The “twos” were an afternoon’s work, back before my first tournament, absorbed simply by reading them and then writing them down in sequence, starting with AA (a type of Hawaiian lava) and ending with YO (as in, “Yo, Adrian!”). But even after a few months, I discover the twos still are not second nature. Is BO good or is it OB? (BO, as in beau, is acceptable.) Is it HM or SH that plays? (Both.)
The 972 “threes” are more challenging. I read down the columns on the cheat sheet and cross out those already familiar. About 70 percent go on the first pass. I try to assimilate the rest through repetition. No one has suggested a better method; the threes, I’m told, you just have to learn cold. So on a separate piece of paper, I write down the 300 or so unfamiliar words. AAL, ABA, ABO, ABY, AFF ... I read each word over and over, staring at it in hopes the image will imprint on my brain. When I’m confident it has, I place a check mark next to the word. Then I write up a new list: 180 words this time. I place a dot next to a word when it feels secure, a dash when it doesn’t.
What I don’t do is write down the definitions, which could help, and might be interesting. But definitions, for the limited purpose of playing Scrabble, don’t matter. One could even argue that the words really aren’t words at all. They are strings of letters, dancing across the board, an array of lines and arcs and circles. The strings usually represent language, but the letters that comprise them really are nothing more than, as G.I. Joel Sherman crudely puts it, “scoring tools,” which must be juxtaposed in a fashion deemed acceptable by a source or else rejected from the playing field. They could be random shapes or colors or buttons or widgets that must be placed in a regulated order. They just happen to be “letters” forming “words.”
That’s a formalistic view but a useful one for now. Over time, I will come to discover the beauty of words like FLOKATI and GANTLOPE and SEADROME and PANTOFLE and PERDU and OUGUIYA and SNAFUED and SIEROZEM and OQUASSA. I will fall in love with seven-letter words that take an eighth letter in front of them: LEVATOR-ELEVATOR, LEADERS-PLEADERS, ESTIVAL-AESTIVAL-FESTIVAL, INCITES-ZINCITES, ONETIMEZONETIME. And I’ll even look up their meanings and be better for it.
For now, though, I understand only that while definitions can be interesting, they’re not necessary. It’s just about impossible to play high-level (or even low-level) competitive Scrabble if you’re hung up on the game’s use of odd words. The two most common refrains of living room players are the incredulous “That’s a word?” and the indignant “That can’t be a word!” Because how can something be a word if I’ve never seen it before? The answer, I decide early on, is that there are lots and lots of words (hundreds of thousands, actually) that even the most highly educated person doesn’t know.
To play competitive Scrabble, one has to get over the conceit of refusing to acknowledge certain words as real and accept that the game requires learning words that may not have any outside utility. In the living room, Scrabble is about who has a better working vocabulary. It’s a sort of crossword puzzle in reverse. But in the tournament room, Scrabble has nothing to do with vocabulary. If it did, I—an Ivy League–educated professional journalist, for crying out loud—would rule. But I can only dream of competing with the champions. No, Scrabble isn’t about words. It’s about mastering the rules of the game, and the words are the rules.
Some players—like Matthew, the poet from Washington Square Park, or Ron Tiekert—seem to know the meanings of almost every word they play. They are curious word lovers who accept, if not agree with, the oddities and contradictions of the language. Some players, usually not the better ones, dumb down the game to their level. Words are weird—until they learn them. Then they aren’t. On the e-mail Scrabble discussion forum Crossword Games-Pro, or CGP, which I join, one player sums up the contradiction nicely. The issue “appears not to be the number of words, nor the strangeness of words, but the realization that one’s opponent gets to use the strange words he/she knows, but you don’t get to use the ones you know.” I save the post. In Danbury, playing “new” words like QUOIN, QANAT, and OUTVIES is almost as satisfying as winning.
“Each dictionary,” Dmitri Borgmann writes, “no matter how comprehensive, no matter how ‘unabridged,’ has selected a comparatively small number of words from the enormously large mass of words that make up the language.” Lexicographers get to decide what goes into dictionaries, and then people decide which dictionaries to use as sources. Another wordplay giant, Ross Eckler, writes that “each person must draw his own line between words and nonwords and, once having done so, communicate carefully to others what stockpile of words he is using. There is no right answer.”
The Scrabble world decided that The Official Scrabble Players Dictionary, first published in 1978, would, for the purposes of the game, answer the question: What is a word? The latest edition (the third, or OSPD3) includes words found in at least one of ten editions of five major U.S. college dictionaries. The Scrabble world was riven by a decision by Hasbro in 1994 to delete about 200 “offensive” words from the OSPD. As part of a compromise, the company agreed to have Merriam-Webster publish a book for use in sanctioned play which lists every word, dirty ones included, but without definitions or parts of speech. Just the words. The new book, which is published as I’m beginning my Scrabble career, is called the Official Tournament and Club Word List, and is known as the OWL.
Generally speaking, if a word is among the 120,302 two- through nine-letter words (plus inflected forms) listed in the OWL, it’s good. If not, it isn’t. Scrabble players argue about whether certain words should be deleted or others added; the NSA has a dictionary committee that is supposed to discuss such matters. While I’m quickly learning that the book isn’t without its inconsistencies, I decide that there’s no point worrying about it. I’m no lexicographer.
In addition to eschewing definitions, I also don’t write down which of the three-letter words can be pluralized and which can’t. That can come later. The magnitude of even this relatively short list feels overwhelming enough. A couple of mnemonic tricks, though, help. All of the letters in the phrase BETSY’S FEET, someone tells me, can be appended to the two-letter word KA (the spiritual form of a human being in Egyptian religion) to make a three-letter word (KAB, KAE, KAT, KAS, KAY, KAF), while all of the consonants in the phrase KNIGHT SWAM can be placed before AE (Scottish for “one”).
I buy some study aids, including a Franklin, a handheld electronic device the size of a credit card and the thickness of a cigarette case—and as addictive as either. It’s loaded with the now out-of-date OSPD2. Made in the early 1990s by Franklin Electronic Publishers, the gizmo apparently is available only in the Scrabble underground; I pay $43.25 to a guy in San Francisco who has bought much of the leftover inventory. He’s a black marketeer, but it’s worth it. Type in a word, press ENTER, and the Franklin tells you if the word is acceptable (and if so gives a definition) or not (in which case it suggests corrections). Type in a rack of letters, and the Franklin lists every acceptable word in the rack. Type in four question marks and the Franklin lists every four-letter word, in alphabetical order. Seven question marks followed by a rack of letters yields all of the bingos in the rack.
At tournaments, players whip out their Franklins as soon as games end to see what they missed. In one game, I have AEEEST? on my rack and there is an open H on the board. I play EH to unburden myself of an E. Afterward, I wonder whether I could have bingoed, and the Franklin mocks me with cyborg efficiency: AESTHETE. When I don’t know what to do with the promising-looking rack of ADENOPR, the Franklin points out my shortcomings in quadruplicate: APRONED, OPERAND, PADRONE, PANDORE.
For $20, I buy The Complete Wordbook and The Complete Blankbook, two oversized, multi-hundred-page tomes filled with words categorized in ways I have not yet seen. The psychedelic blue Wordbook is a classic, cocreated by Mike Baron, an expert player who was instrumental in the evolution of word lists for Scrabble study: Words with 70 percent vowels. Eight-letter words containing five vowels. Words of fewer
than seven letters containing J, Q, X, or Z. Seven-letter words arranged according to the most probable six-letter combinations plus a seventh letter. The same for eights. Four-letter words made from three-letter ones. Fives made from fours. Every three-, four-, five-, six-, seven-, and eight-letter word arranged alphabetically by alphagram. The Blankbook is even more bizarre: nothing more than two lists in alphagram order—words formed from six letters plus a blank and words formed from seven letters plus a blank.
The books are a blurry mass of capital letters that neatly categorize the language, but also make it seem impossible to digest. Column after column after column of words. No definitions or embellishments. Just the words. But when I open the books and began riffling the pages, I feel a pulse. It feels as if I have found the secret to Scrabble success.
Mike Baron’s bingo blueprint starts on page 24 of the Wordbook, but it actually began long before the days of clubs and tournaments, when competitive Scrabble was played mostly in game rooms ruled by chess hustlers. The most famous was the Chess & Checker Club of New York, better known as the Flea House. Scrabble had been played there since the fifties, when the game first took the country by storm.
Players learned words mostly by over-the-board osmosis, assisted by Funk & Wagnalls Standard College Dictionary, a liberal volume of 150,000 entries that was the Flea House word source (because a Funk & Wagnalls editor played there regularly). Sometime in the late 1960s, though, one of the players determined that the six letters in the word SATIRE offered enormous opportunities to make bingos. Poring through the dictionary, he determined that eighteen of the twenty-six letters in the alphabet combined with SATIRE to form sixty-seven seven-letter words.
In those days, there were no cheat sheets or computer programs that spit out words. No one practiced anagramming in the way Matt and Marlon do, barking out words in rapid succession. At the Flea House, the Scrabble players would swap anagrams, and spend hours or even days trying to solve them, “stuff that most intermediate players would just swat away these days,” Lester Schonbrun, the best New York player in the 1960s, and still one of the game’s best, tells me.
In the early 1970s, Selchow & Righter formed a national players organization and began publishing a newsletter. SATIRE was introduced in the second issue of the Scrabble Players Newsletter. The third issue posed another productive six-letter group: AENRST, dubbed SANTER*. More “stems” followed: SETTER, ENTERS, RETINA, SALTER, and, finally, TISANE, which combined with twenty-three letters forming fifty-nine words. (Both SATIRE and TISANE, a.k.a. SATINE*, now make more words thanks to additions to the Scrabble dictionary.) From seven-letter words, the quest moved on to eights.
The word searchers applied mathematical logic to the game. They knew that prefixes and suffixes were a key to making bingos. So it followed that letters like A, E, I, N, R, S, and T were good potential combiners. D, G, L, N, and O also seemed to show up a lot. The players’ intuition was backed up by history and linguistic research. In the 1890 edition of his wordplay book Gleanings for the Curious from the Harvest Fields of Literature, Charles Carroll Bombaugh published a chart listing the proportional frequency of the letters of the alphabet, which “have been pretty accurately determined.” In order, the first twelve were ETAOINSHRDLU, the sequence on which the standard Linotype keyboard was based. (The two left-hand columns of the keyboard produce the sequences ETAOIN and SHRDLU.) Academic studies of letter frequency in published American English confirmed the order. In Scrabble, it wasn’t surprising to find the G tossed in, because of the common occurrence of words ending in -ING. The H and U could be tossed out because there are only two H’s in a set and because the U is the clunkiest of the vowels.
The early lists, which were compiled by hand, weren’t always reliable. Almost every time one appeared in the newsletter, players wrote in with corrections, a sign of how the word-obsessed community was growing: “John Turner missed GANTRIES and PANTRIES in his RETAINS list. Make that at least 62 eight-letter bingos.” “Joe Cortese’s Crazee Eights anagrams need some help. ACONITES—CANOEIST and SONICATE. ASTERISM—MISRATES and SMARTIES. PSILOTIC—POLITICS and COLPITIS.”
The publication of the 662-page OSPD made it easier for players to search for words, but they still had to do it manually. One of the most prolific word searchers was David Shulman. A cryptanalyst of Japanese codes in the army during World War II and contributor to the Oxford English Dictionary supplement, Shulman took the study of bingo words to its first computational level. In 1979, he assigned a rating to the “top” fourteen six-letter stems by adding the number of letters in the alphabet with which the stem combined and the number of bingos that were formed; SATIRE was first with a rating of 85 (18 letters plus 67 bingos), followed by SATINE with an 82 (23 letters, 59 bingos).
It was as if sex or chocolate had just been discovered; the players couldn’t get enough. By 1980, the newsletter was dominated by lists; tournament results and club news were secondary. It was all about mastering the game, pulling the sword from the stone that was the OSPD.
Serious players learned that the word game was really a math game, as the pages of the renamed Scrabble Players Newspaper reinforced. In the February 1980 issue, an expert named Albert Weissman, a Connecticut psychologist, conducted the game’s first computerized mathematical experiments. Weissman calculated the probability of drawing certain racks to start play. The least likely combination of letters was BBJKQXZ, where the B could be replaced by any tile of which there are two in the bag (B, C, F, H, M, P, V, W, Y, and the blanks); the probability of drawing such a rack was about 1 in 16 billion. The single most probable rack, he found, was AEINORT, with an expected frequency of 1 in about 9,530 draws from a fresh bag.
But there was no acceptable seven-letter word in AEINORT (there still isn’t), so Weissman figured out the most and least probable bingo-producing draws. The former was (and still is) AEEINRT, which made RETINAE and TRAINEE (ARENITE, a kind of rock, was added to the second edition of the OSPD), with a probability of 1 in about 13,870 opening draws. A word like ERRATIC would come up 1 in 91,743 opening pulls. POACHES, 1 in 588,235. MUZJIKS, 1 in 55,555,555. Finally, MUUMUU?, where the blank is an S, could be expected to appear once every 8 billion opening turns, making it the least probable opening bingo.
The point was probability. As a game progresses, the number of letters from the finite pool of one hundred tiles is reduced toward zero, increasing or decreasing the odds of certain mathematical occurrences: drawing an S or a blank, playing a bingo, getting stuck with the Q, pulling a letter that would form a bingo with SATINE. If you could determine the letters most likely to be extracted from the bag, you could figure out which words were best to learn.
Weissman published the 125 most-probable seven-letter racks, 220 words in all. He concluded that players should study words based on the probability that they will show up during a game. Not everyone agreed; after all, 220 wasn’t many words. And the stems could be “built” deliberately during a game through expeditious play. The newsletter didn’t take sides. It just transmitted the information received. Next came a refinement of Shulman’s research. Rather than how many different letters of the alphabet combined with a six-letter stem to form a bingo, what mattered, it was decided, was how many of the ninety-four tiles that remained in the bag after drawing the stem could be used to make a bingo.
So the word searchers examined all 21,734 seven-letter words in the OSPD. One of them, a reclusive dictionary lover named Joseph Leonard, identified some two hundred worthwhile six-letter stems. The searchers combined the number of tiles in the bag which can be used to form bingos with the total number of bingos formed. SATINE took over the top spot, with a “Bonus Power Rating” of 150: Ninety of the remaining ninety-four tiles in the bag (every letter except J, Q, and Y) could be added to SATINE to make sixty different words. SATIRE was second at 142 (67 words plus 75 usable tiles), followed by SANTER and CLEARS.
It was a deluge. And, for players, manna from heaven.
Mike Baron
gave them even more. In the early 1980s, Baron was a psychologist at the University of New Mexico. His father and brother played in an early Scrabble tournament on Long Island, in 1973, and Mike and a few friends took up the game. He formed a club in Albuquerque, and as a result was invited to play in a qualifying tournament for the 1980 Nationals, the second of its kind.
Mike was seeded sixty-fourth in the field—dead last. He went 0–5 on the first day but thought, I can play this game. I just need to learn some words. On his flight home Mike started to circle every four-letter word in the OSPD containing K. After a few pages, though, he decided to tease out all of the short JQXZ words as well. Then he culled all the threes, and with the twos created the first cheat sheet, which he distributed to club members. Then he generated a list of “two-to-make-threes”—three-letter words formed by placing a letter in front or behind two-letter words—and then three-to-make-fours and four-to-make-fives.
His lists were a revelation: With all the emphasis in the newsletter on flashy bingos with their 50-point bonus, short words hadn’t gotten their due, and no one had determined their relative importance to play. “I was aware of certain parameters of the game,” Baron tells me. “The J, Q, X, and Z prevent you from bingoing, and you need to play two- to five-letter words to score. It was obvious to me, but there hadn’t been anything in the Scrabble News about it.”
Word Freak : Heartbreak, Triumph, Genius, and Obsession in the World of Competitive Scrabble Players (9780547524313) Page 6
