My Wordle system does its thing

Mary Lou left a week ago to meet Jenny in Athens, Greece at the beginning of a 3-week tour. They'll be gone two more weeks during which I'm about aa free as a person can be. In the past I might have seen it as a chance to go on some secret adventure. Now, I'm content to relax at home. The extreme heat and humidity deters me from spending time outdoors. 

This morning, I beat the Wordle Bot 3-4. The solution was CRANE. My first guess  STARE, left 8 possible answers, because 6 possible solutions had previously been eliminated. I guessed GRAPE, taking out 4 words starting with G and leaving 4 options. I chose CRANE as the most common of the 4. The Bot, however, guessed LEAST, then CRAVE CRAZE, both of which were previous solutions. I would have thought it would go with CRANE second, winning 3-2 but it's algorithm must have determined CRAVE and CRAZE were more likely. My system won the round!

My dream life is very active. The UGA campus is often the setting for my dreams. This morning I was trying to get Jenny, Lauren and Mark to pick me up from a class at a downtown UGA building. I went down some stairs and exited onto the street. I walked around the side through an alley after deciding it probably was safe. I had two cell phones. I used an old model to call Jenny after remembering her number is 202-4554. Somehow I ended up sitting in the back seat of a female stranger's car talking on the phone. The woman started driving and I realized I shouldn't be in her car. I asked her to let me out but she kept going. I was worried she was taking me to the police. "I can get out here," I said. She finally let me out as I said, "I can give you some cash. Is $20 OK?

I recall some fragments from the past few days: abducting a baby tiger, Mary Lou backing into a ditch causing the car to flip trapping us underneath, alternately watching and being in a strange movie about spies and an alien spacecraft. I enjoy having curious dreams.

How the B-52s first played together at my parents' hoiuse

I've told this story many times, possibly somewhere in this blog. I was interviewed last year for a documentary about the B-52s commissioned by the band. My friend Bill King asked me to relate it for a book he's writing. Here it is:

In 1976 I had stopped playing guitar in bands and re-enrolled at UGA as a Psychology major. My friend circle was a loose group of eccentric characters including Jeremy Ayers, Ricky Wilson, Keith Strickland, Kate Pierson, Robert Waldrop, and Rick “the Printer” Hawkins. I had an apartment near campus and rode a bicycle for transportation. I had a job as a ticket agent at the Southeastern Stages bus station on Broad Street. The manager and bookkeeper were Keith’s parent, Julian and Jo Strickland. 

Sometime in the Fall I had nearly completed my degree and was in the process of applying for graduate school in Clinical Psychology. My parents had gone out of town and I had access to the family home at 275 Milledge Terrace and one of their cars. One evening, I had gotten together with Kate, Keith, Ricky and Cindy Wilson and Fred Schneider and we decided to go to Hunan’s Chinese restaurant and have some tropical drinks, Mai Tais, Pina Coladas and a Flaming Volcano, a house specialty. 

Afterwards, we were in good spirits and we returned to my folks’ house. One of the group suggested going to the basement and jamming on some musical instruments I had left over from my band days. I wasn’t up for it, having set my course in a new direction but I was happy for the others to have fun jamming. They went downstairs and I retired to the little study where my Dad had an old Royal typewriter I used to type papers for my classes. I was working on grad school applications while they played around with my gear. A short while later I heard some intriguing sounds coming up from the basement- a nasty rock groove on guitar, conga drum rhythm and Fred chanting “Buzz buzz buzz buzz buzz buzz buzz buzz- KILLER BEES!” accompanied by falsetto whoops from the ladies. They sounded quite good and I smiled at the combination of rock music and clever satirical humorous lyrics. 

 The music had the basic elements of the future B-52s iconic songs. This random occurrence led to the five friends deciding to form a band and begin practicing and developing their early repertoire. I went to some of their practices and was impressed by how quickly they were progressing. I particularly remember “Devil’s in my car” as a favorite of mine. Meanwhile, I was accepted into the PhD Clinical Psychology program at West Virginia University where I started in August 1977 the week Elvis Pressley died.

A week with Lauren

Lauren invited me to visit for the past week as she closed on a lovely 1905 house on Meade Street in Pittsburgh. Mark was at home in Davenport, Iowa so I was the only person staying with Lauren in her rental house on South Euclid this week.

I have had occasional visits with Lauren since she moved from Baton Rouge to Davenport quite a few years ago, and none where it was just the two of us. Mary Lou and I have been to Yosemite, Italy, France and Africa with Lauren and Mark. These were all memorable experiences. I spent 5 days in Pittsburgh with them about a year ago, during two of which I suffered from incapacitating food poisoning. 

Jenny was with Lauren just before I arrived. Jenny has a business event in Greece after which she will be joined by Mary Lou in a few days. They will tour Greece and Turkey for three weeks, quite a long time for ML to be away from me, Maureen and the grandchildren. 

Jenny and Lauren are cat people. Jenny left Manchan (Manny), her beautiful gray British short-haired cat, for Lauren to board during the Greek excursion. Lauren's cat, Bert, is Manny's twin brother. The two of are almost identical and are very photogenic. Their fur is luxuriant, extremely thick and soft. I like to take photos of the cats with my Samsung Galaxy and share them with Jenny and Lauren, who never tire of viewing them. To be honest, neither do I. 

Manny has visited us with Jenny several times and I've taken care of him at least twice when she was traveling. Manny clearly is attached to me. He sleeps in my bed when Jenny is gone and allows me to cuddle him pretty much any time. Jenny misses him and greatly appreciated the photos I took and sent to her each day. 

The week in Pittsburgh with Lauren was very special and meaningful. I toured the new old house, went to the closing and celebrated afterwards with some of her close friends. We went to the Warhol Museum where a Velvet Underground with Nick exhibition was in progress. We had breakfast at Square, coffee at Redstart and steak at Gaucho. I watered the many plants in and outside her rental house. 

Wordle

I've played Wordle over 500 times. To win a daily game, the player must identify an unknown five-letter word in 6 or fewer guesses. After each guess, Wordle uses color to reveal which if any letters of a guess are present in the solution correctly positioned (green), present but not is the correct position (yellow) and which are not in the solution at all (black).

I keep detailed records of my wins (currently 500/507 games or 98.6%), the number of guesses made for each game (1-6 to solve and 7 if unsolved after 6 tries) and the results of direct competition with the NYT Wordle Bot (W 69, L 83 T 102 since beginning to compare). 

Over time, I've developed an elaborate system of play that has resulted in significant improvement in my performance. I always begin with STARE, a good starting word but not among the highest rated according to the Bot which gives it a 95 skill rating out of 99. The Bot always starts with LEAST, a 99 skill rated word. I chose STARE because the highest number of possible words start with S, a large number end in E and A, R and T are very common letters, as well. Over all 507 games, I've failed to solve 7 times; but, my current streak of solving the game in 6 or less is 261 straight.  

The Bot works from a dictionary of over 3,000 five-letter words. After each unsuccessful guess, the Bot culls all the words that are no longer possible in light of the information gained from a wrong guess. It divides the remaining possible words into groups based on shared patterns of letters. According to the Bot, "On average, more and smaller groups mean faster solving." It then uses an algorithm to select the next guess based on its estimate of statistical probability and how many possible words it expects to eliminate. 

In my head to head experience (254 games), the Bot solved Wordle in an average of 3.56 guesses. My head to head average over those games is 3.66. On average, the Bot requires 10 less guesses than me every 100 games (366 to 356).

My method also involves starting with a dictionary of possible five-letter answers. The Bot's analysis of each game includes a listing of every word it considered after the first guess. Checking these lists for words I didn't have on my lists has been very helpful in expanding my dictionary. I've broken my dictionary down into a set of alphabetized lists based on what I learn from my first guess, STARE. For example, I have a list of words that have no letters in common with STARE (e.g., BLOCK). When I find after my first guess the solution has no letter in common with STARE, I choose a word from the "No letters in common with STARE" list for my second guess. With each wrong guess, just as the Bot does, I use the information gained to eliminate more words from the list, reducing the number of possible solutions. Other lists contains all words ending in E, all words with A in position 3 that don't contain STRE and so on. Currently, I have 12 mutually exclusive lists.

My method is an informal version of the Bot's algorithm. The Bot has a major advantage in that its list is more comprehensive. My lists don't always contain the solution because I've failed to include all possible 5 letter English words. When that's the case, I eliminate all the words on my list without solving the game. I then have to figure out what other word or words are possible solutions. In a recent example, the scenario emerging after 4 guesses was

     _ U _ I _ D. 

The solution, HUMID, wasn't in my dictionary, resulting in me losing to the Bot 5 guesses to 3. However, the most common outcome in the competition  (22%) is a 3 to 3 tie. Second most common (16%) is me winning 3 guesses to 4 followed by me losing 4 to 3 (15%) and a 4-4 tie (14%).  Additionally, twice the Bot has failed in our 254 mutual games to find the correct word with 6 guesses. During this stretch, I've solved it each time in 6 or less. 

I do have one big advantage over the Bot. Wordle never uses the same solution twice. To account for this, I keep an alphabetized note with every past Wordle solution. (currently 715 words).  Each day, I delete the solution from my lists and add it to the eliminated words list. Thus, I never guess a previous solution word. Every word I guess might be the solution. The Bot, however, doesn't take past solutions into account and frequently makes guesses I've eliminated that can't possibly be correct. I've had games where I "beat the Bot" because it made several guesses I'd already eliminated. 

I have a second, perhaps smaller, advantage. The Bot can't make a subjective judgment as to which of several possible solutions is more likely. The solution to today's Wordle was HATER. After one guess, I knew the solution contained TARE but not S. My second guess was LATER, then WATER followed by RATER. In the Bot's analysis of my game, it suggested a better guess for turns 3 and 4 was EATER. My subjective judgment was that EATER was less likely than the other two I guessed. When was the last time you used or encountered the word "eater" in a sentence? Maybe "meat eater?" I solved the game on my fifth guess. The Bot solved it in 6 after guessing TATER and EATER and a previous solution, CATER.

A few games back, guessing STARE resulted in R being eliminated, A being in the correct position (3) and STE being present but out of position. This left only 3 possible answers, BEAST, LEAST and YEAST. FEAST was ruled out as a previous solution.

I immediately guessed LEAST in case the Bot solved the game on the first guess. It was wrong so I guessed BEAST next which was correct. The Bot, however guessed BEAST second, defeating me 2 guesses to 3, despite it not having ruled out FEAST. That was OK with me since 3 guesses lowers my average number slightly. The next time this scenario comes up. I'll guess LEAST followed if necessary by YEAST while I expect the Bot to go with BEAST again if LEAST isn't the solution, even after I've eliminates it as a possible solution. 

As time goes by and my dictionary gets smaller, I expect my advantage to allow me to beat the Bot more often and close the gap between our average number of guesses. Even now, my average result is only 0.1 guess more and I tie or beat the Bot two thirds of the time. Stay tuned.

Jungian dream analysis

When I was in my early 20s attending college and playing in the Zambo Flirts, I happened across a paperback copy of Man and his Symbols at the UGA Bookstore. It was a pivotal moment in my life that triggered my interest in the field of psychology and led eventually to my long career as a therapist.  Within a few months I had read The Portable Jung edited by Joseph Campbell and started in on the Collected Works in English translation. I taught myself Jungian dream interpretation and became adept at analyzing my friends' and my own dreams in Jungian terms. 

All of this came back to me this morning when I awoke with a perplexing dream. I was having an argument with Mary Lou who had stayed out all night. We were in a hotel room, perhaps attending a conference. Some of our children were wirh us- we were much younger than now, maybe in our 30s or 40s. I confronted her about staying out. Her attitude was haughty and defiant. She admitted she had been having sex with another man. I kept grilling her with questions and she unashamedly told me the man's name. She intended to divorce me. I was stunned and hurt but not entirely displeased. I started thinking about what would happen and what I needed to do. I tried to look up the man using my cellphone but I couldn't type accurately because the phone didn't work prooerly.

On its surface, the story line depicts the complete opposite of reality. My relationship with Mary Lou is very positive and she is highly appreciative of me. With the stress of Jane's recent death and Maureen's dependence of her for help with childcare, Mary Lou leans on me heavily for emotional support. The dream is also out of line with our personal history. Reversing roles would be more consistent with long past periods of conflict. 

A basic principle of Jungian theory is all the characters and objects in a dream represent aspects of the dreamer just as every element of a story reflect aspects of the storyteller. This concept provides a way to circumvent resistance to awareness of disowned (ego dystonic)/repressed personality traits. I'm short, I am both myself and Mary Lou.

Looked at in this way, Owen in the dream is my more rational masculine dimension whereas ML would be my more emotional and passionate femine self. I recognize this internal conflict easily. Divorce viewed psychologically is the futile attempt to split the psyche. The goal of Jungian therapy is the reconciliation of seemingly opposite inner selves. Imagining a dialog between the two forces is a standard technique for achieving inner peace and wholeness. I will give it a try.