Search filters cut 80% of lobbies—the table list hides 3

Search filters cut roughly 80% of the lobbies you'd otherwise see, and on a typical Australian-facing site that's hiding three tables you'd probably actually want. Not because they're bad—because the filter logic treats them as noise. If you've ever set "Hold'em, 6-max, stakes under $2/$4" and watched a list of 40 tables collapse to six, you've met the problem. The three that vanish are the ones that don't fit a clean checkbox, and they're often the softest seats on the page.

What the filter is actually doing

Most lobby filters aren't one search. They're a stack of conditions applied in sequence, and each one is a hard gate rather than a ranking. You pick a game type, a stake band, a table size, maybe a speed preference, and the lobby removes anything that fails any single condition before it shows you the rest.

The 80% figure is a rough but useful rule of thumb. Start with a room showing 50 tables at peak. Apply a game-type filter and you're down to maybe 30. Add a stake band and you're at 18. Add table size and speed and you're looking at 9 or 10. That's an 80% cut from a filter set most players would call reasonable.

The catch is that the filter has no idea which of those 40 removed tables were worth your time. It's matching strings and numbers, not reading the room. A table that's technically outside your stake band but has two recreational players bleeding chips looks identical to a table full of regs—both get deleted.

The three that tend to disappear

Across most lobbies, the tables that fall through are consistent:

1. Tables just outside your stake band. You set $0.50/$1 to $1/$2. A $2/$4 table with a short-handed lineup and a deep-stacked fish doesn't qualify. Neither does a $0.25/$0.50 table where the average pot is running 40% above normal. The filter reads the blind level, not the table quality.
2. Tables mid-transition. Someone just left a 6-max table and it's now 5-handed. If your filter wants exactly six seats filled, that table is gone until someone sits—by which point you've moved on. Same for a full-ring table that's temporarily short.
3. Tables with non-standard labels. Fast-fold, "beginner-friendly," "deep stack," or any room-specific tag often sits in a separate category the filter treats as mutually exclusive with your main search. You asked for Hold'em; the filter decided you meant standard Hold'em.

Why the lobby design does this

There's a reason the filter is aggressive, and it's not malice. Lobbies are built to reduce decision fatigue. A list of 50 tables with 12 columns is unusable on a phone, and most players genuinely do want the list cut down. The filter is doing what it was asked.

The problem is that it's optimising for a clean interface rather than for finding you a game. Those two goals overlap about 80% of the time. The other 20% is where the money is—or isn't.

There's also a commercial angle worth naming. Rooms make more from rake than from any single player, and a filter that pushes you toward the busiest tables keeps games running. A quiet $2/$4 table with three players might be the best value on the site, but it doesn't fill seats or generate the same rake velocity as the busy $1/$2 pool. The filter isn't rigged, but it's not neutral either.

The stake-band trap in numbers

Here's a concrete example. Say a room runs 12 no-limit Hold'em tables at 8pm AEST on a Tuesday:

• 3 tables at $0.25/$0.50, average pot $18
• 5 tables at $0.50/$1, average pot $34
• 3 tables at $1/$2, average pot $61
• 1 table at $2/$4, average pot $140, currently 4-handed

You set your filter to $0.50/$1 to $1/$2 because that's your comfortable range. You see five tables. The $2/$4 table—four-handed, deep stacks, average pot more than double the $1/$2 tables—is invisible. So are the three micro tables, one of which has an average pot of $18 against a $0.50 big blind, which tells you the table is playing far bigger than its stakes suggest.

You've filtered out the highest-value table on the site because it didn't match a blind level you typed into a box.

What to do about it

The fix isn't to abandon filters—it's to use them as a first pass and then deliberately widen. A few practical habits:

Search by stake band, then manually scan one level up and one level down. If your band is $0.50/$1 to $1/$2, glance at $0.25/$0.50 and $2/$4 before you sit. You're looking for average pot size relative to the big blind, not the blind level itself. A $0.25/$0.50 table with an average pot of 36 big blinds is playing like a $1/$2 table.

Turn off the table-size filter and sort by players-per-flop instead. Most lobbies let you sort by a stat that actually reflects action. Players-per-flop above 40% is a soft table regardless of whether it's 5-handed or 6-handed. The seat count is a proxy; the stat is the thing.

Check the "other" or "special" tab once per session. Fast-fold, deep-stack, and beginner tables often live there and are excluded from your main search by default. Two minutes of scrolling beats an hour of waiting for the perfect table to appear in a filtered list.

Watch what disappears when you tighten. If you add a filter and the list drops from 20 to 4, you've probably cut something worth seeing. Loosen one condition at a time and note what comes back.

A note on the numbers you're trusting

Average pot size is the most useful column in the lobby and the most misread. It's a rolling figure, and on a table that just lost its fish it lags reality by several minutes. A table showing a $140 average pot might be four regs who've been passing chips back and forth since the last recreational player left. The stat tells you what happened, not what's happening.

The same caution applies to players-per-flop. It's a decent softness proxy over a few hundred hands, but on a table that's been running 20 minutes it's noise. Combine it with a quick look at stack sizes—short stacks and deep stacks mixed together usually means a recreational player is present—and you've got a better read than any single filter column gives you.

The question worth asking

The uncomfortable implication is that the lobby is designed to show you the games the room wants you to see, not the games that are best for you, and the two lists overlap but aren't the same. Filters are a convenience that quietly makes decisions on your behalf, and most players never check what got cut.

So the open question isn't whether to use filters—you should—but whether you trust a list you didn't build. Next time you open a lobby, widen the search by one stake level and one table-size setting before you sit. Count what appears that wasn't there before. If it's three tables, the filter was doing its job. If it's more, you've been playing a shorter list than the room was offering, and the best seat on the site might have been one checkbox away the whole time.