Weird HPFAP blind defense strategy inconsistencies

[Nigel]

So I'm taking a day off from playing and was bored this morning so I went back to read some HPFAP and found some things a little strange. I thought I'd ask about it...

On page 46 they advise stealing from the SB with any 2, when opening, if the BB will fold as low as 30%. I think this fits most players folding tendencies, especially if you follow S&M's own defense recommendations. So should we be raising any 2 in the SB when opening? I am not a fan of playing total trash out of position, so I usually show some restraint in the SB when opening. But like I said, I think most BB's will fold over 30% here, so maybe I need to open up my game in this spot?

Then on page 47, concerning the BB, they go on to repeat that the SB will most likely show an automatic profit if you fold your BB fold's even 30% here. So what does that leave you calling with, 80% or 90% of your hands? Doesn't seem like there is much room for error, so shouldn't we just call with basically everything since we have position and most players are so, so bad HU?

So far no problems, until we get to the shorthanded section on page 185, in which they contradict their earlier advice, and suggest 50% as the minimum number at which you must call with in a HU match. Maybe the BB being out of position is the reason for the discrepancy, but I don't see how position changes the mathematics behind the SB showing an automatic profit from the BB folding too much pre-flop. And, to confuse things even further, 2 pages later on page 187, they steer away from that 50% number and now suggest you defend at least 40% (so folding 60% is ok???) and then give a hand range totalling 39%. All this after they have just told us that folding even 50% (or is it 30%?) will give the SB an automatic profit. What gives?

Lastly, on page 44, concerning SB defense from a steal, they suggest you might want to reraise groups 1-6!!! Since ASB's of well over 40% are now all the rage, we can basically consider any CO or button raise a "steal attempt", and therefore S&M are suggesting we 3 bet hands group 6 hands like 75s out of the small blind? By that logic, if you are going to play 75s, or J8s, for 3 bets OOP from the small blind, what hands shouldn't you 3 bet *with position* on the button when the CO open raises??? Might as well 3 bet with 1/2 the deck!

If I made changes to my game to be consistent with the above advice (assuming I can determine what they are actually advising), my SB/BB defense would go from 85/50 to probably 60/30 and my overall VPIP would probably go up 15%. Were S&M hitting the pipe when they wrote this stuff? This can't be right. Maybe I'm just being an idiot and missing what they are trying to say, but can someone explain this to me so it makes sense, or is there a reason I don't pick this book up much anymore?

Cheers,

Nigel

[Victor]

[ QUOTE ]
On page 46 they advise stealing from the SB with any 2, when opening, if the BB will fold as low as 30%. I think this fits most players folding tendencies

[/ QUOTE ]

i contend that most 10/20 and 5/10 regulars call nearly 100% in the after a sb raise.

[stoxtrader]

nice work summarizing this stuff nigel - it's insanely important I think, and poorly understood - one of the last "frontiers" of theory that seems to be missing a lot.

There are two ways to approach this problem, I think one is to find a general guideline, which HPFAP has done somewhat (I think the advice is actually pretty good overall, and hits on the major points, but is definitely inconsistent). The second way to to isolate specific ranges and determine what ranges to play versus "that".

This stuff is so difficult because even if I gave you a relatively specific range for some players (and with large databases we can approach that), it is still debateable what the correct play is even for pre-flop, forget about post-flop.

Let's say a typical SH player open raises 50% from the SB, what should your calling, folding and 3 betting ranges be in the BB?

How about if you are in the SB and know that the BB defends with the "optimal" 70% of the time, what should your folding/limping/raises standards be?

further, to what extent should blind structure change the ranges?

someone really really good could come close to filling a book worth of material on this stuff, anyone who could without getting lost or contradicting themselves is a genius.

questions answered with more questions. sorry.

[Nigel]

[ QUOTE ]
[ QUOTE ]
On page 46 they advise stealing from the SB with any 2, when opening, if the BB will fold as low as 30%. I think this fits most players folding tendencies

[/ QUOTE ]

i contend that most 10/20 and 5/10 regulars call nearly 100% in the after a sb raise.

[/ QUOTE ]

Ok, so I actually went and checked. It's about 25% at 30/60 and higher for me. When I played 10/20, the BB folded a little closer to 30%.

That would bring you pretty close to wanting to steal with any 2 if their 30% calling requirement is correct, which is what one of the key things I'm questioning in their advice.

[Victor]

ok. it seems to me there is a large class of players that defend any 2 there. however, there are clearly some tighter players that fold as high as 50%. hell, i am one of them.

[TStoneMBD]

against a weaktight BB i openraise pretty much any 2 at 15/30 with its 2/3 structure.

if the BB is a tag i give him the $10 when im holding some low unsuited crap.

if the BB is high vpip/low pfr i complete with any 2 and bet almost any flop if hes not the chasing type.

against a SB openraise from a guy with 8+PFR i call with 67o/68o and almost any 2 suited.

are any of these guidelines leaks i dont know, let me know if you do.

[Nigel]

TStone - Those guidelines are pretty similar to my standards, almost the same actually.Your guess is as good as mine as to how "correct" this is, but personally think I give up too much not defending or stealing quite enough. It's why I decided to break out the books and hit up a little Google today, to see if I could revamp my stealing and defense strategies.

Stox - give me a few as I want to write out a longer reply to your post. I'm glad you posted in this thread because you probably have as good a grasp as anyone on this.

[The Truth]

[ QUOTE ]
nice work summarizing this stuff nigel - it's insanely important I think, and poorly understood - one of the last "frontiers" of theory that seems to be missing a lot.

There are two ways to approach this problem, I think one is to find a general guideline, which HPFAP has done somewhat (I think the advice is actually pretty good overall, and hits on the major points, but is definitely inconsistent). The second way to to isolate specific ranges and determine what ranges to play versus "that".

This stuff is so difficult because even if I gave you a relatively specific range for some players (and with large databases we can approach that), it is still debateable what the correct play is even for pre-flop, forget about post-flop.

Let's say a typical SH player open raises 50% from the SB, what should your calling, folding and 3 betting ranges be in the BB?

How about if you are in the SB and know that the BB defends with the "optimal" 70% of the time, what should your folding/limping/raises standards be?

further, to what extent should blind structure change the ranges?

someone really really good could come close to filling a book worth of material on this stuff, anyone who could without getting lost or contradicting themselves is a genius.

questions answered with more questions. sorry.

[/ QUOTE ]

Great post.

I have actually been in the process for the last month of trying do just this. I didn't intend to write a book, but I have so much Excel data spreads and word documents I might.

It is very complex, and one of the most difficult parts is coming to a conlusion about how post flop works in isolation with ev of one hand vs a range. I have read some good post from Peter_rus, and even wrote some queries myself to get an idea of positional advantage modifers, and this is actually working out well.

It is really computational and time consuming, but I think I will be able to draw some good conclusions soon.

This actually makes preflop more clear cut from every position, including the blinds.

The bad part is, as I start adjusting for more players per pot and just the overall more complex situations, the math becomes even more tedious.

If anyone wants to join in and help, let me know.

-blake

[Subfallen]

Andrew Prock already knows all this stuff exactly, I guarantee it.

Edit - no, he isn't going to help you. E.g. here and here and here.

[Nigel]

Hey Stox,

You know, you are one person I was really hoping would respond here. In fact, I was thinking of crossposting this at your site, or PMing you similar questions, because I imagine this is an area you shine.


[ QUOTE ]
it's insanely important I think, and poorly understood - one of the last "frontiers" of theory that seems to be missing a lot.

[/ QUOTE ]

I agree completely. Far too little literature for something that probably holds the key to truly world class high stakes limit play. I think it pretty much defines your win/loss rate.


[ QUOTE ]
There are two ways to approach this problem, I think one is to find a general guideline, which HPFAP has done somewhat (I think the advice is actually pretty good overall, and hits on the major points, but is definitely inconsistent). The second way to to isolate specific ranges and determine what ranges to play versus "that".

[/ QUOTE ]

We do have some nice general guidelines, like S&M's stuff and Abdul's work, to name a couple. However, I get concerned that I might be "losing the race before it even starts", so to speak, if I am failing to cover all my bases mathematically, which is why I was so interested in the inconsistencies that Sklansky puts forth as "must defend" and "must steal" percentages. Anyway, I just ran some numbers over my last 50k hands at 30/60 and was pretty scared as to what I discovered. I don't want to say too much more about it yet, but I think I can significantly improve my earn.


[ QUOTE ]
This stuff is so difficult because even if I gave you a relatively specific range for some players (and with large databases we can approach that), it is still debateable what the correct play is even for pre-flop, forget about post-flop.

[/ QUOTE ]

It's overwhelming, especially since potential post-flop mistakes (which, I imagine are made in abundance for those who aren't true HU specialists) can dictate so much about how we might want to aproach pre-flop strategy. However, looking at pre-flop from just a strict numbers point of view, I think I'm coming up short. Basically, I aspire to be more where you are - your SB VPIP is slightly higher than mine, and you defend your SB and BB slightly more than I do. I think small improvements in these areas can make massive differences in one's earn, both short-handed and full, so I am focused 100% on making adjustments over large sample sizes to see what unfolds. Unfortunately, as we all know, there is more to this than just mimicking stats.


[ QUOTE ]
Let's say a typical SH player open raises 50% from the SB, what should your calling, folding and 3 betting ranges be in the BB?

How about if you are in the SB and know that the BB defends with the "optimal" 70% of the time, what should your folding/limping/raises standards be?

further, to what extent should blind structure change the ranges?

[/ QUOTE ]

Well, this is where what happens post-flop becomes very interesting because, from what I can see, and as I alluded to earlier, there are probably fundamental, post-flop, mathematical errors being made by many players in these HU situations, especially multi-tablers playing very fit or fold. I would think these errors, if they are indeed being made to the degree I suspect they are, could potentially negate differences between pre-flop blind structures and lean towards a never limp strategy from the SB (or the button, for those into that sort of thing) when trying to formulate pre-flop guidelines. As for defense, it becomes quite complicated because, if we are not careful, it becomes very easy to fall victim to the same things we might want to exploit as the stealer.


[ QUOTE ]
someone really really good could come close to filling a book worth of material on this stuff, anyone who could without getting lost or contradicting themselves is a genius.

[/ QUOTE ]

Well, if Wall Street ever gets old, maybe you have your calling [img]/images/graemlins/wink.gif[/img] Seriously though, the best part is (well, maybe not for the author) that nobody would probably read it. Surprisingly, there just doesn't seem to be a huge interest amongst players to try and get to the meat of this stuff - maybe it's just too complex to try and tackle. Who knows.


[ QUOTE ]
questions answered with more questions. sorry.

[/ QUOTE ]

No apologies. At least your questions lead to more questions, whereas I just ramble. Also, I know you play under your screenname, so it's understandable you might want to be guarded with how much information you reveal, but what you do share is appreciated. I hope you'll have more comments to make.

Cheers,

Nigel