Fastest solvers: An experiment

[Deanna48]

I can't imagine why anybody would waste their time on this site cheating...I mean, what's the point? Wouldn't people have better things to do? Bansai and Lacrossesticks (I believe) posted youtube videos proving that these cryptos can indeed be solved very quickly.

[momn8r]

Agree with you on AI. I would gladly check off one of those little "i am not a bot" every time I sit down to play.

[Susan.Root]

I am a very good Boggle player, and I moved to Wordtwist as an alternative. But I'm only a 'low' scorer in Wordtwist. I'm pretty sure that I"m playing against artificial intelligences in this game. I'd love to compare my scores to human players, where I suspect I'd have the same record of high scores i routinely get in Boggle. I imagine cryptograms are similarly readily solvable by AI. I don't mind having the AI's compete, but I'd prefer it were done openly.
--thanks for your statistical understanding!, Susan Root (rootsusan74@@yahoo.com)

[Lurker]

I'm not totally sure, but I think Madbury is Munchlet. I know she also uses MoreMadMunch. I'm also Lurker Smurf. There's a player who usually uses Widowmaker but has about 5 userIDs. I don't see any of them on the current list.

That was an interesting study in statistics. Thanks.

[puffybob]

LLapp -- I struck four puzzles that I knew for sure I had already encountered in the survey. There were about four others that I thought I might have already seen, but I wasn't sure, so I left them in.

Deanna48 -- Thanks! You were one of the names I saw an awful lot of!

[Deanna48]

That is really cool! What a great idea. And so neat to see those results.

[LLapp]

Puffybob, this is so interesting! I love your observations and conjectures, and I especially appreciate your historical perspective from the early days of the site. "If this site had a stadium, it would be named after him." -- I love that.

Should we assume the list of records may include some recurrences of the same puzzle, or did you somehow keep track of recurrences?

[puffybob]

So over the holidays I did a little experiment. I did a thousand cryptograms, and for each one I made a note of the record holder for the fastest solve time. When there was a tie, I counted multiple record-holders (which is why the numbers don’t add up to 1000.) Here’s who held the records:

nanrich 255
bansaisequoia 207
LacrosseSticks 168
munchlet/MoiAussi 111
twskinner 108
will0416 76
Deanna48 68
Descifrador 49
defeaterdefeater 24
Annamariah 20
Altoid701 19
GramBow 18
rotorguru 13
DMVJDCPA 12
azhang941 9
FloatLikeButterfly 6
Lurker 6
kb83 5
wings35 5
shaquille 4
nedzeppelin1453 3
SippyGurl 3
contrary 2
henrietta 2
kayen 2
kgun5 2
LLapp 2
Madbury 2
opallady 2
priestp 2
Unwinding_at_The_Falcon 2
xenia 2
Zachman285 2
4pouncers 1
bassood 1
bellaxocampo 1
caa 1
damsel22 1
debzhaus 1
guavamango 1
jimdgar 1
kat 1
Laura 1
LilyH 1
Nautilus 1
patsy86 1
puffybob 1
tchen 1

A few observations:

1. Of course, not knowing the algorithm, I have no idea whether this was a truly random sample; but if it were, I believe the margin of error would be around plus-or-minus three percent.

2. This really shows how amazing some of our fast solvers are. I barely ever get to see the plane fly anymore, but nanrich apparently sees (or has seen) it on one out of every four puzzles! (And that doesn't even count the records she has set that have subsequently been broken by others.)

3. Until I did this, I didn’t realize just how much bansaisequoia had decreased his presence here over the past few years. Since so many of his records remain intact, it felt like he was still here. But I see now that although he continued to rack up trophies until late 2017, the great majority of his record times came before 2012. It’s amazing how many of them are still there, sometimes ten years later.

Those of us who have been around this site since its early days can remember when it seemed like bansai held the record on almost every puzzle. As I was doing this experiment, I did a little mini-experiment: For a hundred puzzles in a row, I looked at who had the fastest pre-2013 solve time. Two-thirds of the time, it was bansai. If this site had a stadium, it would be named after him.

4. I was surprised at how few ties there were. Out of a thousand puzzles, only 211 had a tie for fastest solver. I guess I thought that as time went on, the fast solvers would all eventually converge on the “fastest feasible time” for each puzzle. But the power of the bell curve is strong, I guess; the great majority of time, someone managed to be the outlier (if only by a second).

5. For counting purposes, I assumed lacrossesticks, LacrosseSticks, lacatorcesticks, and lacrosseballs were the same person. Same with bansaisequoia and banzaisequoia. (His several other pseudonyms — impededredwood is my favorite — did not appear in the results.) Same with JCarlosCuevas and Descifrador, based on what I’ve heard here. If there are other multiple pseudonyms that I should account for, just let me know. (Just corrected this to combine munchlet and MoiAussi.)

6. I suppose it would also be possible, in theory, to gauge how many of these records each player had *per total number of puzzles they’ve played*. But that was beyond what I felt I could reliably do, and is especially complicated by the fact that some players have multiple usernames. Also, let’s face it, weird things have happened over the years. (For example, the site says that GramBow has solved only 61 puzzles, which is simply impossible, given how often she appeared in my tiny sample.) But, for what it’s worth, it does look like twskinner, LacrosseSticks, and will0416 have extraordinary rates of records-held-per-puzzles-played.

7. Of course, I guess Stephen could probably run a script and tell us the exact number of records held by everyone. I’d be curious to see it! Having been playing for ten years, I’ve seen the plane fly more times than I can count, but I strongly suspect that virtually all my records have fallen. A clumsy two-finger typist like me will never be among the fastest. (My one appearance on this list was on one of those new puzzles that hasn’t been played much yet.)

Anyway, I hope you are all enjoying a nice holiday season. Good luck with future solving!