One of the biggest differences between theoretical poker and the poker you see at low and micro stakes is that theory contains far more lines than most aspiring players ever consider. It’s not that these lines are too complicated for someone playing 25NL or 50NL to understand. They simply aren’t on their radar.
So today, let’s put more tools in your toolbox. Let’s look at theoretically standard lines that are absent from most 25NL six-max games online. By getting these into your game, you can move toward the next theoretical level in your poker career.
1. Polar 3-Bets with BB vs SB
You might already be three-betting polarized to some extent. Maybe you find the obvious value hands in this range, and you add in hands like 6♠5♠ as natural bluffs. But are you three-betting , , , suited, or suited? And do you understand why you could?
I’m not saying if you copy these frequencies that some magical formula will let you climb stakes instantly. That’s not how it works. But three-betting is very lucrative in a game like 25NL because most small blind opponents are not skilled at playing three-bet pots out of position. They don’t defend enough by four-betting, and you gain a big edge by three-betting here from the big blind.
Not only do you pick up fold equity preflop, but you also face opponents who rarely find the right counter lines postflop. You can bet, stay aggressive, and often run people over in this spot.
At Carrot Corner, when we teach theory, it’s not to turn you into a robot copying solvers. We want you to understand the baseline so you can exploit better. And here, with fold equity upwards of 60% in theory, it’s no surprise that a hand like off, which is a marginal call, becomes a profitable three-bet.
As long as you’re not facing an opponent who four-bets aggressively, three-betting becomes the highest EV action. If you’re at 25NL in the big blind and an unknown opens, hands like off are clear three-bets. Solvers say you can fold, call, or three-bet—but in practice, three-betting is often best.
The regs who climb stakes learn to do this against weaker players and recreationals. The players who get stuck at 10NL or 25NL often avoid these hands because they “don’t look good enough.” They fall into the better hand fallacy—comparing their holding to one that feels more comfortable. This is a silent killer for progress.
Instead, focus on where the EV comes from. If a solver can make these lines work against another solver with no skill edge, you can absolutely make them work against your pool. Even if a hand is close to break-even, remember rake exists, and population mistakes will swing EV in your favor.
Money flows toward the in-position aggressor in these spots—both in theory and even more in practice. So don’t shy away from three-betting hands like off here.
2. Non-Intuitive C-Bet Raises
Now let’s node lock this spot. Say the small blind opens, the big blind defends, and the flop comes .
Instead of assuming the small blind c-bets 60% at one-third pot, we’ll adjust based on population tendencies. At low and micro stakes, many players simply bet too often here. Recreational players also overinvest in this node compared to GTO.
If you don’t punish this, they’ll get away with it. Against an average 25NL opponent, the EV loss of range-betting may look small—but you want to be the wolf in sheep’s clothing, not the sheep.
So the counterstrategy is counterintuitive raising. Let the small blind bet, then design the big blind’s defense strategy around more aggressive raising.
For this example, I plugged in a slightly smaller raise size to allow more raising. Against a range bet here, the raise frequency jumps to 17%. That’s a lot.
Are you actually raising 17% in this spot? Probably not. You’re likely raising with obvious value hands and some bluffs like or . But are you finding raises with hands like suited with a backdoor?
If not, ask yourself what your mental process is with a hand like that in this spot. Is it simply: “bad hand, fold”? Is it something like “I’d rather have X”? That’s a lot of our thought processes when we’re in a humdrum spot like this. But we have to wake up. We have to realize that our opponents are not able to defend the strategy they’re using out of position in single-raised pots with very wide ranges. We have to go about punishing that—not just by mimicking a solver and raising 17% of the time in this spot, but by actually dialing it up further and raising more often.
3. Low Equity Second Barrels
Let’s move over to Button versus Big Blind. The button opens, the big blind calls, and the button c-bets small on . That’s the sizing most of you will pick. Big blind defends, and the turn comes .
This is a pretty good turn for the button’s range. A lot of the big blind’s Qx combos get shed on the flop. This is going to be a high-EV spot. We have an EV of 4.05 chips in a pot of 7.38—so a healthy majority of the pot. And you can see the solver has built a 2x pot bet here.
You probably don’t do that because 2x pot is an AI preference. It’s actually something we built into the new course I’m working on—but more about that later. You probably go B150 or B75 here. I think B150 is a bit better.
This is a spot where you have a big enough nut advantage that the hands you bet for value are very strong. The solver range is significantly ahead of the big blind’s continuing range. Nut advantage is clear.
Now there are some very obvious bluffs you probably bet every single time that solver doesn’t like—for example, T9. Solver looks at T9 and thinks, “I can bluff, I can check back, I can realize equity, I won’t get raised often.” Checking is okay, betting is okay. But solver prefers some counterintuitive bluffs.
For example, . That’s just an over-card to the board. But it’s a pair draw, it has low showdown value, and it unblocks the folding range. Many Ax folded on the flop. The four also unblocks folds, and you still have the ace as an out. This is a completely normal bluff—even for an overbet.
You’re probably missing this because it’s counterintuitive. Humans prefer putting money in the pot with equity. But that’s a trap. The correct mental process is not “I’d rather have hand Y than hand X.” The correct process is to compare the EV of options A and B with the hand you do have. Leave hand Y out of it.
When you’re betting the turn with an overbet, or facing an overbet, population tends to be too equity-driven. They don’t include enough low- or no-equity bluffs.
Another example: 5♦4♦. Yes, it’s a straight draw, but more importantly the five and four are good unblockers because villain rarely reaches the turn with those ranks.
So in this spot, almost anything with a little equity or more can go B150. Find these hands in sims. Understand that you’re generating fold equity in a favorable environment. Realize the ace is an out. It’s not just air. And get more of those low-EV turn barrels into your strategy.
4. Turn Overbet Probe
This is a 3x pot probe. The AI actually chose this itself. I just said: come up with a sizing for when the same Button versus Big Blind texture goes check-check on the flop of . And the AI decided that 3x pot was a good sizing to lead on the turn , after the big blind checked the flop and the button checked back.
So big blind here is going hyper-polar. They’re leading out with a lot of stuff. And it makes perfect sense that they would want to do that because this time, when you look at ranges, it’s the big blind range that’s the leader here. They’re the leading player.
So yeah, when you look at the sim again, it’s obvious when you see it, right, that your value hands want to shovel a lot of money into the pot. You have a nut advantage, but not a range advantage. You’re leading here basically for the sake of the pot growing when you have the nuts, and you’re building your bluffs around that. Value hands are driving the bus. The bluffs are the passengers. That’s how I like to think about it.
So what are you finding here? You might be finding T9, you might be finding 96 open-enders, flush draws, etc. What you’re not finding are hands like A5 or A6—these lower equity ones. But as soon as it’s a big card out there—like a king, a queen, or an ace—you’re just going to want to build around a more nutted range, and that involves finding the bluffs and finding the value.
So yeah, turn probing—very important thing. Get the overbets into your game. This is a good way to get acquainted with overbetting in a very safe and natural spot for it, where it can’t really be wrong as long as you do it with the right hands.
5. Huge Check-Raises of Delayed C-Bets
So we can actually take the very same spot here ( ), but this time say that we went for a check and Button went for something like B75, which is a normal sizing to use after flop has gone check-check. They’re a bit more capped now. They don’t have an absolute nut advantage anymore. They do still have a lot of EV and they want to do a lot of betting, but we still have the nut advantage here—as long as we build our range in a protected way.
We can slow-play nutted hands here. So you’ll notice that we do get here with a bet of and we check-raise that hand. And when we check-raise here, we go absolutely huge, because let’s face it—we have a lot of lost ground to make up. The pot has remained quite small. Now that we’re reopening out of position, we’re handing it back to Villain to do some three-betting if they want to. They’ll do very little, but we don’t really want to be doing that thinly.
We also don’t have tons of nutted hands in our range. We want to check-raise pretty sparingly here, and we want to do it really nutted and with some decent equity semi-bluffs. The standards now for what we need to bluff are higher. You’ll notice that we don’t do this with anymore or anything like that—or , or a naked gutshot. We’re more around the sort of combo-draw, pair-plus-draw, bottom-pair-and-redraw-plus-draw vicinity.
So these hands need to do that. They need to make up for that lost ground. Get that giant check-raise into your game. You do have to be a bit fussier about the blockers and the equity you do it with now. But yeah, really cool tool to throw in there.
6. Donkbets on the Turn
Go for it. Sometimes the turn card is so stark—it changes the texture to such an extent—that you want to go ahead and build a donk bet.
Right, so let’s say that Button bet the flop, big blind called. What do you think is a turn card that’s going to mean you should build some donk bets?
So if we put in the now (I’m just guessing here, I haven’t simmed this yet) and we give them a smaller sizing than, like, a giant sicko overbet donk, we should see some amount of wanting to do this. Yeah—27% of the time once you plug in a small sizing.
So donking the turn is nice. It’s a really nice thing to do because it’s a very good way to say, “I know more than you about poker. I studied this node. Do you know what’s up here? I do, ‘cause I looked at solver.”
Very good way to say that to your opponent.
And there are many hands here that you can donk bet that you wouldn’t think, because a small donk bet is not actually that polar in action. So it’s not just that you’re building around the eight that you’ve just hit. This is not like turn probe where you’re building around the eight. You’re actually just trying to stop it going check-check in what’s a very high-EV world for you.
You become the leading player again in this spot. And therefore you lead, and you make sure that the pot doesn’t stagnate at an overwhelmingly high frequency. And that’s good for you.
So, find some 10x, find some 8x. You can even go a bit mergey here, and you can lead nine-seven of hearts. Like, the seven with the flush draw is creeping in there. Then you find your usual array of bluffs. Anything you floated flop with—like —still has a pair-draw right above second pair. That can bluff here.
You don’t need to be fussy in this node because you’re the player with the equity advantage. You don’t have to be ridiculously handcuffed. Sure, there will be some standards—it’s not an overwhelmingly amazing node for you, and you are out of position. You’ll need to watch what you lead here to some extent. You can probably find examples of hands that just aren’t allowed to do it ever, because they’re too miserable. But it won’t be that many. Most stuff with anything going on—especially flush draws—can lead.
So donk leading is great. It prevents people from checking back in spots where they’re meant to check back quite often. Because if you check here, from Button’s perspective? Okay—again, solver wants to go nuts with some huge sizing. The AI is obsessed with building around 2x pot in this spot. But let’s drag it back to Earth. If we give Button something human, like B75, solver says it’s meant to bet about 40% of the time.
In reality, you’ll never face that. As big blind against Button, you probably face a bet a third of the time, maybe a quarter, and it’s equity-driven and under-bluffed. So don’t let the turn go check-check.
7. River Check-Raises
This doesn’t happen at the micros. It feels counterintuitive and unnatural. But let’s stick with the same action sequence: big blind check-calls Button’s c-bet on the flop , turn is the , and it goes check-check. Now on the river , when you have a flush, you’re antsy, right? Pot is small, you’re itching to get money in. Most people just snap bet or overbet, and Villain folds about ten billion percent of the time. It feels awful.
But there’s another option. You can check-raise.
Take . It’s a great example because it blocks some of Button’s natural bluff-catchers (K7, Q7). Those 7x hands often check back if you check. So by holding the , you’re blocking the check-back range and unblocking the betting range. That’s what we call a polarizing blocker.
For instance, . Amazing check-raise. Why? It has a massive investment ceiling. The pot can balloon. You’re rarely re-raised here because when you hold the four of hearts, Villain can’t have the ace of hearts to bluff-raise, nor can they have lower flushes like to value-raise. But they’ll still value-bet worse hands, and they’ll still bluff smaller with non-nut blockers. That’s why ace-four of hearts is a pattern hand for this spot: a reminder that slow-playing through the check-raise can be more profitable than fast-playing.
8. Huge River Overbets
This time, let’s flip it around. Button c-bets flop on , big blind calls. Turn is the . Button overbets turn—2x pot—and gets called. River is the .
At this point, our value range is basically the whole nutted region. We polarized so heavily on the turn that by the river, we’re not really value-betting many hands that aren’t super nutted. The one exception is 87 suited, which lives on its own little island: too strong to check back, not quite strong enough for the shove.
Everything else is just, you know, too thin or is going to be an all-in. So when you have a hand like king-eight here, the highest EV play at equilibrium is to go all-in.
In practice, I think it’s just a really good thing to do because you can talk yourself into: “I won’t get called,” or “they’re going to overfold,” or whatever. And, you know, that might be true. But you are relying to some extent on being raised when you go smaller, right? When you go smaller, there’s some frequency where they’re meant to bluff-jam or thin-value jam against your small bet. And that doesn’t really happen so much. So putting it all in for value yourself is an important thing to be able to do.
And if it is the case that people are overfolding, and that puts you off, let me just entertain this theory: I’m not saying it is, but if it were the case that your pool was really overfolding here, and you don’t overbet for value with king-eight—but then you also say, “I don’t bluff overbet because they’ll just call me”—it’s very possible to be temporarily insane in poker and to make a polar opposite argument when you have it compared to when you don’t have it.
