This is a good point to keep in mind: a good mindset is nice, but without the necessary skill and knowledge, it's of little use. This sounds confident, but sometimes poker players forget it, and assume if they can just get their mind in a certain state the results will follow.
+1...... and I think it works both ways. Some often forget that that the knowledge is important while others forget that the mental game and the ability to implement the knowledge is important. I think I fall squarely in the first group.
Oct. 28, 2016 | 6:24 p.m.
Awesome. Will check it out. GL
Oct. 24, 2016 | 1:26 a.m.
Hmm. I am really out of touch with what is going on in the poker world. So let me see if I understand:
- You are saying that off the table work on strategy + implementation of that strategy is a better approach than the at the table improv / deep in-game focus / PG style. ???
If I understand you correctly, why do you think that is the case?
Is it because the the latter approach is just not sustainable / volume conducive for the majority of people?
Or is it more that it just doesn't produce as good of bb/100 results as the other approach?
Oct. 17, 2016 | 12:23 a.m.
Based on associations from our school days, "playing" sounds like a lot of fun and "studying" sounds like a drag... but more and more I'm finding that in my poker work day, studying is the interesting, creative, and fun part, whereas playing mostly feels like going through the motions.
I feel this as well. But I had never explicitly thought about it.
And I think I agree with you reason why this might be (less figuring everything out on the fly and more trying to implement your strat and not [email protected]#$ up).
Just thinking out loud here, but maybe another solution might be to play fewer tables to allow yourself to focus more in depth and on details of hands. There is usually a lot more going on in a hand than 'our strategies' have taken into account. If we take more time and focus on those details and nuances maybe some of the 'playing' / fun mindset will return because again we will be trying to figure stuff out on the fly.
Oct. 15, 2016 | 10:10 a.m.
What I learned:
- I like using notepad, but for something like this it is not a very good way to stay organized
- PPT is more powerful than I thought (For example DirtyD pointed out the I messed up the calculation of for how often all three players fold to a flop bet. Turns out that there is a function in PPT that answers exactly that. Faster and more accurate.)
- After doing this I am more comfortable with the two different ways of writing an 'EV equation'. One where you include your bet in the total pot and then subtract it back out in the 'cost of the bet' part of the equation. And the other, where you don't include your bet in the total pot size and the 'cost of the bet' only occurs when you lose the pot.
- Having people read your work and point out what is wrong with it is a very efficient way to learn.
About the hand:
- About half our EV comes from winning the pot right now with a bet.
- Seems like the story of "Winning a big pot right now or get it in with 40%" seems to due better than "We have a strong hand but we are never ahead when we get called. And it is a 4 way pot so we will get called often"
- The little math I did is not super useful with out tweaking some of the assumptions and seeing things change.
Oct. 12, 2016 | 12:01 p.m.
Goal: I am going try to set in stone a few variables to see if checking back or betting would be better under those conditions. While trying to make it simple enough that it does not take me all day.
First let me make some assumptions. While they may be highly questionable, but we can debate and change adjust them later.
i- all three opponents have ranges of: 50%!AA
ii- all three opponents still have 100% of their AQxx hands after checking
iii- if we face a turn bet: we know that it is from AQ and we will act appropriately
iv- a flop bet will get Q9 to fold if it doesn't have 2p+ to go with it
v- a flop bet will get action from the following worse hands: TT+:Q
vi- Pot = 1; stacks = 2
vii- All bets are committing shoves
EV of a shove = (fold %)(pot) + (equity when called)(pot size)(1-fold %) - (cost of the bet)(1-fold %)
Fold % translates to "how often nobody has AQ, Q9:(KK,JJ,TT,KJ,KT,JT), (KK,JJ,TT):Q
Fold % = 56.5%
Equity when called for all three cases of how we can get it in on the flop:
- vs one opponent we have 39% and it occurs at 37.3% frequency 
- vs two opponents we have 38% and it occurs at 6%
- vs all three opponents we have 37% and it occurs at 0.2%
STORY CHECKS OUT as 37.3 + 6 + 0.2 + 56.5 = 100
EV of a Shove = (.565)(1) + (.39 * 5 * .373) + (.38 * 7 * .06) + (.37 * 9 * .002) - (1-.565)(2) => .59 of the pot
or you can write it like this:
EV of a Shove = (.565)(1) + (.39 * 3 * .373) - (.61 * 2 * .373) + (.38 * 5 * .06) - (.62 * 2 * .06) + (.37 * 7 * .002) - (.63 * 2 * .002)=> .59 of the pot
- before estimating the EV of checking back, lets divide the turn cards into three groups (and then assume how each group will play out)(and at what frequency):
i- board pairs (we win entire pot)(7/45)
ii- turn we turn a FD to go with our set (we check back if checked to or we call with 39% equity vs AQ)(18/45)
iii- blank (we check back or we fold facing a bet with only 23% equity vs AQ)(20/45)
i- no more action on board pair
ii- realize exactly our equity when calling or checking back turn
iii- only up against one player on the turn if we face bet (doing this for simplicity)
iv- we face turn bet 1/2 of the time (this is about how often AQ shows up in someones hand, so it seems reasonable)
vi- AQ is still in their ranges after checking twice (we have around 38% on random turns that don't pair the board or give us a FD, and 50% with a FD)
EV of checking back:
-> (board pairs) + (call on FD and win) - (call on FD and lose) + (win on FD after checking back) + (fold on brick) + (win when checking back on brick)
-> (7/45)(1) + (18/45)(.5)(.39)(2) - (18/45)(.5)(.61)(1) + (18/45)(.5)(.50)(1) + (20/45)(.5)(0) + (20/45)(.5)(.38)(1) => .37 of the pot 
Given these assumptions and the (hopefully correct) math it looks like betting outperforms checking.
Thanks for pointing out my sloppy work and thinking. Please let me know if you see anything else.
Oct. 11, 2016 | 8:15 p.m.
Troubling b/c my intuition fell on the side on Ephedrine's check
Seems like there are quite a few variables here. So I will try to fix a few and see if maybe some math can 'prove' my intuition superior!!!
i- 3 players ranges = 50%!AA
ii- Still have AQxx in ranges after checking
iii- Know that a bet on the turn is from AQ and will make the proper decision
iv- Q9 folds to shove w/o 2p+ to go with
v- TT+ w a Q will also call
vi- Pot = 1; stacks = 2
vii- All bets are committing shoves
EV of a shove:
i) how often you get folds
- Opponent 1 has it 16.5
- When he dont, Opponent 2 has it 17.5
- When both dont, Opponent 3 has it 18
-> we get...... 100 - (16.5 + 17.5 + 18) = 48% folds
ii) equity when we do get it in
- vs one oppenent we have 43
- vs two opponents we have 40
- vs all three opponents we have 38.5
Dont feel like figuring out the proper math, but I think this is close
- up against one .83 * .83 * .17 + .83 * .17 * .83 + .17 * .83 * .83 = 35
- up against two about 1/4 as ofter = 9
- up against all three about 2
-> that equals 46 so we need to add 6 percent to get us to the 52 percent of the time we get it in. Let's add 3, 2, and 1 respectively. So we are getting it in vs one, two, and three with 38, 11, and 3 opponents respectively.
EV of a Shove:
(folds) + (one) + (two) + (all three opponents) - (cost) = EV
(1 * .48) + ((5 * .43 * .38) + (7 * .40 * .11) + (9 * .385 * .03) - 2(.52)) = .67 psb
or you can write it like:
i) what can happen (outcomes)(probability)
a- board pairs (we win entire pot)(7/45)
b- blank FD card(we check back or we call with 39% equity vs AQ)(18/45)
c- blank off-suit card(we check back or we fold with 23% equity vs AQ)(20/45)
i- no more action on board pair
ii- realize exactly our equity when calling or checking back turn
iii- only up against one player on the turn if we face bet
iv- we check back again facing no turn bet
v- we face turn bet 1/2 of the time
vi- AQ is still in their ranges after checking twice (we have around 38% on random turns that don't pair the board or give us a FD)
ii) EV of checking back:
-> (board pairs) + (call on FD and win) - (call on FD and lose) + (fold on brick) + (win when checking back on brick)
-> (7/45)(1) + (18/45)(2)(.39) - (18/45)(-1)(.61) + (20/45)(.5)(0) + (20/45)(.5)(1)(.38) = .80
I believe that my math is correct (if not please let me know).
If so and you still think that shoving is the play. Where are my assumptions out of line.
Oct. 10, 2016 | 5:59 p.m.
+1 that RFI by position is a good one to have on HUD
Does anyone use CC % (flat vs open) by position on HUD or does looking at the spread between VPIP and PFR give you all the information you want?
Oct. 6, 2016 | 10:44 a.m.
Interesting thought about check-calling:
- Let's assume BTN bets pot and it folds back to us.
- Assume also that his range is simply JT+
- Assume we over shove our bluff cards and he always folds worse than the nuts
How will it play out:
i) Nut outs (6)
ii) Bluffing cards (9) clubs and (4) Qo or 7o
iii) Strange cards - (2) Aces
iv) Blanks - 45 - 6 - 9 - 4 - 2 = 24
a) Folding = 486
b) Calling with a PLAN
-> (i * 6/45) + (ii * 13/45) + (iii * 2/45) + (iv * 24/45)
EV of cases:
i) 486 + (2 * 112.5) = 711
iii) Let's assume we win 75% of the pot -> (486-112.5) + (112.5 *3 * 0.75) = 627
iv) We check, he bets, we fold -> 486 - 112.5 = 374
-> (711 * 6/45) + (711 * 13/45) + (627 * 2/45) + (374 * 24/45) = 527
-> or +8bb
Wait I forgot to account for times when he has nuts when we bluff shove
- very low percentage of the time on Q or 7 he has straight (3%*)
- now the weighted sum equals:
-> (711 * 6/45) + (711 * 9/45) + (711 * 4/45 * 0.97) + (627 * 2/45) + (374 * 24/45) = 525 (negligable difference)
Looks like your plan of calling and bluffing has merit. However all the assumptions that I made are in your favor and starts calling with some non-nut hands on the turn when you bluff shove I would guess floating quickly becomes worse than folding.
Looking at hands that would make a 'good candidates' to call in his shoes. They are few and far between. For example on 3c turn he only has 6% flushes**
I like the float here vs the right opponent. High variance !!!
-* really closer to 4% but rounded down to account for the equity we still have when called
-** assuming his pfr is 20% and flop range is JT+
Oct. 5, 2016 | 6:31 p.m.
Think this is pretty math oriented spot assuming you don't have reads. Like you said the situation is static, because him betting this size is effectively all in and never folding. No future planning necessary.
Quick and dirty math:
- you are risking 67.5 to win 67.5 *2 + 42
- need 37%
- and vs a reasonable range of K5+, and AK+:ss you have 36%
- so I would say it don't matter much
- I you feel strongly that his range is much stronger: fold
Think calling is bad.
Oct. 4, 2016 | 5:24 p.m.
No real math or knowledge in this spot. But I agree with sorcloud for the most part.
Although I think it makes some sense to include your weaker trips (meaning hands like 8[7-][7-] in your ch range.
- Not getting it in crushing (might even be an underdog idk)
- If you cbet all of your trips (then when you check) on blank run outs opponent can use large bet sizes and include more bluffs in his range because he knows you always have a bluff catcher
- A lot of you weaker 8xxx hands will have cards that boat up when a straight completes, so that negates the need for protection a little bit.
Oct. 3, 2016 | 12:19 p.m.
Been wondering about this spot and would love to hear others thoughts.
Flatting opens OTB with Axxx + single suited to the A + (not having another component or having another component that is weak)
Examples: AQ96, A532, AT77, AK22 (ass single suited to the Ace)
While these hands are all slightly different, I think thinking about them as one category for the sake of simplicity might be a good idea. Could be wrong here tho.
And while the decision to flat or not to flat these hands on the button might be close and therefore a waste of time, they also occur at a high frequency, which might make the spot worth thinking about.
- 3betting is out of the question vs any reasonable opening range. We have to fold to 4b and the hand does not posses the qualities you want in a hand in a 3bp. It does better in higher spr situations because of the nut and nut draws it flops (however infrequently).
- Getting flatted behind cuts both in both directions. Positive is that our nut flush draws gain value. Negative is that facing a cbet a) it is tougher to float and b) calling with naked top pair and A high boards becomes less appealing
- I think facing a squeeze behind is not that bad of an outcome. While we will have to fold on the flop very often we have 2 things going for us. One, when we flop our NFD we have a very profitable situation. Two we have the best absolute and relative position which will allow us to pick up more than our fair share of pots post-flop.
- While we are a dog vs almost all opening ranges, we aren't that big of a dog. And having a hand that has a nutty component + position + deep stack-to-pot ratio is a good combination. However w/o much anything else going on we have to fold quite frequently when we don't flop a clear continuing hand (flush, nfd, 2p+).
- Think having a cold caller in-between the opener and us might be very important to the value of the play: a) our nut flush draw goes up in value and more importantly b) we get better information on the flop. Namely pfr cbets less frequently and more face up. More frequently we will get to see if our back-door flush-draw or our 2p comes in on the turn fo' free.
My hem numbers (way too small of a sample to mean much I think):
- of 8000 opportunities to flat on a raiser (w or w/o callers) OTB, 530 where hands like I am talking about (6-7%)
- I VPIPed on 300 of the 530 chances (50-60%); 55% vs only a raiser and 70% vs a raiser and caller(s)
- have a -55bb/100 loss rate; -70 vs only a raiser and +12 vs raiser and caller(s)
Aug. 11, 2016 | 6:29 p.m.
no math, but would bet quite a bit that it is a stack off
think it would help if you swapped the suits of your 9 and 5, to make combo-draws more likely.
thinking about hands the UTG might shove:
- you are in fine shape vs top two pair
- have ok equity vs sets
- are in very good shape (55-65) against some combo-draws
- and if he spazzed out with KKss you are in great shape
Aug. 11, 2016 | 5:10 p.m.
Is this a standard 3b for you guys?
- We are more than 100bb deep so the spr is a bit deeper (5.5), which is better for IP
- Hand doesn't mind if the BB flats behind, as we flop top sets and nut flush draws
- We block a K and therefore the more likely hand to 4b us lite- KK
- We get a very profitable spot on the off chance BB squeezes
Aug. 11, 2016 | 5:03 p.m.
I think I understand your question, but let me know if you meant something else.
- I do it this way because I find it to be the simplest way.
- There are only two outcomes when we get-it-in (lose or win) and we know what our stack will be when each event occurs.
- So to find out how often each one needs to happens (ie your equity needed), just set it equal to your stack size if you fold
Aug. 8, 2016 | 7:37 p.m.
assuming you only shove or fold to the 4bet (although flatting seems like reasonable option though, folding on flops where you don't hit a pair or a draw) :
folding = 188.46 - 9.50 => 178.96
5b/call = (188.46-107.5)(x) + (188.46+107.5)(1-x) = 178.96
where x is the % of the time you lose
x = .544
so you need 45.6% equity to stack off
assume he is 4betting linear
you have the right equity when he gets to 4betting wider than the top 11%
how wide should we assume he is 4betting?
think this might be a good situation to use Bayes theorem.
since the main hand you are worried about running into is AA (because you have 49% vs top 11% excluding AA and only 31% vs AA) we look at bayes theorem in reguards to AA.
- if he is opening 50% and you hold an A he has 3.5% AA hands prior to 4betting
- he has 4b 2/6 time
- we will assume he is always 4betting when he has AA
Probability of holding AA after opening then 4betting = (1 * .035) / .33
The 11% range that you need him to have to stack of profitably has AA 20 % of the time (23% when you don't block an A) almost double as often as we now assume he actually has AA.
Note: Villain having an observed 4b % of 18% would put him at the 20% AA threshold
So we should be alright to stack off !!!
July 31, 2016 | 3:15 a.m.
Thinking about RFI rules of thumb from the button.
Think the equation for opening our weakest hands looks like this:
-2.5(x) + 1.5(y) + ((pot size when called - average rake taken)(z) - cost of getting to the flop)(1-x-y)
where x is how often we get 3b (assuming we fold to all 3bets)
y is how often the steal works
z is the % of the pot we realize when called
Now with the bottom 15% of hands we have 40% equity vs top 65% hands minus a 10% 3b range.
If we assume that we realize that 40% (which seems reasonable because the factors of having a hand that does a poor job of realizing equity opposes the fact that we have position and the 'uncapped' preflop range) and that our opponent doesn't adjust to our strategy, we get the following break-even points:
(Cumulative 3b%/Steal-success rate)
12/39, 15/42, 18/45, 21/48, 24/50
Since the numbers seem to scale together we can come to the rule of thumb that:
If Steal Success Rate - Cumulative 3b% > 26 or 27, then opening any 4 cards assuming opponent is not adjusting is better than folding
Assumptions: 2.5x raise, only the bb calls (5.5bb pot), and .75bb average rake taken
Also interesting to think about is how the graph of % of hands we should open vs opponents overall defending % given a constant 3b% would look.
My guess is something like this (using 18% cumulative 3b (what it actually is in my hm database)):
Any thoughts appreciated
July 30, 2016 | 3:48 a.m.
Can you expand please.
Just so we are on the same page....
When I say top-down I think I mean something like: forming the best overall strategies from facts about the game like combinatorics, position, stacks etc.
When I think bottom-up I think looking at characteristics of a specific spot (say how often often opponent does xyz or equity when called), and then creating an overall strategy from what you find.
June 2, 2016 | 3:10 a.m.
So say we don't know what we should do on the flop in this spot. Does it help to take a look at what turns we want / should be continuing on ?
Agree that if opponent is betting turn very often and we can hardly ever continue that it makes the flop a fold.
June 1, 2016 | 7:23 p.m.
Just read the free parts of the book and heavily considering 'investing' in the rest of the it (part one and two)
I like the thought of (as he points out) approaching the game from a top down perspective (how a computer would 'think' about the game. I have always just copied what I have seen and done work from the bottom up approach- thinking about a specific hand / spot and then trying to generalize.
If anyone is interested buying the book and going through it together let me know (think it will take quite a bit of time to do it properly)
Tom (if you're still here), I have a question for you.
Your thoughts on if it might be better for me to skip right to the second volume in my situation (slight winner in small-midstakes (200-600 6m), playing professionally plo professionally for around 4 years, but not too well versed (outside of playing and watching vids) in combonatoric and textures?
June 1, 2016 | 5:17 p.m.
Thanks. I struggle with this area of my poker a lot.
A good book that it sounds like you might enjoy is Deep Work: by Cal Newport
May 30, 2016 | 2:54 p.m.
Like you said it is good that you have the T blocker.....
Say your hand didn't have the T blocker
- you are now getting:
assuming your equities remains the same (it will go down a bit in reality I think)
(1)(.47) + ((1+ 2/3 + 2/3)(.68) - 2/3)(.31) + ((1 + 3 + 3)(.26) -3)(.22)
* which is less than the estimated EV of checking back
May 26, 2016 | 8:44 p.m.
Set-up (from turn): stacks = 3, pot = 1
cb) check-back and realize X% of the pot
b-c) bet and either get a fold, get peeled or get big-johned on
cb: Equity on turn = 70% vs overall range
Say you realize 100% on pairing rivers (10/44), 20% on heart rivers (7/44), and 50% on las otras (27/44)[a]
=> (10/44)(1) + (7/44)(.2) + (24/44)(.5) = .53[b]
Folding: all hands worse than T7+, sets, 2p+OESD, 2p+NFD will fold = 51.5%
Peeling: assume the hands above that are not QT will peel your 2/3psb = 30.25%
- these hands you have good equity great equity against and are pushing a lot of value
- around 68%
Shoving: 18.25% of the time you will get in your stack with 26% equity
EV Folding (.515) + EV Peeling (.3025) + EV Shoving (.1825)
=> (1)(.515) + ((1+ 2/3 + 2/3)(.68)[c] - 2/3)(.3025) + ((1 + 3 + 3)(.26) -3)(.1825)
.58 > .53
so maybe betting is better
BB range pre = 60%!$3b8o
BB range for flatting flop = (hh:4+,T8+,AJ+,J:T7+)!(Ah:hh:J+,99+)
[a] Just guessing here
[b] Seems a bit low, but in the ballpark I think
[c] I am assuming you will win your equity on bet-call turns
May 26, 2016 | 8:35 p.m.
still thinking about doing it?
May 26, 2016 | 1:58 p.m.
anyone know of a hand replayer or converter that works with 5 card plo hands.
- Deuces cracked doesn't work
- Pokerhand.org sort of works, but has some bugs and doesnt show the 5th card in hands
May 25, 2016 | 7:18 p.m.
I think folding is too tight vs a complete unknown.
If you are folding this hand you will be folding over 50% and his lead is profitable with any 4 cards
And you have 2 pretty disguised nut outs, and are in position.
Going to be tough to play vs future aggression if you brick. But again vs until you have some reads I wouldn't fold
Nice video, ty.
Min 13, table 3, in 3way srp in the BB: "I don't think I can lead on 55T, but if it were 22T I think I should"
I thought I caught you making a false claim as I believed 5s and 2s were going to be about the same weighting in tight MP opening and SB cc ranges. But you are right, 5s are about twice as likely as 2s.
Would you still only lead 22T boards and not 55T boards if it were HU ?
Did the fact that the flop was rainbow play into you decision not to lead at all ?
Nov. 22, 2016 | 1:43 p.m.