On my Monday livestream I ran a single-channel neurofeedback session on my own head and then walked through the myths I hear most often after years in this field. People either think neurofeedback does nothing or think it does everything. Both miss the mechanisms. Here is what I covered, cleaned up into a reference you can actually use.
A note on attribution: the questions came from a live audience. I have stripped names and kept the substance.
What was I actually doing in the session?
I trained C4 referenced to the left ear (A1), a sensorimotor rhythm protocol at 12 to 15 Hz. C4 sits one inch up and back from the right ear, at the vertex split. I rewarded SMR (12 to 15 Hz) and inhibited two bands: slow activity at 4 to 7 Hz (theta) and fast activity at 22 to 34 Hz (high beta).
The reward logic is straightforward. When my SMR amplitude climbs above its threshold and my theta and high beta stay below theirs, the game advances and I get a beep. The software auto-thresholds every 30 seconds, so when I fatigue it makes the target easier, and when my brain moves fast it makes the target harder. We aim for roughly 65% reward on the up-train band and about 75% pass on the slow inhibit, because those rates tend to maximize learning.
C3 and C4 are the deadlift of neurofeedback. If you could only run one protocol, this central pair would still move executive function, sleep, and stress for most people. If you want the underlying mechanism, I wrote it up in SMR Neurofeedback: Train Sleep, Focus, and Self-Control.
Myth 1: Is neurofeedback just placebo?
The evidence against a placebo explanation is strong. SMR neurofeedback was discovered in cats at UCLA in the 1960s, where conditioning sensorimotor rhythm made the animals resistant to seizures (Sterman & Friar, 1972). Cats are terrible at following instructions, so a placebo response cannot account for a change in seizure threshold.
Placebo effects also fade. Neurofeedback effects hold for the regulatory targets across months and years. And I ran one of the first double-blind placebo-controlled neurofeedback studies myself, in the mid-2000s, as part of my doctoral work. You can find it under my name in the open-access dissertation archive.
Here is the design. I trained C3 and C4 with both real and sham (non-contingent) feedback while recording 64 channels of DC-coupled EEG with a BioSemi cap. I anchored each reward event into the ongoing EEG and averaged a couple thousand of those two-second clips into an event-related potential. Anything phase-locked to the beep survives the averaging; everything else washes out.
The brain produced a burst of activity in the exact frequency I was rewarding, at the exact site I was training. Reward SMR on the right, you see an SMR bump on the right. Reward beta on the left, you see a beta bump on the left. Run sham and the burst disappears. That burst is an event-related spectral perturbation, and it shows up inside the first five minutes of training, well before the person consciously feels anything (which usually takes three to five sessions). The brain binds to the contingency long before the mind notices.
Myth 2: Does neurofeedback work the same for everyone, and for everything?
It works well for almost everyone aimed at the right goals, but it resembles personal training more than medicine. You aim at goals and elicit effects, then iterate session to session. A bicep has predictable machinery. A brain reacts variably, so two people with similar maps and similar goals can respond differently to the same workout.
The reliably movable targets are the regulatory features: executive function, sleep, stress and anxiety, speed of processing, social processing, sensory integration, and stability phenomena like seizure and migraine suppression. The research on ADHD, trauma, OCD, alcohol cravings, PTSD, and sensory and social anxiety is reasonably consistent (Arns et al., 2009). Tinnitus is the inconsistent one; when it moves it moves a lot, but it only moves about half the time.
Some tissue locks down after a critical period. If you are autistic with no productive language, or you lost language to a stroke, that tissue is hard to change. Possible for some people, but case by case. This is why I steer people away from one-size-fits-all headsets if they are actually suffering or chasing real performance goals. Get a QEEG brain map and build tailored protocols you can iterate over a few months.
Myth 3: Does neurofeedback zap or stimulate the brain?
Almost no form sends electricity in. Whether you are running SMR, slow cortical potentials, infra-slow, or standard band training, you are measuring the electricity leaving your head, doing math on it, and sending information back through your eyes and ears. A few outliers exist (LENS may involve micro-current, and EMF systems like Neurofield apply fields), but standard neurofeedback is a passive measurement that applauds your brain through the senses.
Myth 4: Does neurofeedback have no side effects?
It absolutely can, and the side effects are your best teacher. Run the wrong protocol and you will feel wired, wiped out, or off, and your sleep that night can shift.
One audience member described a motivation protocol that left him "driven but irritable." That points to fast beta pushing too hard, or executive function getting trained up when the person needed calm and reserve instead. I usually run motivation around FP1, the tip of the left frontal pole near the approach-oriented dorsolateral prefrontal cortex. My old mentor Larry called FP1 the desk-clearing protocol. It can light a fire under someone, and if it lands hard you feel wired.
Temporal lobes are the finicky ones. Train them a little too slow and you downregulate emotional stability, which produces reactivity, a jagged, almost bipolar-feeling response. Effects start transient, so you have freedom to experiment, but do not push through side effects. The MLM-style sellers tell you to keep going. Bad advice. You can build a transient side effect into something semi-permanent in five or ten sessions. Watch how you feel, then iterate. For the anxiety end of this, see Neurofeedback for Anxiety: What the Research Shows.
Myth 5: Does QEEG mean you train the brain toward "average"?
The map exists to find the features that matter to you, and average is nowhere in that goal. People are weird, and weird is often where the gifts live.
The QEEG shows conserved circuits: the vigilance and stabilization functions on the left, the supervisor on the right, hyperfocus in the front midline, threat assessment in the back midline. When those common circuits look cramped in a particular way, that gives us a fast read on what might be driving your stress, fatigue, or distractibility. You then train toward goals and symptom reduction. Sometimes that lands you near typical, sometimes nowhere near it.
All of this is nonlinear. Push one feature and others shift, which feeds back on the thing you pushed. The "linear versus nonlinear" labels in this field are mostly branding. Every neurofeedback system is training a nonlinear system. More on reading these features in Biohacking with EEG Phenotypes.
Myth 6: Do effects wear off, or are they permanent right away?
Both halves miss what actually happens. Effects are transient until they consolidate.
ADHD, PTSD, and alcohol cravings need roughly 30 to 40 sessions. With ADHD, the research and the brain maps point to substantial gains in executive function over about three months, enough to get ADHD out of the way of your life (Arns et al., 2009). Once you reach that point the change tends to be durable. Before that, individual sessions wash in and out.
Myth 7: Is neurofeedback just for ADHD?
There is a lot of executive-function work, but the field was born from seizure control (Sterman & Egner, 2006). There is solid research on epilepsy, ADHD, creativity, trauma, and other anxiety presentations. There is even work on improving surgical skills (Ros et al., 2009). Most of it sits at the case-study level, partly because blinding EEG in a placebo context only became practical recently. For attention specifically, see Does Neurofeedback Work for ADHD?.
Myth 8: Is neurofeedback new or unproven?
EEG was discovered around 1924 (Berger, 1929). Neurofeedback followed in the mid-1960s, with Joe Kamiya at UC Berkeley working on alpha (Kamiya, 1968) and Barry Sterman at UCLA discovering that conditioning SMR made brains seizure-resistant (Sterman & Friar, 1972). That is roughly half the lifespan of EEG itself.
One audience member reported faster resting alpha and improved verbal fluency after training. Individual alpha frequency tracks the handoff between brain regions. The research links faster individual alpha frequency with cognitive performance and processing speed (Klimesch, 1999). IQ is a construct built from speed of processing, working memory, and implicit learning, and alpha speed indexes the speed-of-processing piece directly. In some ways it is a cleaner physiological read than the test score. More on this in Decoding Alpha Waves and on the aging side in The Critical Aging Window.
If you want the literature, our website organizes about a thousand papers by topic, and you can dump the full abstract library to one page and run a text search across it.
Myth 9: Are you controlling the game with concentration?
Conscious effort is not the mechanism here. You cannot feel your brain waves, so you cannot deliberately steer them. Neurofeedback runs through operant conditioning below awareness. The game advances when your brain produces the target pattern and dims when it drifts. Your mind discovers, over the session, that it does not actually do anything.
This is why it works in a coma, works on a nonverbal teenager staring at his phone, works while you are asleep. The late Margaret Ayers built a practice training people at coma bedsides and changed their brain activity that way. Boredom is irrelevant. The brain prefers information over no information, so even feedback you dislike still drives learning. Once you can feel your sessions (usually after four or five), you can do non-social tasks like paperwork during training. Avoid socially engaging feedback. Social processing breaks implicit learning, which is exactly the mechanism doing the work. That is also why I am not a fan of video-based feedback for most people.
Myth 10: Can a QEEG diagnose you?
The map shows your patterns of delta, theta, alpha, and beta, your connectivity, and your resting features. Those are real, but they do not map cleanly onto diagnostic labels. You can see signatures congruent with ADHD or anxiety, but the resolution is about resources, not diagnoses. Done well, brain mapping surfaces things that already make sense to you given your performance, your suffering, and your goals.
Bonus Myth 11: Does neurofeedback change your personality?
It gives you more control over yourself without erasing the self. If your front midline runs hot, you probably carry some OCD traits, and training that activity down does not leave you sedated on a mountaintop. You keep the ability to hyperfocus and choose when to use it. Train up low beta and down theta in severe ADHD, and the person gains the ability to sit through a boring class or shift from a video game to a quiet dinner without it feeling like an amputation, while keeping the fast pattern-matching and hyperfocus for the game or the sport. The double-edged traits stay; you get the dimmer switch. See Biohacking with EEG Phenotypes and Biohacking OCD for how those circuits behave.
How to start with neurofeedback
If you are dealing with real symptoms or chasing serious performance goals, skip the consumer headset and get a QEEG, then build protocols you can iterate over a few months. If you want to see the software in action, the EEGer demo is downloadable from support.eeger.com on a Windows machine; you cannot run live sessions without proper supervision and a license, but you can examine the features. Most of the people I coach train from home with software, an amplifier, and coaching on how to place wires, find locations, and read their own after-effects on sleep, stress, and mood. That self-monitoring habit carries its own benefit. What gets measured gets managed.