53 vs 54 point word rarity in 5x5

[BoggleOtaku]

Thanks for the reply, Naboka.

For me the best thing that has come out of this effort so far is to give recognition to people that have been close to lifetime active participants in wordtwist and the competitions.

There are other ways to give recognition as well. It's a rich game with many different possible facets and combinations of facets that people like to explore and compete in that aren't directly recognized by such competitions. The lifetime list has narrower recognition than the monthy competitions, so there can be recognition for broader facets at a lifetime scale, at scales between monthly and lifetime (especially divided in three to recognize the mass broadening of allowed words that has occurred twice), in combinations such as high word count and word points combined, etc.

I think it is useful and relevant to look at different scoring formulas to try to draw out different facets and different people otherwise unrecognized directly in official competitions, that are excelling, intentionally or not, at some style of play. Especially those long supportive of the game. Some was of looking at the data may not be super useful (like the unique words I tried, usefulness dampened by the masking of higher point words). But press on, as you wrote, toward "relevancy and usability."

[lalatan]

Now that I figured out how to do this, there are a lot of interest aspects to explore. If you or anyone else have an idea for me of something interesting to look at, let me know.

I was interested in who has the most word point records over 18 years. And then how may word point records are for words only on the list one time in 18 years.

These results look really surprising to me, so I'll have to check my code more, but maybe it makes sense.

Word Point Monthly Competition top 40 most recorded words
All words found in 18 years (count) Unique words found (count)*
ranking 18 years wordpoint.png
*Unique words found are those recorded in all 18 years of monthly competitions
only one single time.

I am not quite sure of these results yet and need to do more checking.
Just as an example, here are the words recorded by one person that appears in both lists above:

Unique Words recorded
ACCOMMODATINGNESSES, ALLOTROPICALLY, AMPHITHEATRICAL, ANATOMICOPATHOLOGICAL, ANIMATENESS, CARDIOACCELERATION, DEMATERIALISATION, DEVALORISATIONS, DISCORPORATES, DISEMBARRASSMENTS, DUCTILITIES, ELECTROFILTERING, EXORCISER, EXTRAVAGANTNESSES, FEROCIOUSNESSES, FLOWCHARTINGS, FORMULARISATION, GLUCONATES, HEPATOTOXICITY, HYDROMETEOROLOGICALLY, HYDROXYTESTOSTERONES, HYPEREXCITEMENTS, IMPERTURBABILITIES, INGRAINEDNESSES, INTELLECTUALISERS, INTENTNESSES, KENOSISES, LEANNESSES, NONDEDUCTIBILITIES, NONSTEREOSPECIFICITY, OSSATURES, OVEREXTRAPOLATIONS, OVERPROVOCATIONS, PERENNIALIZATIONS, PHOTOBIOSTIMULATION, PSEUDOHEMOPHILIAS, QUASISPHERICALLY, RECONCENTRATIONS, RELEGITIMISATIONS, REPUDIATIONISTS, RESTRICTEDNESSES, SCENARIZES, SENSITISER, SUPERFICIALIZING, SUPERWORLDLINESSES, TESTIMONIALIZING, TRANSCENDENTALIZING, TRANSCENDENTALNESS, TUTORIZES, UNADVANCEDNESSES, UNCOMPLAININGNESSES, UNIRRITABLENESSES, UNQUALIFIEDNESSES

What seems strange about the first list is that there are many words I and others have found. But maybe they are lower points than other words that the same people often find and so are almost always masked from appearing in the competition.

I haven't read the forum for a few weeks but noticed this thread today. For sure the better long word players have a best scoring word worth more than those in your unique words list. During my search for long word records I always play a game that has a best score of >=44 pts. It gives me practice on those words and also reduces my APPW average. I've been doing that for years so you'd never see me (or Spike and others in the APPW monthly competition) in the unique words list. I searched my screenshots of words I found for the first time, looking for the unique words you listed. They were all valued at 30-38 pts. Last month there were 62 players whose best word played was valued between 40-53 pts. So using your criteria for unique words, none of them would ever show up if they played any of those unique words in your example.

In light of that I suggest that many who have unique words likely played them before long word players such as those I mentioned started doing our thing. And yes, I also have seen words come and go. I played SINTERABILITIES numerous times years ago. I was surprised by how often it showed up. But I haven't seen it for years now.

[Naboka]

I haven't read the forum for a few weeks but noticed this thread today. For sure the better long word players have a best scoring word worth more than those in your unique words list. During my search for long word records I always play a game that has a best score of >=44 pts. It gives me practice on those words and also reduces my APPW average. I've been doing that for years so you'd never see me (or Spike and others in the APPW monthly competition) in the unique words list. I searched my screenshots of words I found for the first time, looking for the unique words you listed. They were all valued at 30-38 pts. Last month there were 62 players whose best word played was valued between 40-53 pts. So using your criteria for unique words, none of them would ever show up if they played any of those unique words in your example.

In light of that I suggest that many who have unique words likely played them before long word players such as those I mentioned started doing our thing. And yes, I also have seen words come and go. I played SINTERABILITIES numerous times years ago. I was surprised by how often it showed up. But I haven't seen it for years now.

I remember you were doing a crazy number of best words per day for a while. How many were you doing? And were you bothering with the games that had low sum best words?

[BoggleOtaku]

Hi lalatan, great to hear from you.

Your reply focused on word points >=44, available in the great big-word point addition of ~2021.
So I redid the list for just the last 60 months, 2021-2025.

The result was 87 people (including you) having a word point monthyly competition entry all 60 months, along with many others that missed 1+ months.

A total of ~30k words were recorded over the 60 months, 10056 of which were only listed one time (i.e., "unique").

Here are 4 examples of unique word lists:
clip1.jpg clip2.jpg


I think this shows, as you were saying, that there are many more opportunities for unique words at smaller (<44) word points, since there will be many more words in the dictionary giving such points. So if people tend to score in this sweet spot, not trying to hunt the biggest of words, they will get many more unique words recorded. (The unique words might be scored other times by other people but be masked by their best word that would show in the monthly competition.)

*Note: I did find one discrepancy in my generated lists, a unique word that appeared to be listed twice. Results need further confirmation.

[Spike1007]

One thing about finding a "new" long word is that, once it's on the lists (especially if the finder posts about it) is that it's more likely someone else will remember it and find it if it shows up again in a different board. A lot of unique words may not stay unique for long.

[Spike1007]

On a related note, it looks like it should be possible to determine the players who were the first to find particular words (at least if a word was found only once in the month it was first found). Thus, you could check to see who had the most "firsts" over time.

[BoggleOtaku]

It would be a bit like archaeology: monthly competitions would provide favorable, sesquipedaliferous strata. You can find the first recorded finding. It, however, could have been found earlier but poorly preserved. For instance (hypothetically): lalatan finds the previously found 54 point word every month and it is recorded. His other 49 words for the month, all of which could have been firsts finds, are masked. It's similar to the case for unique words--only unique in what is recorded, not necessarily in what was actually found. Maybe 100% reliable for the 54-point word and not bad for 50+, but drops with word points.

[Spike1007]

OK, that's true. Screw that idea.

[Naboka]

So, you're dealing with 5x5 which I found too hard because my fingers, wrist and forearms would get so tired after about a minute and a half that I simply stayed with 4x4.

You've mentioned that many different games exist within the basic one presented as Wordtwist. I had to completely reinvent how I wanted to play--and am still modifying my approach.

Now that I've developed anterior interosseous nerve syndrome, I've been exploring low word density boards. Before, I wouldn't touch a game that didn't have achance of being retired. Initially, with the scaled back approach, I was trying to eliminate all the UR words for the game. Then decided to capture only UR words that scored over 15 points--unless some 6 letter word was the best or longest word. Anyone who's played a while knows how you can detiremine much of this in advance of even starting the game.

Pretty sure that some of the little 6 and 7 letter UR words are just as rare as the very long words. When the latest word dump occurred I was finding loads of scientific and medical terms that were previously unplayable.

Apologies for maundering while trying to discover a point just beyond my grasp. Usually, writing helps me find things I was trying to understand, but not this time.

The genus of this wonderment was a game I just played with coelenterazines, a light emitting molecule.

I've long held that within the ultrarare group there are substrata of common, wide, rare and ultrarare ultrarare words. I wonder if there is a mathematical pattern to this. Some fractal pattern, maybe not as elegant as the fibonacci sequence, but a pattern that underlies all ideas and thoughts, possibly as to rarity.

Fortunately for those reading, I've hit the limit of typing with my AIN.

Screen Shot 2026-01-15 at 11.57.31 AM.png

[BoggleOtaku]

Naboka, first time seeing that word!

I think you are right about rarity. There are different classes of rarity, so then we can have intersectional rarity where membership in more classes increases rarity, maybe in a multiplicative effect. It is pretty fascinating to look at something complex like a plane, car, IC chip, etc. and realize to make it, every part had to have a unique name, or biology every part of a cell, every chemical, etc. Or inspection, every different kind of defect has a name. If the object itself is rare or obscure, then the names of its parts may have such intersectionality.

Quick brainstorming on classes of rareness. Rareness increases with:
1. Length of word
2. Complexity of spelling
3. Difficulty of pronunciation.
4. Obsolescence (no longer needed, e.g. descriptive term of old technology that has been replaced. I guess you could also call this historical distance. Example: damascening
5. Linguistic distance: Borrowing from other languages and not widely adopted or incoroprated. Historic borrowing and loss. Example: When trade opened up with Japan I assume a cluster of rare words was incorporated into the English language as a result and to facilitate that market.
6. Cultural distance. Examples: 32 or some odd names for precipitation in extreme northern countries. Different names for items in different dialects. (-ISE vs -IZE)
7. Specificity of profession: terms for things, phenomenae, theoretical concepts (e.g., REORTHONORMALIZATION). I guess when you get into concepts, you could have increased rarity as the distance from known reality increases. Superstring theory probably has a lot of ultrarare ultrarare terms. (#7 covers speaking precisely as needed to facilitate moving the frontier of knowledge forward as opposed to #9 below.)
8. Rareness of entity or process described.
9. A desire to obscure, hide, protect livelihood, sound sophisticated, build group identity, i.e., do something other than communicate simply and clearly. examples: Cryptology, Gang or Teen slang (skibidi bussin lowkey yeet dab), some dialects (e.g., Cockney in part to foster community identity), patent language. Even professional language in clinical settings (e.g., myocardial infarction).
10. Newness. We were recently posting about mediabuzz words like Snowapocalypse and when do they become common enough to incorporate.

One through three are related to Zipf's law of abbrevation: word frequency negatively correlated to complexity.

Seven and eight are great examples of how structure can form in rare words, with clusters of rarity formed around an uncommon entity or esoteric, recondite concept and its related functionality. Patterned clusters occur in professional textbooks (corpus linguistics yield collocation biases).

Wordtwist (especially with Naboka dare I say) is the ultimate blackhole of ultrarare word collocation bias.