Claims are cheap to make and expensive to refute. That gap is the whole engine. The Gish gallop weaponizes it: bury the other side in so many assertions that answering even half of them would burn the entire clock, then point at the unanswered pile as if silence were agreement.
It feels like winning in the moment. It usually isn't. The same speed that makes a gallop hard to answer live also makes it fragile on the replay. This is a playbook for handling it in four settings: a live debate, a panel or podcast, a comment thread, and the recording afterward. It closes with a harder question, which is whether the person galloping is you.
Where the term comes from
The name traces to Duane Gish, a creationist debater known for firing off a rapid stream of claims that his opponents could not address in the time allowed. The observation outlived the debates: the tactic works on structure, not subject. Anyone can do it about anything.
Why it works even when the claims are weak
A gallop exploits three things at once.
The first is cognitive load. An audience can track one argument and follow whether it holds. It cannot evaluate twenty claims delivered in ninety seconds. Past a certain rate, listeners stop checking each point and start judging the overall impression, which is exactly the swap the galloper wants.
The second is the asymmetry of effort. A false or sloppy claim takes four seconds to assert. A real rebuttal takes a minute: define the terms, cite the counter-evidence, explain why it matters. Multiply that by fifteen claims and the responder runs out of clock long before the list runs out.
The third is confidence read as substance. A confident person rattling off many points looks like they have command of the topic. A careful person refuting one point thoroughly can look slow, even evasive, as if they only had an answer for the one thing they chose. Format makes this worse. Short rebuttal windows, timed debate segments, and the optics of being visibly outnumbered all reward volume over accuracy.
None of this makes the claims true. It just makes them costly to clear away in real time. The counters below each attack a different part of that machinery.
Four counters that work
Pick the two strongest and bury them
This is the move for a live debate with a clock. Do not try to answer all twenty. You will fail, and failing in public is the outcome the gallop is built to produce.
Pick the two claims that are most central to their case or most damaging if left standing. Refute those two completely. Then name what you are doing and why:
I could spend the rest of this segment fact-checking the list, but most of those points are either tangential or fail for the same reason the first one did. So let's stay on the two that actually carry the argument.
You have now reframed the contest. It is no longer "answer everything or lose." It is "did the load-bearing claims survive," and you have shown that they didn't. The audience does not need all twenty addressed. They need to see that the important ones collapsed.
Name the move out loud
This works on a panel or a podcast, where there is a moderator or at least a shared sense of fair play. Describe the tactic plainly, without heat:
You've made about fifteen claims in the last minute, and several are false. If I pick one, it looks like I'm dodging the other fourteen. But nobody in this room can actually evaluate fifteen claims at this speed, and you know that. Let's take them one at a time.
Naming the gallop does two jobs. It tells the audience why a single careful answer is the honest response, not a dodge. And it puts a small cost on continuing, because galloping after someone has labeled it looks like exactly what it is. You are not asking permission. You are setting the terms.
Hold the line on one claim
In a comment thread, the clock disappears. That changes everything. You are not racing anyone, so do not act like it.
Pick one claim. The clearest, most checkable one. Ask for a real defense of that single claim and decline to move until you get it. When they reply with another ten points, restate the one:
Happy to get to the rest. First, the thing you said about X. Here's why it's wrong: Y. Do you have a response to that specific point, or not?
Repeat as needed. This feels almost rude in its persistence, which is why it works. A gallop depends on forward motion. Refuse to move and the whole tactic stalls. In transcript after transcript, the side that held the line on one concrete claim reads as the side that won, because there is a visible question on the table that the other side kept running from.
Watch for two escape hatches while you hold. One is the straw man, where the claim they finally defend is not the one you challenged. The other is the motte and bailey, where a bold claim quietly shrinks into a modest one the moment you press it. Both are signs the line is working.
Win on the rewatch
This one works anywhere, because it does not happen in the moment at all. If you cannot refute a gallop live, refute it afterward, on the record, with timestamps.
At 14:32 he claimed X. X is false, because Y. At 14:47 he claimed Z. Z is false, because W.
Live, the galloper had speed. On the recording, you have something better: time, and a reader who can scroll back and check. The viewer who left the live event uncertain comes back to a clean, itemized teardown and changes their mind at their own pace. The gallop is built for the live moment and defenseless against the replay. The replay is where you live.
What makes it worse
Three responses feel natural and all of them lose.
Trying to answer everything is the trap the gallop is designed around. You will run out of time, leave claims standing, and look like you fell short. Volume is their game. Do not play it.
Getting visibly frustrated hands them the frame. The galloper wants to look calm and in command while you look rattled. Annoyance reads as losing even when you are right. Stay flat.
Reciprocating with your own gallop is the worst option. Answer their flood with a flood and you have abandoned the one advantage you had, which was looking like the reasonable person in the room. Now you have both lost, and the audience has stopped listening to either of you. This is also a fast way to slide into whataboutism, where the argument becomes a contest of accusations instead of a question with an answer.
When the galloper is you
Here is the uncomfortable part. A gallop does not always come from bad faith. Sometimes it comes from a quiet worry that no single point would survive on its own, so you reach for volume as cover.
The diagnostic is one honest question. Are you making many claims because you genuinely have many strong points, or because you suspect any one of them, examined closely, might not hold? If a single claim were the entire argument, would you be comfortable defending just that? If the answer makes you flinch, you may be galloping.
The fix is to subtract. Pick your two or three strongest claims and defend each one thoroughly. A small number of well-supported points is more persuasive than a flood of half-supported ones, and it is far more durable. The flood wins the live minute. The two solid points win the transcript, the screenshot, and the reader who comes back tomorrow.
This is the same lesson the responder learns, seen from the other chair. The skill in any argument is calibrating to how it will read later, not how it feels right now. A move that looks like dominance in the moment often looks like flailing on the rewatch.
A gallop is a bet that nobody will go back and check. Make them want to. (For more on naming a pattern without turning it into a brawl, see using fallacy.is in a discussion.)