You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I never know whether to comment on a closed issue or start a new one, so this time I'm starting a new one. From #386 last week - I reported that the AutoSolver in a former iteration had been better able to figure the keyword, so you revamped the AutoSolver and fixed it (yay!).
This improvement will make us want to revisit how the points are assigned. Now that the AutoSolver can find the keyword on this one, it drops it to about 344, which is now too low, because it loses the 300 points for that. If you look through its work, the keyword is both still not immediately obvious to student solvers and also I think the points might want to be restructured so that the wide bulk of the points are just for taking on a Nihilist cryptanalysis and then if the AutoSolver finds the keyword, a smaller reduction. These generally want to be somewhere on par with a Patristocrat or even a bit higher - when it was giving me suggested points in the 650-700 range, give or take, they seemed about right.
We are getting pretty skilled at these and so here are some touch points for your consideration (all the same solver, who is my best at these). I designed these with a variety of factors so I could start learning as a test writer which factors make them harder and easier (I'll offer what I've learned so far as suggested Guidance for your consideration in a different post).
Quote: Gratitude makes sense of our past, brings peace for today and creates a vision for tomorrow.
Crib: CREATESAVISION
PKey: WONDERFUL
Key: HEART
Block size 0
Time to solve: 13:13
This one has both the longest crib (I was trying to choose one that the AutoSolver could recover the keyword, and it took a crib this long to do it) and also a block size 0, which was a large factor in making it easier than the others, as I was trying to figure out in ideal conditions what is the fastest these can be solved. I need to experiment with one where the AutoSolver can find the keyword and it has a nonzero block size to see what the time would look like on that.
Quote: My dog and I both freak out whenever the doorbell rings but we run in opposite directions.
Crib: DOORBELL
PKey: BEHAVIOR
Key: BONUS
Block size 6
Time to solve: 23:30
Was testing a block size different from the keyword length. Student solver said this one was impacted by low variety of unique letters in the crib.
Quote: What matters is not the facts but how you discover and think about them.
Crib: DISCOVER
PKey: FORMATIVE
Key: STUDY
Block size 10
Time to solve: 18:56
Was testing a block size double the keyword length, as sometimes this makes the AutoSolver's table turn white instead of red although it would be the wrong keyword length.
Quote: Do you train for passing tests or do you train for creative inquiry?
Crib: CREATIVE
PKey: EDUCATION
Key: MUSING
Block size 5
Time to solve: 24:55
chose a longer keyword, but this one ended up impacted by so many of the keyword letters coming from the center of the Polybius table.
Hope you find this helpful. I think one thing you might try is to structure the points more like a Patristocrat - add 500(ish?) for cryptanalysis and then smaller tweaks up or down from there based on keyword size, block length, and whether the AutoSolver can recover the keyword, which is probably worth a value of something but not as high as 300 of the points. It will be really interesting if we can start incorporating some of these other factors that affect difficulty, but at least it would get us materially closer to where these want to be.
The text was updated successfully, but these errors were encountered:
I never know whether to comment on a closed issue or start a new one, so this time I'm starting a new one. From #386 last week - I reported that the AutoSolver in a former iteration had been better able to figure the keyword, so you revamped the AutoSolver and fixed it (yay!).
This improvement will make us want to revisit how the points are assigned. Now that the AutoSolver can find the keyword on this one, it drops it to about 344, which is now too low, because it loses the 300 points for that. If you look through its work, the keyword is both still not immediately obvious to student solvers and also I think the points might want to be restructured so that the wide bulk of the points are just for taking on a Nihilist cryptanalysis and then if the AutoSolver finds the keyword, a smaller reduction. These generally want to be somewhere on par with a Patristocrat or even a bit higher - when it was giving me suggested points in the 650-700 range, give or take, they seemed about right.
We are getting pretty skilled at these and so here are some touch points for your consideration (all the same solver, who is my best at these). I designed these with a variety of factors so I could start learning as a test writer which factors make them harder and easier (I'll offer what I've learned so far as suggested Guidance for your consideration in a different post).
Quote: Gratitude makes sense of our past, brings peace for today and creates a vision for tomorrow.
Crib: CREATESAVISION
PKey: WONDERFUL
Key: HEART
Block size 0
Time to solve: 13:13
Quote: My dog and I both freak out whenever the doorbell rings but we run in opposite directions.
Crib: DOORBELL
PKey: BEHAVIOR
Key: BONUS
Block size 6
Time to solve: 23:30
Quote: What matters is not the facts but how you discover and think about them.
Crib: DISCOVER
PKey: FORMATIVE
Key: STUDY
Block size 10
Time to solve: 18:56
Quote: Do you train for passing tests or do you train for creative inquiry?
Crib: CREATIVE
PKey: EDUCATION
Key: MUSING
Block size 5
Time to solve: 24:55
Hope you find this helpful. I think one thing you might try is to structure the points more like a Patristocrat - add 500(ish?) for cryptanalysis and then smaller tweaks up or down from there based on keyword size, block length, and whether the AutoSolver can recover the keyword, which is probably worth a value of something but not as high as 300 of the points. It will be really interesting if we can start incorporating some of these other factors that affect difficulty, but at least it would get us materially closer to where these want to be.
The text was updated successfully, but these errors were encountered: