Wednesday, 4 June 2008
My Boggle Strategy (FWIW)
As a player I tend to focus on small word clusters, rather than on the larger board. As a result, I miss almost all of the longer words, but get a good percentage of the smaller ones. I work up to the longer words, rather than down from them. My strategy is more geared for regular 4x4 boards than it is for 5x5s or the new 4x4(4)s.
I am a slow typist and also need to spend a lot of time looking at my fingers, so I have to see words quickly in order to keep ahead of my fellow players. In order to do this, I think in terms of creating strong and interlocking word networks. In a game you always need to know which word comes next, as it were. Each word needs to lead to the next word or set of words. You shouldn’t have to pause to think about it; pause and it might cost you the game.
Word networks can be formed in many ways. At first my networks were primarily formed by a mixture of letter order, semantics and phonetics. I would mentally say the words to myself as I wrote them, checking to see if they sounded right. If I saw HEAD, I would probably see DEAD. The problem was that I would often miss a word like BEAD, as it was pronounced differently. I still employ this method on occasions, but complement it with many others and don’t rely on it so much.
Primary Opening Strategy
Nowadays (or until I went cold turkey, at least) I rely far more on anagram patterns and hook words.
An example movement might help to illustrate my approach. Having begun by quickly scanning the board for common prefixes and suffixes and taking note of them (if there is an open –ING, or an –ED that looks promising I will work with it immediately), I start off seeing letters like A-R-I and then instantly write AIR/RAI/RIA. I keep my eye out for an S and, if there is one, I type the plural forms, along with ARIS/SARI/SAIR. I then see that a T is connected and type ART/TAR/RAT and AIT/TAI/ITA as the remaining subanagrams of A-R-I-T. I type ARTI/RAIT/TIAR/AIRT and look for an A so that I can have ATRIA/AARTI/RIATA/RAITA/TIARA/TAIRA. At each stage I am looking for possible plural forms of each of these words and typing them as I see them (and adding in words like RASTA and ASTIR, if they are there). I remember that ATRIA, being the plural of ATRIUM, cannot take an S, but that it can take an L at either end, making LATRIA or ATRIAL (I also remember that LARIAT is a word; if I see LARIAT, I also glance to see if SALARIAT is there — the only SOWPODS anagram of my first name). If I see an L at the end, I would immediately move to TRIAL/TRAIL/LIART, RIAL/RAIL/LAIR/LARI/LIRA/LIAR/ARIL and TIRL.
Prefixes and Suffixes
Most of the words listed above cannot take any common suffix (-ED, -ER, -ING, etc.) or prefix (RE-, DE-, EX-, UN-, etc.), but a few can. When I type a word like TRAIL, I will look out of the corner of my eye for such suffixes and prefixes. If they look very promising I will type them immediately and make a mental note to return to the other word chain. Once I feel that I have got the best out of a certain set of letters, I will start again somewhere else on the board.
As I said earlier, each word needs to lead to the next. There are plenty of mediocre Boggle players out there with a far better vocabulary than mine. The difference is to be found in the length and strength of my word threads. When it comes to Boggle, the size of one’s vocabulary counts far less than the strength with which your words are bound together. Consequently, if you want to improve your game this is what you need to focus on.
There are a few places where connections, either between word sets or within them, may be weak. There are some words that you will generally miss as you move between and within word sets. Your goal must be to ‘stick’ such words to their neighbours. For example, while AIR/RIA/RAI may be easy enough, ARIL/LAIR/LARI/LIAR/LIRA/RAIL/RIAL is slightly harder. I remember that particular set by trying to observe loose patterns: in the L anagrams, the R only appears in third and fourth position and the vowels can be reversed when the R is in both positions. A similar pattern helps me to remember RIATA/RAITA/TIARA/TAIRA. Once I have remembered those, the remaining two A words are not too hard (that anagram set has two A words, two R words and two T words).
I remember the number of words in an anagram set and sometimes the number of words in the set beginning with particular letters. I rehearse word chains (like the one illustrated above) in my mind, until I get every word every time.
I divide each anagram set (for 4-letter words and greater) into subanagram sets and learn the most useful anagram sets that can be formed by the addition of a letter. For instance, if I have the letters A-A-L-P, I write PAAL/ALAP, AAL/ALA, ALP/PAL/LAP, but my priority is on looking for an S and another A. If I find the S, I write PALAS/ALAPS/PAALS/PLAAS/SALPA. If I find the A, I write ALAAP/ALAPA.
One of the best ways to strengthen connections is to learn the meanings of words. For instance, having studied Hebrew, I hardly ever miss ZAYIN and AYIN when they come up on a board. Semantically reinforcing key words in your mind makes it easier to remember connected words as well. Once you come to know how your own mental word chains work, you will better know which parts need some ‘semantic greasing’!
Creating New Connections
There are a number of ways in which words can be connected in networks. Anagrams are my primary focus. When learning anagrams I focus on the largest sets for each word length and then work on those which have the most common letters, or those which contain the most unusual words. I can gradually work my way through all of the words of a particular length following such a method.
I also use hook words a lot. SUR, SURA(S), SURRA(S), SURAH(S) and SUN(S), SUNN(S), SUNNA(S), SUNNAH(S) are good examples of useful hook word sets. The importance of hook words cannot be overstated. They are a great way of connecting uncommon words to more familiar ones. For instance, both WHAT and DOWN can take an A at the end to form WHATA and DOWNA (ADOWN is also a word).
I also form sets of words with slightly unusual common endings, e.g. BEARE, DEARE, FEARE, GEARE, HEARE, LEARE, MEARE, PEARE, SEARE, WHEARE. There are various mnemonics that one can create to help one to remember such a list. I use ‘DouGH, (Wh)iFF, BLiMP’ to help me to remember the list above. Rather poor, I know, but it does the job (you might prefer ‘(Wh)ee! FaBLeD PHLeGM!’). I then remember that each one of these words takes an S at the end, except the one with an S at the beginning; all take a D at the end except those with a HuMP in them; all take an R at the end except the GLuMPy ones (and WHEARE). Somewhat cryptic, perhaps, but it usually works.
I try to remember some of the unusual prefixes that certain words can take. For instance, LITH can take a number of prefixes, making OOLITH, EOLITH, NEOLITH, ZOOLITH, TALLITH, OTOLITH, TRILITH, and UROLITH. I also create some sets of similar looking words in my mind: e.g. ANOESES, ANOESIS, ENOSES/NOESES, NAOSES, NOISES and ENOSIS/NOESIS.
I also group words with alternative spellings together in my mind. DALETH, DALEDH, DALED are all words for the fourth letter of the Hebrew alphabet (Hebrew letters generally have at least one alternative spelling). Often if one is on a board, one of the others will be there as well.
Learning the meaning of some exotic words is a good way to create new word sets and chains. A ROJI is a Japanese tea-garden. When I see that I look for RIOJA, then JAR/RAJ, JAI and JIAO.
Words with really odd sets of letters or odd letter orders can really prove valuable on occasions. For example, SDEIN, SDAYN, SDEIGN and SDEIGNE are all SOWPODS words, being archaic forms of DISDAIN. The oddness of such words can make them easier to remember as they tend to stick out in your mind.
Learning lots of obscure word trivia is a proven method of familiarizing yourself with words. Semordnilaps (words that can be reversed to spell other words) are one example of useful word trivia — REWARDER/REDRAWER, DIAPER/REPAID, IMAGED/DEGAMI, LAMINA/ANIMAL, DESSERTS/STRESSED, REKNITS/STINKER, etc. If you know your semordnilaps, you won’t miss REKNITS next time you find STINKER on a board. Lists of famous people’s names that are in SOWPODS, lists of words of Yiddish or Welsh origin, and that sort of thing can also be very useful in improving your Boggle vocabulary.
Learning words with unusual letter sets is important, starting with all of the vowel-only (EUOUAE, EUOI, AUA, AUE, EAU, AIA, etc.) and all of the consonant-only words (PHPHT, CRWTH, CWTCH, TSKTSK, PHT, BRRR, BRR, CWM, etc.), and moving on to slightly less unusual letter sets. On tough boards these are the sort of words that can win games.
It is worthwhile to set yourself the target of learning at least two new words (from one of your opponents' lists or from the missed word list) every game. It really isn't that hard and the words add up quite fast. I have dozens of sheets on which I have jotted new words to learn over the last few months.
A tool like Zyzzyva can also be extremely helpful for learning new words. Zyzzyva allows you to set up quizzes for yourself. I used to use Zyzzyva's quizzes to test myself on my anagram knowledge.
Once you have been playing for a while you need to start working on compound words. Words ending with -MAN and words beginning with OUT- would be a good place to begin. There are over 50 6-letter words connected with both OUT and MAN. The list is far, far greater for 7-letter words.
I could say far more about my preferred strategy (about how to vary it when playing particular players or types of boards, etc.), but, as an introduction to the broad contours of my general approach, the above comments are about as detailed as it gets.
Posted by Al at 15:13