Boggle's scoring table pays 11 points for an eight-letter word and 1 point for a three-letter word, so the usual advice is to hunt long words. We rolled and solved 100,000 boards to check. The boards mostly disagree.
Every Boggle strategy page repeats the same tips: go long, add prefixes and suffixes, read words backward. We couldn't find one that tests them against real boards. Atay Games makes Word Money, a Boggle-style word game, which gave us a reason to build a solver and put the tips on trial.
One rule for reading every number here: it's what the board contains, found by a computer. It isn't what a person finds in three minutes.
We rolled and solved 100,000 Boggle boards in 2026. The average board holds about 98 words worth 140 points, and 51% of those points sit in 3- and 4-letter words. Only 22% of boards hide an 8-letter word, while one word in five ends in S. Sweep short words from the center out, and chase S first.
Atay Games board simulation, 100,000 boards, public-domain ENABLE word list, run September 26, 2026.
- Short words carry half the score. 51% of the points on an average board come from 3- and 4-letter words. Counting only everyday words, it's 73%.
- Long words are rarer than the scoring suggests. Just 22% of boards hold any 8-letter word, and 58% hold one of 7 letters or more. With everyday words only, it's 3% and 16%.
- S is the hook that matters. 21% of all words end in S. ER and ED cover 2% to 3% each. ING covers 0.2%, because only 2 of the 16 dice carry a G.
- Boards showing an S hold 52% more points: 158 on average, against 104 on boards with no S.
- Look in the middle first. Each center tile sits in about 46% of a board's words; each corner tile in 13%.
- Quality still counts when time is short. A board's 10 best words are worth 31 points on average. Any 10 short words are worth 10.
- Two old tips hold up. 15% of words spell another word backward along the same path, and a board's longest word hides about 3 shorter words.
All figures: Atay Games simulation, September 2026. Words available on the board with the ENABLE list, not words a person found.
What is the best Boggle strategy, according to the data?
Sweep short words from the center out, hang an S on everything, and give long words one quick look. In our 2026 test of 100,000 solved boards, 3- and 4-letter words held 51% of the available points. Each center tile sat in 46% of a board's words, against 13% for a corner (Atay Games simulation).
That boils down to four habits, in this order:
- Start in the middle. Scan the four center tiles first, then the edges. Leave the corners for last.
- Sweep 3- and 4-letter words. They're worth 1 point each, but there are lots of them.
- Add S to every word that ends next to one. It's the cheapest second word on the board.
- Look once for a long word, early. If nothing long jumps out in a few seconds, go back to the sweep. When you do find a long word, mine it for the shorter words inside.
Here's how the popular tips fared. The middle column comes from a computer that knows every word in the list. The verdicts are about habits a person can copy.
| Tip | What the boards show | Verdict |
|---|---|---|
| "Focus on long words" | Words of 8+ letters sit on 22% of boards and carry 3.2% of all points | Partly. Look once; don't hunt. |
| "Start with short words" | 3- and 4-letter words carry 51% of points (73% with everyday words) | Supported |
| "Add S to everything" | 21% of words end in S; boards with an S hold 52% more points | Strongly supported |
| "Hunt ED, ER and ING" | ER 3.0%, ED 2.4%, ING 0.2% of words | ED and ER modest; ING rarely there |
| "Use prefixes: RE, UN, DE, DIS, PRE" | RE 2.5%, DE 1.2%; UN, DIS and PRE 0.3% or less | Mostly not on Boggle's dice |
| "Try words backward" | 15% of words reverse into another word | Supported |
| "Look for words inside long words" | About 3 per board's longest word | Supported |
| "Look for unusual words others might miss" (Hasbro's own tip) | This one is about rivals and the cross-off rule, not the board | Can't test here; see the surprises below |
All of this assumes you can scan quickly. If that's your weak spot, here's how to train your eyes to scan faster. The rest of this post shows where each number comes from.
The short answer. The best Boggle strategy, by the numbers, is a sweep. In Atay Games' 2026 test of 100,000 solved boards, 3- and 4-letter words held 51% of available points, center tiles sat in 46% of words, and 21% of words ended in S. Only 22% of boards held any 8-letter word.
How did we test 100,000 Boggle boards?
We rolled the 16 widely documented Boggle dice 100,000 times, solved every board against the public-domain ENABLE word list and scored each word with Hasbro's official table. In 2026, the results tracked earlier studies closely: about 98 words and 140 points per board, against roughly 104 and 150 in Bananagrammer's 2013 test.
How we ran it
- The dice. The 16-dice set published by Bananagrammer (2013), Robert Gamble (2016) and Stanford's CS106B course: AAEEGN ABBJOO ACHOPS AFFKPS AOOTTW CIMOTU DEILRX DELRVY DISTTY EEGHNW EEINSU EHRTVW EIOSST ELRTTY HIMNUQu HLNNRZ. It's the widely documented modern set. Hasbro doesn't publish an official dice list, and sources disagree on when this set arrived (1987 or 1992), so we can't promise it matches the box on your shelf.
- The roll. Each board shuffles the 16 dice into the 16 cells and shows one random face per die, the procedure the Stanford assignment describes.
- The rules. Hasbro's: words of three or more letters, built from letters that "adjoin in sequence horizontally, vertically or diagonally in any direction," with no cube used twice in one word. "The QU counts as two letters." Plurals and words inside words count.
- The word list. ENABLE, a Scrabble-style list that its readme places in the public domain, researched and compiled by Alan Beale. Our copy, from Peter Norvig's site, holds 172,724 words of three or more letters.
- The count. Each distinct word counts once per board, however many paths spell it. One player and no cross-off, so this measures what's available, not what anyone finds.
- The scoring. Hasbro's table: 3 or 4 letters = 1 point, 5 = 2, 6 = 3, 7 = 5, 8 or more = 11.
- The everyday-words check. A second run kept only the 15,021 ENABLE words that also rank among the 20,000 most common words on the web (Norvig's count_1w list, from the Google Web Trillion Word Corpus). It's a rough stand-in for words most people know.
- The run. September 26, 2026. 100,000 boards per run, one fixed seed per board (base seed 20260925), so every rerun gives identical results. A trie-and-depth-first-search solver in standard-library Python 3.14.4, 9 seconds on 14 processor cores. A self-test matched it against a brute-force check on 25 boards. The script itself isn't published.
How do our results compare with earlier studies?
| Measure | Our run | Earlier study | Source |
|---|---|---|---|
| Words / points per board, modern dice | 98.1 / 140.3 | about 104 / 150 | Bananagrammer 2013, 10,000 boards |
| Words / points per board, original dice (1976 to 1986 set) | 90.9 / 125.7 | about 93 / 128 | Bananagrammer 2013 (our check: 10,000 boards) |
| Points from 3- and 4-letter words | 51.0% | 49.3% | Gamble 2016, one billion boards, SOWPODS list |
| Extra points from the dice redesign | +11.6% | +12.9% | Gamble 2016 |
| Most common words, in order | TOE, TEE, TEN, NET, TEA, SET | TOE, TEE, TEN, NET, TEA, SET | luke-g.com, 2 billion games (top words by points per game, all worth 1 point) |
| Best board | 1,002 points in our sample | 3,625 points, proven best possible | Vanderkam 2025 |
Our averages sit a little below Bananagrammer's and well below the 133.62 words and 193.88 points in Robert Gamble's 2016 analysis, because the word list sets the count. Gamble used SOWPODS, which his own table puts at 267,651 words against 172,820 for ENABLE. The shape of the results matches anyway, and so does the size of the dice-redesign effect.
We didn't invent this kind of study. Bananagrammer, Gamble and luke-g.com published board data years ago, and Dan Vanderkam proved the best possible board in 2025. What's new here is the strategy angle: hooks, positions, letters, the tip tests and the everyday-words check. The limits are covered near the end.
How the test worked. Atay Games rolled the 16 widely documented Boggle dice 100,000 times in 2026 and solved every board against the public-domain ENABLE word list, scoring with Hasbro's official table. The results track earlier studies: about 98 words and 140 points per board, against roughly 104 and 150 in Bananagrammer's 2013 test.
How many words and points does a Boggle board hold?
More than you'd guess. In our 2026 test, the average board held about 98 words worth 140 points with the ENABLE list. The median board held 90 words and 119 points, because a few rich boards pull the average up (Atay Games simulation).
The spread is wide. The middle 80% of boards held between 42 and 165 words, and between 49 and 258 points. Only 8 of the 100,000 boards held no words at all. Counting only everyday words, the average board held about 44 words worth 52 points.
Points available per Boggle board, 100,000 boards
Atay Games board simulation, September 26, 2026 (first-party data). ENABLE word list, Hasbro scoring, no cross-off. The last bar collects every board worth 500 points or more.
The ceiling is far higher. Our best random board held 401 words worth 1,002 points. The best possible 4x4 board, proven by exhaustive search in 2025, holds 3,625 points from 1,045 words with the ENABLE2K list (Vanderkam, 2025; the proof is in arXiv 2507.02117). Back in 1982, Alan Frank's Word Ways article listed a board with 769 words worth 2,047 points.
Vowels matter more than you'd think. Boards showing 5 to 8 vowels (A, E, I, O or U, not counting Qu), about 67% of them, held the most: between 148 and 159 points on average. A board showing 3 vowels held about 100. You can't pick your board, but a thin board is thin for everyone at the table.
Qu drags a board down, too. Boards showing the Qu face, about one in six, averaged 121 points, against 140 overall.
What a board holds. An average Boggle board holds about 98 words worth 140 points with the ENABLE word list, but the median board holds 90 words and 119 points, according to Atay Games' 2026 simulation. Counting only everyday words, the average falls to about 44 words. The best possible board, proven in 2025, scores 3,625.
Should you hunt long words or sweep short ones in Boggle?
Sweep, then look once for length. In our 2026 test of 100,000 boards, 3- and 4-letter words held 51% of all available points, and a median of 58% per board. Only 22% of boards held any 8-letter word at all, so the 11-point prize usually isn't there to win (Atay Games simulation).
Short words are most of the words and half the points
Atay Games board simulation, September 26, 2026 (first-party data). Share of all words and all points available across 100,000 boards, ENABLE word list, Hasbro scoring.
The scoring table pays steeply for length, but the board rarely offers length. Words of 8 or more letters made up 0.4% of all words and 3.2% of all points. Even the single best word on a board averaged 5.4 points, typically about 4% of that board's total.
Gamble found the same split in 2016 with a billion boards and the bigger SOWPODS list. Together, he wrote, 3- and 4-letter words "comprise about 70% of the words and 50% of the points on an average board." Our run confirms it with a different word list.
When does length still pay?
When time is short, quality counts. A board's 10 best words were worth 30.7 points on average, against 10 points for any 10 short words. So a long word you can see is worth grabbing. A long word you're hoping for isn't. And on boards that did hold an 8-letter word, short words still carried 93 of their 238 points.
Does your vocabulary change the answer?
Yes, a lot. With everyday words only, 3- and 4-letter words carried 73% of the points, and just 3% of boards held an everyday 8-letter word. Players who know thousands of Scrabble words gain more from hunting length. Everyone else should sweep harder.
The practical rule: spend a few seconds early on one long word, then sweep. Our Word Money guide gives the same advice for its grid: nothing long in a few seconds, move on.
Long vs short. Short words beat long words on most Boggle boards. In Atay Games' 2026 test of 100,000 boards, 3- and 4-letter words carried 51% of available points, while 8-letter words appeared on just 22% of boards. Counting only everyday words, short words carried 73% and 8-letter words appeared on 3%.
Which Boggle hooks are worth hunting: S, ED, ER, ING or RE?
S, by a mile. In our 2026 test of 100,000 boards, 21% of all words ended in S, and 16% were words whose shorter form sat on the same board. ER covered 3.0% of words, ED 2.4% and ING just 0.2% (Atay Games simulation). Most tip lists treat these endings as equals. The boards don't.
A hook here means a letter or ending added to a word you already have, like SET to SETS. The stem is the shorter word. When the stem and the longer word are both on the board, you get two words from one trace.
| Hook | Share of words | Share of points | Stem on board, share of words | Everyday words, share of words |
|---|---|---|---|---|
| Ends in S | 20.7% | 24.6% | 15.9% | 14.9% |
| Ends in ES (part of S) | 5.0% | 6.9% | counted in S | 3.1% |
| Ends in ER | 3.0% | 4.7% | 1.9% | 2.0% |
| Starts with RE | 2.5% | 3.1% | 0.6% | 2.6% |
| Ends in ED | 2.4% | 3.6% | 1.8% | 1.8% |
| Ends in ING | 0.2% | 0.4% | 0.1% | 0.2% |
| Any of S, ED, ER, ING, RE | 28.1% | 35.4% | 20.1% | 20.9% |
S dwarfs every other hook on Boggle's dice
Atay Games board simulation, September 26, 2026 (first-party data). Share of all words available across 100,000 boards, ENABLE word list. Endings and prefixes are matched on spelling alone.
Why does ING almost never show up? It's countable. Only 2 of the 16 dice carry a G (AAEEGN and EEGHNW), and then I, N and G have to land touching, in that order. S sits on 5 dice with 6 faces between them, and at least one S showed on 68% of our boards. Tip lists recommend ING, UN and DIS as if they were as common as S. On these dice, they aren't.
The S effect shows up in the totals. Boards showing an S held 158 points on average, against 104 on boards without one: 52% more. With everyday words, the gap was 32%.
The practical rule: find the S tiles first. Any word that ends next to an S may hand you a free second word. Don't spend time hunting ING.
Hooks, ranked. S is the only Boggle hook that pays much. In Atay Games' 2026 simulation, 21% of all words ended in S, while ER and ED covered 3.0% and 2.4% and ING just 0.2%, because only 2 of Boggle's 16 dice carry a G. Boards showing an S held 52% more points.
Where on a Boggle grid should you look first?
The middle. In our 2026 test of 100,000 boards, each of the four center tiles sat in about 46% of a board's words, each edge tile in 25% and each corner tile in 13%. That makes a center tile about 3.5 times as productive as a corner (Atay Games simulation).
Share of a board's words that pass through each tile
Atay Games board simulation, September 26, 2026 (first-party data). Average share of a board's words that use each tile, 100,000 boards, ENABLE word list.
It's geometry. A center tile touches 8 neighbors, an edge tile 5 and a corner tile 3. More neighbors means more paths through a tile, and more paths means more words. The pattern held for everyday words too: 42.5% for a center tile against 12.3% for a corner.
Which letters are worth building from?
Average number of a board's words that pass through a tile showing each letter. Atay Games board simulation, 100,000 boards, ENABLE word list (first-party data).
The letter on a tile matters as much as where it sits. A tile showing S sat in 43.6 words on average, then E (39.2), A (38.6), R (33.9) and T (31.9). The weakest were Qu (3.8), Z (3.9), J (4.1) and X (4.7).
The practical rule: start your scan on the center four. If an S or an E sits in the middle, build outward from it. Check the corners last.
Where to look. Look in the middle of a Boggle board first. In Atay Games' 2026 test of 100,000 boards, each center tile sat in about 46% of a board's words and each corner tile in 13%. Tiles showing S sat in about 44 words on average; the Qu tile sat in about 4.
Do the classic Boggle tips hold up?
Mostly, yes. In our 2026 simulation, 15% of words spelled another word when read backward along the same path, and a board's longest word held about 3 shorter words inside it (Atay Games). Both are cheap second words, because you've already traced the path. A few other results surprised us.
Reversals are pairs like TEN and NET, or TON and NOT: the same tiles, read the other way. Words inside words follow the example in Hasbro's instructions, where SPARE also holds "spa, par, are, spar, pare." With everyday words only, the reversal rate was 13.8%, and a longest word held 1.7 shorter words on average.
What surprised us?
- ING is almost absent. It covers 0.2% of words, for the dice reason above, yet it appears on most tip lists.
- The dictionary changes everything. Hasbro's own example board, which its instructions say holds "more than 70" words, holds 202 in ENABLE by our reading of its figure, and 99 everyday words. Several of the most common ENABLE words, such as TAE, ETH and OES (each on 15% to 17% of boards), are ones most players would challenge.
- The cross-off tip is about people, not boards. Hasbro's rules cross off "any word that appears on more than one player's list," and its tip is to "look for unusual words that others might miss." The words that show up most often (TOE, TEE, TEN, NET, TEA and SET, each on 17% to 19% of boards) are exactly the ones rivals will write down too. That's our inference; we didn't model opponents.
Playing a digital version, such as Netflix's Boggle Party? Check its rules screen, since apps can score differently. The board habits carry over. Letter-wheel games work differently again; for those, try a two-pass routine for letter-wheel puzzles.
Old tips, new data. Two classic Boggle tips hold up. In Atay Games' 2026 simulation, 15% of words spelled another word when read backward along the same path, and a board's longest word hid about 3 shorter words. The dictionary matters too: Hasbro's example board, which Hasbro says holds "more than 70" words, holds 202 in ENABLE by our count.
How do these Boggle habits carry over to Word Money?
The habits carry over; the math doesn't. Word Money is our Boggle-style game. Its App Store listing, checked in September 2026, says to "connect letters in any direction to create words" and that "the longer the word, the bigger the points." Unlike Boggle, each tile also shows its own point value.
Word Money runs on Skillz, operated by Firy Inc (NYSE: FIRY, formerly SKLZ). The listing adds that "everyone plays on the same board" and that "the top score takes the win." We won't state its round length, minimum word length or exact scoring formula here, because the in-game tutorial is the final word. Our guide to how Word Money scores and plays covers what's confirmed.
How does tile scoring change the math?
Boggle's table jumps from 1 point to 11 as words get longer. Tile scoring, where a word is worth the sum of its tiles, grows in a straight line with length instead. To see what that does, we re-scored the same 100,000 boards two made-up ways. These are hypothetical rules for illustration, not Word Money's formula.
| Word length | Boggle's table | Hypothetical: 1 point per tile | Hypothetical: random tile values 1 to 5 |
|---|---|---|---|
| 3 letters | 25.5% | 27.2% | 27.4% |
| 4 letters | 25.5% | 36.3% | 36.3% |
| 5 letters | 25.3% | 22.5% | 22.4% |
| 6 letters | 14.2% | 10.1% | 10.0% |
| 7 letters | 6.4% | 3.2% | 3.1% |
| 8+ letters | 3.2% | 0.8% | 0.8% |
| 3 and 4 letters together | 51.0% | 63.5% | 63.7% |
Under straight-line scoring, 3- and 4-letter words carried about 64% of the points, against 51% under Boggle's steep table. And under the hypothetical random tile values, the top-scoring word on 23% of boards wasn't one of the board's longest words. So if a game scored by tile values alone, the sweep would matter even more, and the longest word wouldn't automatically be the best one. Read the numbers on the tiles.
What carries over: center-first scanning, S hooks where plurals count, reversals, words inside words and one early look for a long word. What doesn't: Boggle's scoring table and its cross-off habit, since the listing describes no cross-off. We also don't know Word Money's letter mix, so the ING finding is a Boggle-dice result, not a Word Money one.
Is using a solver in a cash match cheating?
Yes. Skillz's User Terms (updated May 22, 2025) list "the use of unauthorized or altered software or hardware to assist play" as abuse in section 7.3. Sanctions can include closing the account, voiding winnings and "a fine of up to $1,000." Using a solver in a cash match is cheating. Look words up after a match, never during one.
One more rule worth knowing: stay in the app until the match ends. Skillz's help center says that when a player crashes or disconnects, the match is marked "Aborted." In its words: "The opposing player is automatically awarded with winning the match and any corresponding prizes." Here's what happens if you leave a match.
From Boggle to Word Money. Boggle habits carry over to Word Money, but the math doesn't. Word Money's App Store listing says longer words score more, and every tile shows a point value. Under a hypothetical flat per-tile rule, 3- and 4-letter words held 64% of points, against 51% under Boggle's table (Atay Games, 2026).
What can't a Boggle board simulation tell you?
How many words you'll actually find. Our 2026 numbers are what the board holds, found by a computer with a 172,724-word list and unlimited time. We found no published study that measures how many words real players find in three minutes, so treat every figure here as the size of the target, not your score.
- No people. We didn't model typing speed, time to find a word or how many words anyone spots.
- The dictionary. ENABLE is a Scrabble-style list full of words most people would challenge. The everyday-words check is a rough filter, not a vocabulary test.
- The dice. We used the widely documented modern set. Hasbro publishes no official list, so a box bought today could differ.
- No opponents. Real Boggle crosses off shared words; our count doesn't.
- Made-up tile scoring. The tile-value results are hypothetical.
Three questions we'd like answered next: what share of each word length real players find in three minutes, whether center-first scanning measurably raises human scores, and how Boggle-style apps generate their letters. We ran a similar same-deal test for solitaire, and the same caution applied there: a model shows the shape of a game, not a person's result.
The limits. A board simulation measures what a Boggle board holds, not what a person finds. Atay Games' 2026 counts come from a computer with a 172,724-word list and unlimited time, with no opponents and no cross-off. No published study we found measures how many words real players find in three minutes.
Frequently Asked Questions
What is the best strategy for Boggle?
Sweep short words first, starting from the four center tiles, and add an S to every word that touches one. In our 2026 test of 100,000 solved boards, 3- and 4-letter words held 51% of all points, center tiles sat in 46% of words, and 21% of words ended in S.
Is it better to find long words or short words in Boggle?
Mostly short, with one quick look for a long word. On our 100,000 simulated boards, 3- and 4-letter words carried 51% of the available points, and only 22% of boards held any 8-letter word. When one is there, it's worth 11 points, so a few seconds of looking pays.
How many words are on a Boggle board?
In our 2026 simulation, the average board held about 98 words worth 140 points with the ENABLE word list, and the median board held 90 words. Counting only everyday words, the average fell to about 44. Those are words the board contains, not words a person finds in three minutes.
Does Qu count as two letters in Boggle?
Yes. Hasbro's official instructions say "The QU counts as two letters," so QUIET scores as a five-letter word, worth 2 points, even though it uses four cubes. The Qu cube is still the weakest tile: in our test, a tile showing Qu sat in about 4 words per board, against 44 for an S.
Is using a Boggle solver cheating?
At a kitchen table, that's for your group to decide. In a cash match, it's cheating: Skillz's User Terms list "the use of unauthorized or altered software or hardware to assist play" as abuse, with sanctions that can include account termination, forfeited winnings and a fine of up to $1,000. Look words up after a game, never during one.
How did you get these numbers, and can I cite them?
We rolled 100,000 boards with the 16 widely documented Boggle dice and solved each one against the public-domain ENABLE word list, with a 3-letter minimum and Qu as two letters, on September 26, 2026. Yes, please cite: Atay Games, "Boggle Strategy by the Numbers," ataygames.com, September 2026.
The bottom line
A Boggle board rewards a sweep, not a treasure hunt. Start in the center, hang an S on everything, give long words one early look, then mine your long words for the shorter ones inside. That order follows where the points actually sat on 100,000 boards.
One honest caveat: these habits raise the share of a board you capture, but a computer will still find far more than any three-minute list. If you play word games mainly to stay sharp, here's more on word-building games as a brain workout.
Want to try the routine on a live grid? Practice free first in Word Money, our Boggle-style word game on Skillz, before you decide on a cash match. It's on iPhone and iPad through the App Store, and on Android as an APK from the official Atay site. Here's why real-money games aren't on Google Play. Real-money play is 18+, and whether it's offered depends on where you live.
Data appendix: the study in one table
| Measure | All ENABLE words | Everyday words only |
|---|---|---|
| Words per board, mean (median) | 98.1 (90) | 44.0 (41) |
| Points per board, mean (median) | 140.3 (119) | 52.0 (47) |
| Points from 3- and 4-letter words | 51.0% | 73.2% |
| Boards with any 8+ letter word | 22.0% | 2.9% |
| Boards with any 7+ letter word | 57.9% | 15.8% |
| Words ending in S | 20.7% | 14.9% |
| Points, boards with vs without an S | 157.6 vs 103.9 | 56.3 vs 42.8 |
| Words through one center / edge / corner tile | 46.4% / 24.9% / 13.3% | 42.5% / 22.8% / 12.3% |
| Words with a reversal partner | 15.2% | 13.8% |
| Shorter words inside the longest word | 3.26 | 1.74 |
| Best board in the sample | 1,002 points (401 words) | 273 points (185 words) |
How to cite this study: Atay Games, "Boggle Strategy by the Numbers: What 100,000 Boards Show," ataygames.com, September 2026. Data from our 100,000-board run on September 26, 2026; we'll update the figures if we re-run it.
Sources
- Atay Games, Boggle board simulation (first-party): 100,000 boards per run, base seed 20260925, ENABLE and everyday-words runs plus a 10,000-board check with the original dice, standard-library Python 3.14.4. Run 2026-09-26. A model of what boards contain, not a measurement of people or of any app; method described in this post.
- Hasbro, Boggle instructions (item 04601, © 2008): rules, scoring table, "The QU counts as two letters" and the example board. Retrieved 2026-09-26. instructions.hasbro.com
- Bananagrammer (Surly Teabag), "The Boggle cube redesign and its effect on the difficulty of Boggle," 2013-10-17. Retrieved 2026-09-26. bananagrammer.com
- Robert Gamble, "A Programmer's Analysis of Boggle," 2016-01-03. Retrieved 2026-09-26. robertgamble.net
- luke-g.com, "Boggle Word Frequencies" (undated page). Retrieved 2026-09-26. luke-g.com
- Dan Vanderkam, "After 20 Years, the Globally Optimal Boggle Board," 2025-04-23. Retrieved 2026-09-26. danvk.org
- Dan Vanderkam, "A Computational Proof of the Highest-Scoring Boggle Board," arXiv 2507.02117, submitted 2025-07-02. Retrieved 2026-09-26. arxiv.org
- Butler University Digital Commons, Alan Frank, "High-Scoring Boggle," Word Ways 15(3), 1982 (board figures as quoted by Vanderkam, 2025). Retrieved 2026-09-26. digitalcommons.butler.edu
- Stanford University, CS106B Boggle assignment specification (undated course page). Retrieved 2026-09-26. web.stanford.edu
- Peter Norvig, "Natural Language Corpus Data: Beautiful Data" (hosts the ENABLE word list, plus count_1w.txt, which is derived from the Google Web Trillion Word Corpus). Retrieved 2026-09-26. norvig.com
- ENABLE word list readme (README-enable2k), mirrored in the BartMassey/wordlists repository, GitHub. Retrieved 2026-09-26. github.com
- Skillz, User Terms and Conditions of Service, "Updated as of May 22, 2025," section 7.3. Retrieved 2026-09-26. skillz.com
- Skillz help center, "What is an Aborted match?", article 8778934866580 (updated 2026-08-01). Retrieved 2026-09-26. support.skillz.com
- Apple App Store, Word Money Game Win Real Cash (id6744853139, version 1.66, released 2026-09-10). Retrieved 2026-09-26. apps.apple.com
- Trademark note: Boggle is a trademark of Hasbro. Atay Games and Word Money aren't affiliated with Hasbro; we use the name only to describe the game we studied.
About this study. Our simulation is a model of what Boggle boards contain with one word list, not a measurement of people, of any retail Boggle set or of any app. The tile-value results use made-up scoring rules, not Word Money's formula. Boggle is a trademark of Hasbro, and this site isn't affiliated with Hasbro. In cash play, better habits raise your odds over many games but never guarantee a result, and you can lose entry fees. Real-money play is 18+ and isn't available everywhere.
Practice Free, Then Decide
Try the center-first sweep in Word Money's free practice first. If you enter a cash match later, pick an entry fee you're comfortable losing. Cash play is optional, skill-based and 18+.
Play Word Money
