronel
      Joined: 13.07.2011 Posts: 1,186

      As far as I am concerned, the Starting Hand Chart for BSS is a preliminary guideline for novice poker players. In other words, a rough estimate of recommended plays with specific hands.

      I can understand that from one perspective the chart is meant to help the player avoid doing some major mistakes (like overplaying little pockets on pre-flop). However, from the other point of view, I keep wondering how accurate (including statistical analysis) the chart is.

      I will start this thread with one example. I was playing NL4 full-ring (10-seated table). I was dealt "the German virgin" (pocket nines) on the button. One player in the MP limped. My first reaction was to raise. However, since being a novice player myself, I decided to consult with the chart and, to my surprise, the chart stated a call. I followed the recommendation, although, regardless the outcome in that hand, I would like to get some clarification on the matter.
      The chart states that pocket tens, the so-called middle pair, is a sufficient hand to raise when one player limps in before you. The pocket nines, the so-called low pair, is a calling hand in the same situation. When I tried to reason the logic behind this cutoff point, I first refered to equity analysis.

      a) When TT is against a random opponent whose range/VPIP is 10%, TT has a winning equity ratio of 53.16% to 46.84%. The nines in the same scenario have a losing equity ratio of 48.62% to 51.38%.

      b) However, in my opinion, 10% is pretty low number for an average player on microstakes, even on full-ring tables: beginners tend to be much more loose (an average of above 30% can be projected, I guess); and by default, the recommended VPIP for a regular play is 10-15%. So, in this light, 99 are actually a much more better winning hand when we take into consideration the average range of full-ring table players on microlimits, and thus, a potential raise hand preflop versus random hand with that average range.

      Getting to my questions and discussion,

      1. I understand that in every hand there are many criterias you should be following before making any specific move. Who limped in, from what position, what is his stack, and so on, and so forth. More of a rhetorical statement...

      2. However, is there a calculation based on some bigger hand sample that would show the average range/VPIP of the entire table or NL2-4 fullring? What would the % be? For example, for PartyPoker community? :)

      3. When comparing to TT and conditions described previously (fullring on microlimits), is raising with 99 actually more losing move?

      4. Hypothetically speaking, what would be more realistic cutoff point for the so-called lower pairs to actually call instead of a raise?
  • 4 replies
    • Jan217
      Joined: 02.01.2009 Posts: 626
      Adjusting vs. oponents ranges >> charts

      As you say the charts are really at best a rough guideline, and are usefull for novices to get an idea for how relatively strong starting hands are. Once you know someone open limps a ton of hands then 99 is an easy isolation raise, as is 88, 77 and sometimes 66, provided you have a good plan for how to play on many uncomfortable flops. Most of the limp calling players do raise TT/JJ+, so the oponent is very unlikely to have any of hands that dominate you, which will increase your preflop equity a lot.
    • supergaijin
      Joined: 03.09.2010 Posts: 133
      IMO the idea behind the BSS starting hand chart has a lot more to do with post flop playability than preflop equity vs probable opponent ranges. It is when you look from this point of view that 99 becomes much less 'valuable' than TT.

      when you have a mid pocket pair like 9's or T's there are 3 (Although there are obviously other issues impacting board texture such as suitedness etc but these factors impact 9's and T's fairly evenly and so can be put aside for this specific discussion) possible post flop scenario's:

      1) you hit a set - (woo hoo) regardless of if you have 9's or T's you have flopped a monster and the hand plays roughly the same for either hand... protect and extract max value

      2) There are over cards - this is where things begin to deviate between the 2 hands. with 9's there is one more overcard to be concerned with (T's) meaning that there will be at least one over card on the flop 7% more often when playing 9's Vs T's. Now although your equity with 9's goes up Vs a villain with 30% calling range rather than 10% the number of hands in their range that hit those flops goes up considerably. In fact 21% of a 30% range has a Ten in it. Furthermore; given the unpredictable post flop play of a generally poor player like hypothetical villain it becomes significantly more difficult to deduce if they have hit the flop and if so to what extent. Conclusion; against a poor player with a wide range of hands there are significantly more times that you are behind after the flop with 9's than T's.

      3) You have an overpair - In this scenario although you have a larger chance of having the best hand now than in 2) if you are ahead at the moment you are significantly more vulnerable with 9's than T's as explained above and villains are much more likely to play over cards on high card 8 boards than high card 9 boards as their is an impact to the likely hood it hit YOUR hand. This is even more so in a raised pot. Also, with all cards ranked 8 or lower the number of boards that give potential straight draws is in creased Vs flops with all cards ranked 9 or lower. adding additional scare cards to be concerned with if villain sees the turn.

      This is why I believe the recommendation in pots in which others where others have already acted for 9's is to play more passively and play primarily for set value. As players post flop abilities improve and are better able to read board texture and opponent ranges and betting patterns it becomes more advisable to widen your raise range in limped pots because of the additional equity you have Vs loose opponents.
    • supergaijin
      Joined: 03.09.2010 Posts: 133
      just want to say I like the discussion you raised!
    • giliNL
      Joined: 13.07.2011 Posts: 56
      Charts are always going to be a guideline for beginning players. Before you develop your own playing style and get a better sense for the value of your hand on that particular table, rather than in general.

      Obviously, Tom Dwan is going to raise with 99. And he's probably going to take down a big pot with it, even if he gets outflopped. But that doesn't mean that it's a statistically sound decision to play that way, or a good decision for every player at every level.

      It's the big difference between NL and limit, and what makes NL fun and exciting and challenging rather than a boring mathematical grind. It's not JUST about the numbers since you can push people around much more. However if you don't know 100% what you're doing, it's pretty helpful to follow some guidelines. Of course then the next step (that you're taking now) is to try and work out why those guidelines are the way they are. And I think apart from the other reasons mentioned or that I will mention, there's also a point where a starting hand chart becomes so complicated that it's unusable. That's why a lot of pros have argued that it's impossible to make a starting hand chart for NL, since there are so many variables. But as a starting point, it's still a pretty good idea I think and perhaps also importantly, it will pretty much guarantee winning play at lower stakes.

      For me as a relatively low level player, I find that sticking more or less to the chart helps protect me from destroying myself on the flop, like supergaijin points out. If your postflop play is not really strong then you can lose a lot of money playing small pairs or things like KQs out of position.

      I came across this article a while ago that might be interesting to you. It's a simulation run for Sklansky&Malmuth's groups of hands done for a Phd thesis (I think). The original article is not available anymore but there's an archive copy here: A New Guide To Starting Hands (2007)

      While this isn't specific to a certain setting like you ask, i.e. NL2/4, it might be interesting to read nonetheless.