Game Guide
A walkthrough of all five Arithmetic Blitz demo levels, with the specific shortcut each question set rewards and what to do when the clock beats you.
Arithmetic Blitz is named after anzan, the Japanese discipline of mental calculation performed without the abacus in front of you, visualising the beads rather than moving them. The game is a long way from competition anzan, but it shares the constraint that matters: there is no time to write anything down, so you need methods that fit inside your head at speed.
The free browser demo runs five levels of five questions each, twenty five in total, with a seven second ring draining around every question and a keypad for your answer. Seven seconds is generous for a times table and brutal for anything you try to solve by column arithmetic. Here is every question and the shortcut that fits.
Here is the MyRin short for Arithmetic Blitz. This one is worth watching before you play, because the draining timer ring changes how the game feels in a way text cannot convey.
The five questions are 7 plus 5, 9 minus 4, 3 times 4, 8 plus 6 and 10 minus 7. Every one of these should be recall rather than calculation, and if any of them requires counting, that is the thing to practise before worrying about later levels.
Use this level to learn the keypad instead. Where the digits sit, how the entry confirms, how much of your seven seconds the typing itself consumes. On later levels the typing is a real fraction of the budget, and fumbling it costs questions you had already solved.
13 plus 8, 15 minus 6, 6 times 4, 18 divided by 3 and 7 times 7. Two of these deserve a method.
For 13 plus 8, do not add eight. Add ten and take back two: 13 plus 10 is 23, minus 2 is 21. Adding a round number and correcting is consistently faster than carrying, and it scales to much larger numbers where carrying does not.
For 15 minus 6, count up rather than down. Six plus what makes fifteen? Nine. Counting up from the smaller number, the same method Balance Scale is built on, is faster for most people and it makes fewer errors, because you are moving forwards through the number line rather than backwards.
24 divided by 6, 9 times 8, 45 minus 17, 56 divided by 7 and 12 times 6. The subtraction is the one that eats clock.
For 45 minus 17, round the subtrahend up to 20: 45 minus 20 is 25, then add back the 3 you over subtracted, giving 28. Notice the direction of the correction, because getting it backwards is the most common error in this method. You removed too much, so you add back.
For 9 times 8, use the ten minus one trick: 10 times 8 is 80, minus 8 is 72. This works for the whole nine times table and it removes the single most commonly misremembered row of the multiplication table from your critical path.
15 times 4, 144 divided by 12, 37 plus 48, 100 minus 63 and 9 times 11. Four of the five have clean shortcuts.
15 times 4: double twice. 15 doubled is 30, doubled again is 60. Doubling is the cheapest operation there is, and any multiplication by 4, 8 or 16 should be done by repeated doubling rather than by recall.
37 plus 48: round the 48 up to 50, giving 87, then subtract the 2 you added, giving 85. Same method as level 2, larger numbers, identical cost, which is the whole point of learning it early.
100 minus 63 should be immediate rather than calculated. Complements to 100 are worth memorising as a set: 63 pairs with 37, 45 with 55, 28 with 72. Subtracting from 100 appears constantly and it is the easiest family of facts to install.
9 times 11: use eleven as ten plus one. 9 times 10 is 90, plus 9 is 99. Multiplying by 11 by splitting it into 10 and 1 is faster than any remembered rule.
13 times 13, 256 divided by 16, 48 times 5, 121 minus 46 and 17 times 8. Seven seconds is not enough for column arithmetic on any of these.
13 times 13: split the second 13 into 10 and 3. 13 times 10 is 130, 13 times 3 is 39, total 169. Splitting one operand into a round part and a remainder is the general method behind almost every fast two digit multiplication.
256 divided by 16: halve repeatedly. 256 halved is 128, 64, 32, 16, and dividing by 16 means four halvings, so the answer is 16. Alternatively recognise that 16 times 16 is 256, one of the handful of squares worth knowing outright.
48 times 5: multiplying by 5 is halving then multiplying by 10. Half of 48 is 24, times 10 is 240. Always prefer this to adding 48 five times.
121 minus 46: round to 50. 121 minus 50 is 71, add back 4, giving 75.
17 times 8: split the 17. 10 times 8 is 80, 7 times 8 is 56, total 136. Or double three times: 17, 34, 68, 136. Both fit comfortably in the time.
You will lose questions, and how you lose them matters. The failure mode that ruins a run is not a timeout; it is carrying the frustration of a timeout into the next question, which arrives immediately. One lost question costs one question. One lost question plus a rattled thirty seconds costs four.
The practical discipline is to abandon early. If you are three seconds in and have no route to the answer, you are not going to find one in the remaining four seconds, and typing a rough guess is strictly better than typing nothing: it costs you the same and it clears your head for the next item. Deciding to let a question go is a skill, and it is the one that separates a good Blitz run from a good arithmetic ability.
What this game trains is narrow and genuine. Speed of retrieval for arithmetic facts improves with practice, and so does the automatic selection of the right shortcut, which is really a pattern recognition skill dressed as maths. Neither makes you better at reasoning in general. Both make the mental arithmetic that appears in ordinary life, bills, tips, unit prices, noticeably less effortful, and the installed game extends the same format to 600 levels if the twenty five questions here are not enough.