The Occasional Jam No. 05

No. 05 strategy

Getting Better at Numbers

6 min read

Numbers pays you for reaching a goal and pays again for reaching it the short way. I play it every day, and most of what I've learned comes down to an order: bank all three goals, then chase par.

It assumes you know the rules. If you don't, the how-to has them, with an example animation. Every figure below comes from working out every shortest route on real boards, goal by goal, and the sources at the foot of the page say how.

What the score rewards

Three numbers decide most of what follows: 100 for reaching a goal by any route, 45 more for matching its par, and 15 off that for each move over.

What one goal can pay, in points. Reaching it is most of the value.
What one goal can pay, in points. Reaching it is most of the value.
ItemValue
Reaching the goal100
Matching par+45
A One-Away+25

Two things follow from those numbers. An unsolved goal is worth more than par on both of the others put together, 100 against 90, so reach every goal before you polish any of them. And a long solve is never wasted. Solve the same goal again in fewer moves and you win back 15 for every move saved, so a solve two over par, later cut to par, ends up exactly where a par solve would have.

Bank every goal first, by any route. Par can be fixed later. A missed goal can't.

A One-Away stays yours if you go on to reach that goal, so landing one off early costs nothing. And because scoring tiers count points, not goals, a missed third goal doesn't cap your day: two goals at par, with their One-Aways, can still reach Wow.

Every goal gets all six numbers

Solving a goal doesn't use up the board for the others. Reset puts all six numbers back and keeps your solved goals and your points, so each goal starts fresh. Undo does the same one move at a time. Neither ever takes a point away, so there's no reason to protect a board you've finished with.

Par only counts the moves that built the goal. A detour you abandoned, a move whose result never fed into the answer, isn't counted against you. Try whatever you like and leave the mess where it is when you go for the solve: it costs you time and nothing else.

On a laptop the board is typeable. The keys 1 to 6 pick a tile, reading left to right and top to bottom. The keys +, -, *, and / pick an operation, Backspace undoes, and Esc lets go of a pick.

What most answers look like

Across real boards, the shortest routes have a strong family resemblance. These figures count only those routes, the ones par is measured against, so they describe what the boards allow and leave out how anyone actually played.

Moves at par for the median goal
4
Three or four for nearly three goals in four. One move for about one in eighty.
Of goals need a × to be done at par
96%
About 24 goals in every 25.
Of goals can end with + or − right after a ×
77%
A product near the goal, then a nudge.

That last figure is the pattern to learn. Most goals can be done at par as a product that lands near the goal, then a nudge with + or −, and the product almost always leans on a big number (97% of the time). Divide the goal by the big number in your head, and the answer tells you roughly what multiplier to build from the small ones.

One goal from a real board, worked

Goal      287
Numbers   2 · 4 · 11 · 50 · 75 · 100
287 ÷ 75 is a little under 4, so try 75 × 4.
75 × 4 = 300
11 + 2 = 13
300 − 13 = 287
Thirteen over, and two of the leftovers make 13.
Three moves, which is par. The 75 and the 4 get close, and the 11 and the 2 cover the gap.

Work from the goal, not from the numbers. Six numbers and four operations offer dozens of first moves, and most lead nowhere near the goal. When nothing jumps out, start from a multiplication. Pick the big number with a multiple closest to the goal, and build the multiplier if it isn't already on the board. Then the gap tells you what to keep back. If the product lands 13 over, keep whatever makes 13.

For half of all goals, the best possible two-move start lands within 4 of the goal. If two moves have left you 60 away, it's usually the product that needs changing. Undo both and try a different multiplier before you start stacking corrections on top.

Reading the board

The goal's last digit is a hint. About two in three goals ending in 0 can finish at par with a multiplication. For odd goals it's about one in six, so with an odd goal, plan the nudge from the start and save a number for it.

Don't leave division out. It's in a route at par for about a quarter of goals, and for about one goal in fifteen there's no route at par without it. When a goal won't come from below, try going over it and dividing back down.

The same day's hardest goal

75 × 50 = 3750
3750 − 2 = 3748
3748 ÷ 4 = 937
937 − 11 = 926
Four moves. Nothing shorter reaches it.
On that board, 926 has exactly one route at par, and it climbs to 3,750 on the way.

After a solve, read the card. "At par" means that goal has nothing left to win. Anything else, like "4 · par 3", says a shorter route exists and exactly how many moves it saves. That's the plainest hint the game gives: you know the target length, and you've already found the numbers that get there the long way.

When to settle, and when to chase

Three minutes for three goals is about a minute each. A typical day needs about ten moves at par across its three goals, and on nearly nine days in ten at least one goal needs four or five. Expect one goal to fight back, and don't let it eat the time the other two need.

  1. Take the goal you can see. The goals come in no order. Start wherever a product jumps out at you.
  2. Bank it at any length. A solve over par is still 100 points, and the shortfall can be won back later.
  3. Move on when you stall. If a goal has had its minute and you're no closer, try another goal and come back to this one later.
  4. Chase par last. Once every goal is banked, go back to any card that shows moves over par. Each move saved is 15 points.

My own start: I glance over the goals and the numbers for anything that gets me within 10 or 20 of a goal quickly, and start experimenting from there. If two goals are close together, I'll often go for those first, since it raises the odds of finding something. I also check my numbers for easy ways to make 1, 2, 3, 10, or anything else useful.

When a goal isn't coming and time is short, land exactly one off it. The One-Away still pays 25 for a goal that got away.

The last thirty seconds

The clock warns you at thirty seconds and again at ten. It only runs while the game is in front of you, so switching tabs or apps pauses it. A move that lands at the buzzer or after it doesn't count, so the last few seconds aren't the time to start a long chain.

Spend that last half minute on points that are already close. Stop chasing par on solved goals while one is still open. Take the open goal by any route, or land next to it. If everything is banked, re-solve a card that shows moves over par, because a shorter route you can already half see is the quickest 15 points left on the board.

Once the buzzer goes, give the answers under your result a minute, while the board you just spent three minutes on is still fresh. They show the shortest route to any goal you missed or made over par, which is where the products you didn't see turn up. A missed day in the archive gives you another board to practice on, and another set to read.

Sources

  1. Released Numbers boards For every goal, par and every route at par were found again by exhaustive search under the board's rules, and checked against the shortest way the game shows after the buzzer. The percentages count goals with at least one route at par of the shape described.
  2. The game's scoring code The 100 for reaching a goal, the step bonus and what each move over par costs, the refund for a shorter re-solve, and the One-Away.