Two candidates can know exactly the same mathematics and finish ten marks apart, because the clock is a section of the paper in its own right. The HAT gives you no per-section limit, which feels generous and is actually the trap: a single analytical puzzle that swallows nine minutes removes the time you needed for three verbal questions you would certainly have answered correctly.
Budgeting is not about hurrying. It is about deciding in advance what each question is allowed to cost, so that the decision costs no thinking during the paper. Athletes do not decide their pace mid-race; they know their splits before the gun.
Model the paper as a budget you spend. You start with 7,200 seconds and every question draws from the same account.
Weights tell you how large each section's bill is. On HAT-1 the marks split 40/30/30, and with 100 questions the papers roughly split the same way: about 40 quantitative, 30 verbal and 30 analytical questions. Spending in proportion to marks looks like this:
| Section | Marks | Approx. questions | Neutral time | Time per question |
|---|---|---|---|---|
| Quantitative | 40 | 40 | 48 min | 72 s |
| Verbal | 30 | 30 | 36 min | 72 s |
| Analytical | 30 | 30 | 36 min | 72 s |
The neutral split is your baseline, not your plan. The plan adds two things: a reserve and checkpoints.
Here is a concrete HAT-1 plan. It front-loads Quantitative because that section is worth the most and because its questions vary most in cost, and it keeps about eight minutes in reserve.
| Elapsed | Checkpoint | Action if behind |
|---|---|---|
| 0-8 min | Settle and run the first pass on the easiest quantitative questions | Skip anything that needs a written setup longer than one line |
| 8-50 min | Quantitative completed to a first pass | Move on with gaps flagged; the second pass returns to them |
| 50-85 min | Verbal completed | Give reading passages a hard 4-minute ceiling each |
| 85-112 min | Analytical completed | Cap each puzzle setup at 6 minutes, then guess and leave |
| 112-120 min | Final sweep | No blanks anywhere; re-check flagged items only |
Three structural facts about the clock deserve their own paragraph:
It is minute 60 with a HAT-1 plan; you have completed 30 quantitative questions, 25 verbal and 5 analytical. Forty questions remain, and there are 60 minutes left. Do you keep the plan?
Arithmetic first: 40 questions with 60 minutes is 90 seconds each, which is 25% more than the neutral 72 seconds - comfortable. What is left is 10 quantitative, 5 verbal and 25 analytical. Allocate 15 minutes to the remaining quantitative and verbal items and 40 minutes to the analytical block, holding 5 minutes as the reserve. That is 60 seconds per quantitative or verbal item and 96 seconds per analytical item - tighter than the neutral pace on the short items, but achievable, because you are now answering only questions you can see your way through.
What changed at the checkpoint: not the plan's shape, but the reserve. Being behind at one checkpoint should spend the reserve, never the accuracy.
The HAT has no per-section time limit. What does that fact actually change?
You have spent three minutes on a quantitative question and still see no route. What is the correct action?
A single analytical setup comes with five linked questions. What is a sensible budget for the whole set?
Without looking back: what is the neutral time per question, and what is the reserve for?
The answer: 72 seconds; the reserve pays for second-pass puzzles and the final sweep that fills every blank.
The paper gives me minutes for 100 questions, which is seconds per question. I hold roughly minutes in reserve, and when I cannot see a route within seconds I flag the question and move on.
You are 45 minutes in. Quantitative is finished, verbal is done except one passage, and you have spent 11 minutes on a scheduling puzzle that is still not solved. There are 75 minutes left and 34 questions untouched. Write the decision.
Leave the puzzle now, with a provisional answer shaded, and finish the reading passage while your reading speed is fresh. Then take the untouched analytical sets in order, capped at eight minutes per set, and return to the abandoned puzzle only if the second pass finishes early. If it does not, the provisional answer stays — a guessed set is worth the same as a blank, and you will not leave it blank.
The numbers behind the decision: 34 questions in 75 minutes is 2.2 minutes each, so the pace is comfortable as long as no single item is allowed to become unbounded. Eleven minutes on one setup is already more than a whole five-question set's budget; the mistake was not the attempt, it was the absence of a cap before starting.
You now have a budget and checkpoints. The next lesson is the other half of pacing: what to do with a question you cannot finish — the elimination arithmetic that makes guessing profitable, and the triage rules that decide which questions to skip on sight.