Sean Bouchard

The first impression of these tiles is based on their appearance. Each tile is roughly square and appears to be made of polished wood. Compared to molded plastic, polished stone, or more exotic materials like ivory, the wooden Scrabble tiles have a color and texture that seems softer and earthier. The tiles are either highly polished or otherwise constructed in a way that makes them extremely smooth, and quite light, seeming to weigh little more than a playing card.

The comparison to playing cards is natural because, functionally, these tiles are similar to printed cards that are commonly used in other tabletop games. Like cards, these tiles have a front and a back, with the front designed to reveal information and the back designed to conceal it. Because cards are common and cheap, using tiles is obviously a deliberate design choice. The block-like nature of tiles makes it satisfying to stack or build with them, and the fact that they have much more depth than cardstock makes them feel more substantive than cards, even though they have a much smaller footprint.

The affordances of tiles are also different than those of cards, especially with respect to how they can be arranged and rearranged on play surface. The size of the tiles allows for many more of them to be placed in a reasonably-sized space, and still be highly accessible. Systems that involve placing and arranging cards, in contrast, overlap the cards in columns to save space. The overlapping convention makes it difficult to access or place cards in the middle of a column, and removing a card is usually done by sliding it rather than lifting it straight up, which requires that some significant space adjacent to the card be empty. Tiles, in contrast, when arranged in a grid, are much easier to place and remove from a crowded play surface.

On the front of each tile is printed a large capital English letter followed by a subscript numeral. This text appears to have been pyrographically burned into the wood of the tiles in a simple sans-serif font, consistent for the letter and numeral. (This text looks very different from the title of the game, which is written in heavy, highly-decorated letters that mix a respectable, traditional serif font with a whimsically curly informal font.) It's functional in its simplicity, and because there's a slight emboss on the writing, it's easy to tell the front of the tile from the back without looking at it.

One hundred of these tiles are included in the game. Aside from the two that are blank, each is encoded with a letter-number combination. A given letter is always paired with the same number, but most of the numbers appear with many different letters on different tiles. If we arrange the tiles into groups according to unique letter-number pairs, there are 27 groups which correspond to the 26 letters of the English alphabet, and the two blank tiles. In most of these groups, there are several duplicated tiles, but the number of duplicates varies widely by group.

Of course, that purely numerical consideration of the distribution of tiles is missing an important piece of the puzzle. Because the entirety of the English alphabet is represented on the tiles, it seems obvious to consider these tiles within the context of the English language. Tiles placed adjacent to each other form strings of letters that can include, notably, English words. Considering them within this context, we can guess that the underlying pattern for the number of tiles in each letter group has something to do with the frequency of that letter in English.

This chart uses known values for the frequencies of letters in English which come from cryptographic studies, compared with the frequencies of tiles in Scrabble. As you can see, the frequencies do not match exactly, but they are similar enough to suggest a connection.

Of course, the frequencies of letters in English writing can vary, depending on the situation, but the variations tend to be minor. Individual style and vocabulary can alter these distributions, but more interesting are the differences that present themselves when letters are counted by unique words, rather than across the text as a whole. When letter frequencies are calculated in this manner for a representative text we see some interesting variation. Specifically, notice that the frequencies of the letters "E", "T", and "H" are lower when only unique words are considered. This is an artifact of the most commonly-used word in the English language, "the".

The distribution of letters on the Scrabble tiles is slightly closer to the distribution of letters in unique words, at least for this particular passage of text. The implication is that the system of Scrabble tiles is probably better suited to the creation of individual words than phrases, sentences, or longer passages. (This is, of course, consistent with how the game of Scrabble is played. Depending on how tiles were used, the fact that there are only 100 tiles could also be a severe limiting factor for constructing text longer than individual words.)

It seems clear, from our analysis thus far, that Scrabble tiles have been designed so that they can be arranged to form English words. Or perhaps it is more correct to say that our analysis confirms an underlying logic to the system as it relates to forming English words, because this action is fundamentally intuitive. Frequency analysis may be fun and enlightening, but it is hardly necessary that it take place before a player begins, without prompting, to put these tiles together into word-like groupings.

Without dealing with the specific rules involved in the game of Scrabble, the fact that the system of tiles so strongly suggests play that involves the creation of words brings up some interesting ideas. First of all, the process of forming words out of Scrabble tiles includes two steps that are independent of any game mechanics: choosing an appropriate word and spelling it. These tasks rely on the player's language skills, specifically knowledge of vocabulary and spelling, such that a player with greater proficiency in these arenas will have greater success. While playing with this system of tiles - particularly solo play - may not be enough to improve the player's language skills outright, it does provide a motivating factor that could prompt an increased effort to improve.

Additionally, the ability to form words out of tiles is inherently intriguing, because words are powerful. Children in particular tend to be fascinated by the idea of forbidden words - words that are culturally, socially, or ideologically barred from use. Swear words and curses may be culturally forbidden; there is a class of words that are considered impolite, and therefore socially restricted; and in certain religions, uttering particular words or phrases is taboo. These words are always part of a socio-cultural system that is both pervasive and complex, and individuals are often tempted to subvert social norms to test the system. Scrabble tiles afford a player the ability to use forbidden words in an ambiguous context. Some individuals might take advantage of this affordance to experience a guilty pleasure at using a word they would not use under other circumstances, and others might take the opportunity to test social response to the ambiguity of the situation. Some might not participate in this practice at all, but still be engaged by the implicit possibility of it. In any case, there is something compelling about a system that allows for this sort of subversive behavior within a context that is both intellectual and playful.

Further, as a system that is implicitly related to language and word construction, Scrabble distinguishes itself from many other social tabletop games that rely more on chance, mathematical skill, or strategic thinking. Certainly this means that this system appeals to a different audience than other games, but it might also imply that the game itself occupies a different socio-cultural space than other games.

Overall, the system of letter tiles seems to have been fairly well designed to implement the core mechanic of Scrabble, but seem to suggest some elements that are notably absent from the official rules of the game. In the official game rules, for example, once a tile is placed it remains fixed for the duration of the game. As discussed earlier in comparison to cardstock game pieces, tiles are relatively easy to remove from a crowded play surface. Because the nature of the English language is such that many common words can become different words through the variation of one or a small number of letters, it may be surprising that the game doesn't make use of this affordance of the game pieces.

It may be likewise surprising that the meanings of words do not factor into the rules of the game. It may be less obvious how this would be incorporated into the gameplay of Scrabble, but the fact remains that the basic gameplay action results in the creation of symbols that have layers of associations in complex but commonly understood systems. Without requiring any changes to the formal system, the game could tap into the understood definitions of words or the grammatical rules that dictate relationships between words, but this potential goes untapped.

Perhaps experimentation with these sorts of variations would prove that the simplicity of the current game rules has an overwhelming advantage over any additional depth that might be introduced by creating new game rules. Still, it seems clear that the system that underlies the game could allow for quite a bit more variety than is reflected in the rules. Playing around with the rules for interacting with Scrabble tiles might prove to be as much educational fun as playing with the tiles themselves.

These graphs and more available at http://www.undefinedbehavior.com/usc/ctin541/frequency/.