Why order matters: every question is worth one point
Here is the fact the whole strategy hangs on. On the Reading & Writing section, a brutal cross-text inference question and a five-second transition question are worth exactly the same: one point. The scoring does not weight difficulty; it counts how many you got right, not which ones. And there is no penalty for guessing, so a blank is never anything but a strictly worse guess.
Put those two facts together and the four-minute question looks very different. Sink four minutes into one hard item while three answerable questions go untouched at the buzzer, and you have traded three near-certain points for a shot at one. That is the time-sink leak in its purest form, and it almost never shows up where it happens. It surfaces downstream, disguised as "careless" misses on questions you actually knew, which is why so many students never trace their wasted points back to the cause.
The section gives you two 32-minute R&W modules of 27 questions each, a little over a minute per question on average. You do not have to hit that on every item. You do have to stop any single item from stealing the budget of the next three.
Pass 1: the 90-Second Flag
Pass 1 is a sorting pass, not a solving pass. You move through the module in order with one job: separate the questions that are giving you an answer from the questions that are giving you a fight.
The rule is the 90-Second Flag. If a question clicks, answer it, commit, and move on. If a question is still not resolving at around ninety seconds, you stop, flag it, guess, and go. Ninety seconds is a trigger, not a stopwatch you stare at. The point is a ceiling on how long any one question is allowed to hold you hostage before you park it and keep banking points.
What counts as "not resolving" is specific. It is the question where you understand the passage but the choices still look interchangeable, or where you keep re-reading the same line and it will not crack. A long, calm question that produces a confident answer is not a stall, so do not flag it just because it took a while. The stall is time spent without payoff. That distinction is the difference between careful work and the leak, and it is the same line the behavioral-leak engine uses to separate genuine stubbornness from a student who is simply thorough.
Flag-and-guess mechanics
Flagging is two moves in one motion, and the order matters: guess first, then flag.
Fill in a real bubble before you mark the question for review. Never flag a blank. The guess is your insurance, because there is no penalty for guessing: a guessed-and-flagged question keeps a real shot at the point even if you never make it back, while a flagged blank is a guaranteed zero the moment the module locks. Same flag, wildly different floor. Make the guess your default, so that running out of time costs you nothing you could have protected.
Bluebook gives you a mark-for-review flag and a review screen for exactly this. Use them as designed. The flag is not an admission of defeat; it is you deciding, on purpose, that this point will be cheaper to earn in ten minutes than it is right now.
Pass 2: triage the flags
When you reach the end of the module, open the review screen. Two things have changed since you flagged those questions. You have recovered time, and you can now see all your flags at once instead of meeting them one at a time.
Do not return in numerical order. Triage by expected payoff: start with the flags you can most likely crack now that the clock pressure is off, then work down toward the ones that beat you the first time. A question you flagged because the choices blurred together often resolves in seconds on the second look, because you are no longer solving it in the shadow of a full module.
Two guardrails on this pass. First, if you revisit a question you already answered and feel the urge to change it, apply the Receipt Rule: switch only if you can point to the exact word that proves your first pick wrong. Second-guessing under pressure is its own leak, first-instinct override, and pass 2 is where it does its damage. Second, if the buzzer is close, confirm that no flagged question is still blank. Insurance guesses beat elegant solutions you did not have time to finish.
What to skip on sight
Some questions telegraph that they will be time-sinks before you have read a word of the choices. A dense quantitative-evidence item with a packed table. A cross-text pair that asks you to hold two authors' arguments at once. A question you have already read twice while the choices refuse to separate. These are candidates to skip on sight: flag and guess immediately, without even starting the ninety-second clock, and let pass 2 decide whether they are worth the return trip.
The thing that makes this hard is not the questions. It is the feeling. Once you have spent two minutes on an item, a voice says I have already invested this much, I cannot quit now. That is sunk-cost reasoning, the exact stubbornness the two-pass protocol exists to override. The two minutes are gone whether you stay or leave. The only live question is whether minute three buys more here than on the three questions still waiting. Usually it does not. Name the feeling, flag, move.
The practice protocol and a self-test you can run today
You cannot install this habit during a real test. The buzzer is not the place to try flagging for the first time. Build the reflex in practice until parking a hard question feels automatic instead of like giving up.
Run this self-test after your next timed module. List every question you spent more than ninety seconds on, then look at the two or three questions immediately after each. If that downstream cluster is where your uncharacteristic misses live, you have just watched the time-sink leak operate: one stall, then a rushed cluster, is the whole signature. It is the same pattern behind the back-half collapse many students see late in each module.
Then drill it deliberately. Take timed sets where you force yourself to flag and move the instant a question stalls, even when your instinct screams to stay. After each set, write a One-Sentence Autopsy of any miss: name the real reason, and ban the word "careless." The goal is not to answer hard questions faster. It is to make banking the cheap points first the thing your hand does without asking.
Practice it now: the free simplesat pacing drill puts you under a live clock and trains the flag-and-move reflex directly, so parking a stall becomes automatic before test day. No signup, works as a guest.