Stats post - so I'm getting reamed in the ass in the BB

[griffey24]

Look at it this way. Your VPIP is pretty much 20% in the BB, so you're defending 20%.
This means 80% of the time you are losing 1bb per hand or 100bb/100.

0.8*(-100)+0.2*x = -47bb/100
x = 165bb/100

This means that when you ARE calling you are running at 165bb/100 (if my math/method is right). So yes it is VERY obvious that you are not defending enough. Like mojo said, your baseline is -100bb/100. You should be playing EVERY hand in the BB that is expected to do better than -100bb/100.

Also you should probably be defending MORE from the BB than SB, given your baseline for the BB is -100bb/100 and the baseline in the SB is -50bb/100. I don't think the times you get to steal BVB should make up for such a big discrepancy.

[BorisTheSpider]

Look at it this way. Your VPIP is pretty much 20% in the BB, so you're defending 20%.
This means 80% of the time you are losing 1bb per hand or 100bb/100.

0.8*(-100)+0.2*x = -47bb/100
x = 165bb/100

This means that when you ARE calling you are running at 165bb/100 (if my math/method is right). So yes it is VERY obvious that you are not defending enough. Like mojo said, your baseline is -100bb/100. You should be playing EVERY hand in the BB that is expected to do better than -100bb/100.

Your maths looks entirely correct, and that is downright shocking.

Do you know of any good blind defence study material?

Do you have any suggestion for the kinds of hands to 3bet from the BB vs a LP open, as opposed to what I should be calling with?

Can I treat it like any other position in terms of 3betting with stuff I can't call with? Obviously, if I need to loosen up a ton in the BB, that also suggests though that I can probably call with quite a lot more than I think I can.

The problem then becomes, I feel like if I can call with a lot more than I think, I then have to play OOP postflop, and that is going to sap a fuckton of profitability out of a lot of the hands that I "can" call with (when looking purely at their preflop equity against a loose LP open), so I think I need to be careful not to over-adjust here.

I guess I feel lost here, because I don't really call OOP like ever unless there is a very specific reason to do so, and so I feel pretty clueless where to start.

The danger on the flip-side, is that if I don't think I can call too much because I don't want to have to play postflop OOP, then I need to 3bet more, but again I don't want to overadjust and start spewing, nor end up playing a lot of 3bet pots OOP.

[griffey24]

Also I've seen those -15bb and -30bb figures you referenced too, but I've never been able to sustain those levels in the long run. I probably suck in the blinds as well though.

[BorisTheSpider]

OK, so I'm working on some ranges now. BTW it seems you can't add stats in the "Opponents" view in HEM, so I can't get an average BTN or CO steal for the player pool that way, but what I was able to do was create a player alias called "Regs >1000 hands" then add one by one all the players I have a big sample on.

That gave me average CO steal of about 25% and about 35% from the button, which sound reasonable, so I am now working on my default defence ranges based on that.

I really have no idea where to start deciding whether I can play a hand more profitably (or less unprofitably) than -100bb/100, although as Griffey said that is clearly the dividing line as to where it's better just to fold. I guess the only way I can start to use that idea, is to say that if BTN and I just got it in preflop every single time, I would need 49% equity to lose 1bb every time a hand was played, but just using preflop equity like this seems to simplistic. I guess at least it provides a starting point. I know I need to make some adjustment to the equity based on having to sometimes play postflop OOP, but how much? I have no idea how to quantify that.

Let's do a little maths based on the 35% average BTN open from my database.

In a typical 3bet defence, if villain opens 3x on the button, the pot will be 4.5bb and we will risk 10bb to win it, so if we were always bluffing, we'd need villain to fold 69% of the time. That would mean they'd play the top 11% of hands and fold the rest.

If they call more often than 31% of the time, they'd be calling with a wider weaker range so our 3bets for value should make more (or lose less), and if they fold more than 69% of the time we do better by taking it down preflop even if we're bluffing with the best hand a lot. Whether these opposing factors balance each other out, I do not know, nor do I have any idea off the top of my head how to further analyse that.

Let's just give him an 11% continuing range, so as to have somewhere to start from.

Leaving aside any 4bet bluffing, that would give them a continuing range something like [AA-TT, A2s+, KTs+, QTs+, AJo+, KQo]

In reality, they can actually continue a bit wider - again, we're not trying to make money here, we're trying to lose less than -100bb/100.

But it does mean we have a starting point for the value section of our 3bet range from the BB - something like [AA-TT, ATs+, KTs+, QTs+, JTs, KQo, AJo+] has better than 50% equity against the above continuing range.

Any ideas on how much of our total 3bet range from the BB vs a BTN open should comprise bluffs as opposed to our widened value range?

Also, if villain calls a lot when 3bet BTNvBB, we can widen the value range above out even more, probably like any broadway, and maybe any suited ace and a few smaller PP if they call really wide. OTOH, if someone folds their steal attempts at an unusually high frequency to a BB 3bet, we can bluff wider with a more polarised range.

Here's my initial stab at some specific default ranges (all assuming an "average" villain):

3bet for value: [AA-TT, ATs+, KTs+, QTs+, JTs, AJo+, KQo] (118 combos, 9% of hands)
3bet as a bluff: [K2s-K7s, 75s-97s] (36 combos, 2.7% of hands)
Call: [88-99, A2s-A9s, K8s-K9s, T8s-J9s, 65s-T9s, A2o-ATo, KTo-KJo, QTo-QJo, JTo] (248 combos, 18.7% of hands]

So then we'd be playing 30% of hands against a 35% opening range, of those hands we did play 2/3 of the time, we'd call (so with a capped range, comprising hands from 10%-30% of total hands). The other third of the time, we'd 3bet, and when we did 3bet 2/3 of the time it would be for value, and 1/3 of the time as a bluff.

Perhaps I should think about whether rather than capping our calling range, we should polarise it a bit, calling with some relatively strong stuff from the 3bet value range suggested above, like KQo, AJo and AQo to give us some relatively nutted calling hands, but then I feel like we'd be calling too much, so I'd want to put some more 3bet bluffs in the bluffing range, and I wonder if the overall 3bet range would become too bluff heavy.

Any input appreciated.