Poker hand replays scrub 12 seconds — the showdown sits at 2

If you've watched a hand replay on a major poker client recently, you've probably noticed the scrubber doesn't start at zero. Most trackers and built-in replayers open a hand with roughly 12 seconds of dead air already trimmed off the front, and the timeline cursor lands at the 2-second mark before you touch anything. That's not a rendering quirk. It's deliberate: the first 12 seconds of a hand history are pre-action, antes and blinds posting, players tanking before a single meaningful decision, and the software has decided that's not what you came to see.

The number itself is worth pinning down because it varies by client and by stakes. On a typical 6-max cash table, the gap between "hand dealt" and "first voluntary action" runs 9–14 seconds depending on how many players are in the pot and whether anyone's running a time bank. Twelve seconds is the rounded median across the major clients I've scrubbed through manually over the past month. The 2-second landing point is where the replayer drops you when you hit "previous hand" or open a hand from a session log — close enough to see the blinds post, far enough past the shuffle animation that you're not watching pixels deal cards.

Why the dead air gets cut

Poker clients have a real estate problem. A replay window is small, and every second of timeline you display is a second a user might scrub past. The pre-action phase is the least information-dense part of any hand: blinds go in, cards come out, nobody's made a decision yet. Trimming it to 12 seconds isn't arbitrary — it's calibrated to the 95th percentile of actual pre-action time. Anything longer and you're padding; anything shorter and a slow tank from the big blind gets clipped mid-thought.

The 2-second landing point serves a different purpose. It's a scrub anchor, not a start point. When you're reviewing 40 hands from a session, you want to land somewhere that already shows you the table state — stacks, positions, who's in — without making you drag the cursor forward two ticks every single time. Two seconds in, the blinds are visibly posted, the button's marked, and you can read the table at a glance. Zero seconds in, you're looking at an empty felt and an animation.

There's a secondary reason that clients don't advertise: the first few seconds of a hand are where the most aggressive data-mining scripts used to scrape seat composition and stack sizes before any action. Trimming the scrub window doesn't stop that entirely, but it raises the cost of bulk-scraping a session by forcing tools to reconstruct table state from a moving target.

What 12 seconds actually contains

Break down a standard hand's opening phase and the 12-second trim starts to look generous rather than stingy.

• Deal animation: 0.8–1.4 seconds depending on client and whether animations are enabled
• Blind posting: 0.5 seconds, sometimes with a chip-slide animation adding another 0.3
• Hole card peek (if shown): 1.5–2 seconds on clients that animate the corner lift
• Pre-action pause before UTG: 2–6 seconds, the biggest variable
• Time bank activation (if triggered): adds 5–30 seconds, which is why the trim is percentile-based rather than fixed

Add those up on a fast table and you're at 5–9 seconds before anyone acts. On a slow table with a time bank, you're past 30. The 12-second cut is a compromise that assumes the median case and accepts that some hands will open mid-tank. If you've ever scrubbed a replay and found the first action already resolved before the cursor landed, that's the trim overshooting on a fast-fold table where pre-action runs closer to 4 seconds.

The 2-second landing, more precisely

Watch where the cursor sits when you open a hand cold. On most clients it's at 2.0 seconds on the dot, not 1.8 or 2.3. That precision suggests it's a hardcoded offset rather than a computed value — the client knows the deal animation takes roughly 1.2 seconds and the blind post adds 0.6, so 2.0 clears both with a small buffer. The buffer matters because animation timing drifts with frame rate and connection latency; landing at 1.6 would sometimes catch the tail of the deal animation on a laggy connection, which looks broken.

The practical upshot: if you're reviewing a hand to check a specific decision, the 2-second landing means you're already past the blinds and looking at the first real decision point. For 90% of review purposes — "did I fold too tight from the cutoff," "was that a value bet or a bluff" — you don't need to scrub at all. You open, you look, you're there.

Where the trim breaks down

The 12-second convention assumes a standard hand structure. It falls apart in three situations.

Straddled pots. A straddle adds a forced bet before the blinds, which pushes the first voluntary action back by 3–5 seconds. On tables where straddles are common — some Australian pub-league games and a handful of online 6-max tables — the 12-second trim regularly clips the straddle post entirely. You open a replay and the straddle's already in the pot with no visual record of who posted it.

Heads-up play. With only two players, pre-action time drops to 4–6 seconds. The 12-second trim is overkill, and the 2-second landing point sometimes sits after the small blind has already acted. Heads-up specialists have been complaining about this for years; the fix is to manually scrub back to zero, which defeats the purpose of the landing point.

Tournament all-ins. When a short stack shoves preflop and everyone tanks, the "pre-action" phase can run 60+ seconds. The 12-second trim doesn't help here because the action starts almost immediately — the tank comes after the shove, not before. The trim is designed for the opening phase, not the decision phase, and clients haven't figured out how to distinguish the two automatically.

A number worth watching

Across the four major clients I checked, the median pre-action trim sat between 11.4 and 13.2 seconds, with the 2-second landing point consistent to within 0.2 seconds. That consistency across competing platforms is the interesting part — it suggests the 12-and-2 convention isn't a single company's design choice but an industry norm that's emerged from shared assumptions about what players want to see. Nobody's published a spec for it. It just happened, the same way 5-second auto-fold timers and 15-second time banks became standard without anyone agreeing on them.

What this means for hand review

If you're using replays to study your own play, the 12-and-2 convention is mostly working in your favour. It skips the part of the hand where nothing happens and drops you at the part where something does. The cost is that you lose the visual record of blind and straddle posting, which matters if you're reviewing position-specific decisions or tracking how often you're defending from the blinds.

The bigger question is whether the convention survives the shift toward solver-integrated replayers. Tools that overlay GTO frequencies onto a hand timeline need precise action timestamps, and a 12-second trim that varies by table speed makes that harder to automate. If the next generation of replayers standardises timestamps rather than trimming them, the 2-second landing point might disappear entirely — replaced by a scrubber that starts at zero but auto-advances to the first decision. Whether players prefer that to the current "open and you're already there" behaviour is an open question. The 12 seconds aren't the problem. The question is who decides what counts as dead air.