Left examples have the keyword "help" on the OEBP.

BPs should be marked "help" when the OEBP wants most examples (at least on one side) to be helpful (not when just one or two uploaded examples are helpful).

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree just by not using the most convoluted unintelligible examples possible.

Examples that are helpful to people are often not particularly helpful to computers.

Any helpful Bongard Problem has a harder, not helpful version. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make the square-ness or non-square-ness of the numbers easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it may become unclear whether the helpfulness is part of the Bongard Problem's solution.

E.g.: Is the left-hand side of BP384 "square number of dots", or is it "square number of dots that are arranged in a helpful way so as to communicate the square-ness"?

See @seemslike, where examples being helpful is an irremovable aspect of the Bongard Problem's solution.

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree just by not using the most convoluted unintelligible examples possible.

Examples that are helpful to people are often not particularly helpful to computers.

Any helpful Bongard Problem has a harder, not helpful version. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make the square-ness or non-square-ness of the numbers easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it may become unclear whether the helpfulness is part of the Bongard Problem's solution.

Does the left-hand side of BP384 show "square number of dots", or is it "square number of dots that are arranged in a helpful way so as to communicate the square-ness"?

Such nitpicking, pedantic interpretations of BPs might affect how a person sorts them within meta-BPs. For example, if helpful arrangements of dots in BP384 were considered essential, a non-helpfully-arranged picture of a square number of dots would no longer belong left, and the Bongard Problem would cease to be @allsorted.

BPs should be marked "help" when the OEBP wants most examples (at least on one side) to be helpful (not when just one or two uploaded examples are helpful).

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree just by not using the most convoluted unintelligible examples possible.

Examples that are helpful to people are often not particularly helpful to computers.

Any helpful Bongard Problem has a harder, not helpful version. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make the square-ness or non-square-ness of the numbers easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it may become unclear whether the helpfulness is part of the Bongard Problem's solution.

Does the left-hand side of BP384 show "square number of dots", or is it "square number of dots that are arranged in a helpful way so as to communicate the square-ness"?

Such nitpicking, pedantic interpretations of BPs might affect how a person sorts them within meta-BPs. For example, if helpful arrangements of dots in BP384 were considered essential, a non-helpfully-arranged picture of a square number of dots would no longer belong left, and the Bongard Problem would cease to be "allsorted" (left-BP509).

BPs should be marked "help" when the OEBP wants most examples (at least on one side) to be helpful (not when just one or two uploaded examples are helpful).

BP Pages on the OEBP where users are advised to upload examples that help people (by hinting at the solution) vs. other BP Pages.

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree just by not using the most convoluted unintelligible examples possible.

Any helpful Bongard Problem has a harder, not helpful version. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make the square-ness or non-square-ness of the numbers easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it may become unclear whether the helpfulness is part of the Bongard Problem's solution.

Does the left-hand side of BP384 show "square number of dots", or is it "square number of dots that are arranged in a helpful way so as to communicate the square-ness"?

Such nitpicking, pedantic interpretations of BPs might affect how a person sorts them within meta-BPs. For example, if helpful arrangements of dots in BP384 were considered essential, a non-helpfully-arranged picture of a square number of dots would no longer belong left, and the Bongard Problem would cease to be "allsorted" (left-BP509).

BPs should be marked "help" when the OEBP wants most examples (at least on one side) to be helpful (not when just one or two uploaded examples are helpful).

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree just by not using the most convoluted unintelligible examples possible.

Any helpful Bongard Problem has a harder, not helpful version. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make the square-ness or non-square-ness of the numbers easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it may become unclear whether the helpfulness is part of the Bongard Problem's solution.

Does the left-hand side of BP384 show "square number of dots", or is it "square number of dots that are arranged in a helpful way so as to communicate the square-ness"?

Such nitpicking, pedantic interpretations of BPs might affect how a person sorts them within meta-BPs. For example, if helpful arrangements of dots in BP384 were considered essential, a non-helpfully-arranged picture of a square number of dots would no longer belong left, and the Bongard Problem would cease to be "allsorted" (left-BP509).

BPs should be marked as helpful when the OEBP wants most examples (at least on one side) to be helpful (not when just one or two uploaded examples are helpful).

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree just by not using the most convoluted unintelligible examples possible.

Any helpful Bongard Problem has a harder, not helpful version. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make the square-ness or non-square-ness of the numbers easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it may become unclear whether the helpfulness is part of the Bongard Problem's solution.

Does the left-hand side of BP384 show "square number of dots", or is it "square number of dots that are arranged in a helpful way so as to communicate the square-ness"?

Such nitpicking, pedantic interpretations of BPs might affect how a person sorts them within meta-BPs. For example, if helpful arrangement of dots in BP384 were considered essential, a non-helpfully-arranged picture of a square number of dots would no longer belong left, and the Bongard Problem would cease to be "allsorted" (left-BP509).

BPs should be marked as helpful when the OEBP wants most examples (at least on one side) to be helpful (not when just one or two uploaded examples are helpful).

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree just by not using the most convoluted unintelligible examples possible.

Any helpful Bongard Problem has a harder, not helpful version. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make the square-ness or non-square-ness of the numbers easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it may become unclear whether the helpfulness is part of the Bongard Problem's solution.

Does the left-hand side of BP384 show "square number of dots", or is it "square number of dots that are arranged in a helpful way so as to communicate the square-ness"?

Such nitpicking, pedantic interpretations of BPs might affect how a person sorts them within meta-BPs. For example, if the helpful arrangement in BP384 were considered essential, a non-helpfully-arranged picture of a square number of dots would no longer belong left, and the Bongard Problem would cease to be "allsorted" (left-BP509).

BPs should be marked as helpful when the OEBP wants most examples (at least on one side) to be helpful (not when just one or two uploaded examples are helpful).

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree just by not using the most convoluted unintelligible examples possible.

Any helpful Bongard Problem has a harder, not helpful version. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make the square-ness or non-square-ness of the numbers easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it may become unclear whether the helpfulness is part of the Bongard Problem's solution.

Does the left-hand side of BP384 show "square number of dots", or is it "square number of dots that are arranged in a helpful way so as to communicate the square-ness"?

Such pedantic nitpicking could affect how BPs get sorted by meta-BPs. For example, if the helpful arrangement were considered essential, a non-helpfully-arranged picture of a square number of dots would no longer belong left, and the Bongard Problem would cease to be "allsorted", left-BP509.

BPs should be marked as helpful when the OEBP wants most examples (at least on one side) to be helpful (not when just one or two uploaded examples are helpful).

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree just by not using the most convoluted unintelligible examples possible.

Any helpful Bongard Problem has a harder, not helpful version. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make the square-ness or non-square-ness of the numbers easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it may become unclear whether the helpfulness is part of the Bongard Problem's solution.

Does the left-hand side of BP384 show "square number of dots", or is it "square number of dots that are arranged in a helpful way so as to communicate the square-ness"?

It is senseless pedantry. Nevertheless, the difference might affect how users on the OEBP sort that BP in some meta-BPs. For example, if the helpful arrangement were considered essential, a non-helpfully-arranged picture of a square number of dots would not belong left, and the Bongard Problem would cease to be "allsorted", left-BP509.

BPs should be marked as helpful when the OEBP wants most examples (at least on one side) to be helpful (not when just one or two uploaded examples are helpful).

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree just by not using the most convoluted unintelligible examples possible.

Any helpful Bongard Problem has a harder, not helpful version. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make the square-ness or non-square-ness of the numbers easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it may become unclear whether the helpfulness is part of the Bongard Problem's solution.

Does left-BP384 show "square number of dots" or is it "square number of dots that are arranged in a helpful way so as to communicate the square-ness"?

(It is senseless pedantry. Nevertheless, the difference might affect how users on the OEBP want to sort that BP page in meta-BPs.)

BPs should be marked as helpful when the OEBP wants most examples (at least on one side) to be helpful (not when just one or two uploaded examples are helpful).

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree just by not using the most convoluted unintelligible examples possible.

Any helpful Bongard Problem has a harder, not helpful version. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make the square-ness or non-square-ness of the numbers easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it may become unclear whether the helpfulness is part of the Bongard Problem's solution.

Does the page BP384 show "square number of dots vs. non-square number of dots," or is it "square number of dots that are arranged in a helpful way so as to communicate the square-ness vs. non-square number of dots that are arranged in a helpful way so as to communicate their non-square-ness"?

(It is senseless pedantry. Nevertheless, the difference might affect how users on the OEBP want to sort that BP page in meta-BPs.)

BPs should be marked as helpful when the OEBP wants most examples (at least on one side) to be helpful (not when just one or two uploaded examples are helpful).

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

Any helpful Bongard Problem has a harder, not helpful version. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make the square-ness or non-square-ness of the numbers easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it may become unclear whether the helpfulness is part of the Bongard Problem's solution.

Does the page BP384 show "square number of dots vs. non-square number of dots," or is it "square number of dots that are arranged in a helpful way so as to communicate the square-ness vs. non-square number of dots"?

(It is pedantry, but the difference might affect how users on the OEBP want to sort that BP page within the meta-BPs.)

BPs should be marked as helpful when the OEBP wants most examples (on at least one side) to be helpful, rather than when just one or two uploaded examples are helpful.

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

Any helpful Bongard Problem has a harder, not helpful version. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make the square-ness or non-square-ness of the numbers easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it may become unclear whether the helpfulness is part of the Bongard Problem's solution.

Is the page BP384 really about "square number of dots vs. non-square number of dots," or is it "square number of dots that are arranged in a helpful way so as to communicate the square-ness vs. non-square number of dots"?

(It is pedantry, but the difference might affect how users on the OEBP want to sort that BP page within the meta-BPs.)

(Note: "square number of dots that are arranged in a helpful way so as to communicate the square-ness vs. non-square number of dots" would be tagged with the keyword "overriddensolution" left-BP56; the original solution is simpler.)

BPs should be marked as helpful when the OEBP wants most examples (on at least one side) to be helpful, rather than when just one or two uploaded examples are helpful.

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

Any helpful Bongard Problem has a harder, not helpful version. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make the square-ness or non-square-ness of the numbers easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it may become unclear whether the helpfulness is part of the Bongard Problem's solution.

Is the page BP384 really about "square number of dots vs. non-square number of dots," or is it "square number of dots that are arranged in a helpful way so as to communicate the square-ness vs. non-square number of dots"?

(It is pedantry, but the difference might affect how users on the OEBP want to sort that BP page within the meta-BPs.)

BPs should be marked as helpful when the OEBP wants most examples (on at least one side) to be helpful, rather than when just one or two uploaded examples are helpful.

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

Any helpful Bongard Problem has a harder, not helpful version. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make the square-ness or non-square-ness of the numbers easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it may become unclear whether the helpfulness is part of the Bongard Problem's solution.

Is the page BP384 really about "square number of dots vs. non-square number of dots," or is it "square number of dots that are arranged in a helpful way so as to communicate the square-ness vs. non-square number of dots"?

(It is pedantry, but the difference might affect how users on the OEBP want to sort that BP page within the meta-BPs.)

The answer is that a BP page on the OEBP can be what it wants. BP384 is called "square number of dots vs. non-square number of dots". The "help" keyword is just a fluffy administrative tag on top of that, which says the page doesn't want users to submit any examples that are hard for humans to parse; that would clutter it with irrelevant noise.

On the other hand, if a user insisted, nobody would stop them from creating another page for that slightly different idea: "square number of dots that are arranged in a helpful way so as to communicate the square-ness vs. non-square number of dots." This would be tagged with the keyword "overriddensolution" (left-BP568); the original solution is simpler. (If a person were given that BP and were told to sort a new, non-helpfully arranged picture of a square number of dots, it would still seem correct to sort it left.)

BPs should be marked as helpful when the OEBP wants most examples (on at least one side) to be helpful, rather than when just one or two uploaded examples are helpful.

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

Any helpful Bongard Problem has a harder, not helpful version. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make the square-ness or non-square-ness of the numbers easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it may become unclear whether the helpfulness is part of the Bongard Problem's solution.

Is the page BP384 really about "square number of dots vs. non-square number of dots," or is it "square number of dots that are arranged in a helpful way so as to communicate the square-ness vs. non-square number of dots"?

(It is pedantry, but the difference might affect how users on the OEBP want to sort that BP page within the meta-BPs.)

Our answer is that a BP page on the OEBP can be what it wants. BP384 is called "square number of dots vs. non-square number of dots". The "help" keyword is just a fluffy administrative tag on top of that, which says the page doesn't want users to submit any examples that are hard for humans to parse; that would clutter it with irrelevant noise.

On the other hand, if a user insisted, nobody would stop them from creating another page for that slightly different idea: "square number of dots that are arranged in a helpful way so as to communicate the square-ness vs. non-square number of dots." This would be tagged with the keyword "overriddensolution" (left-BP568); the original solution is simpler. (If a person were given that BP and were told to sort a new, non-helpfully arranged picture of a square number of dots, it would still seem correct to sort it left.)

BPs should be marked as helpful when the OEBP wants most examples (on at least one side) to be helpful, rather than when just one or two uploaded examples are helpful.

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

Any helpful Bongard Problem has a harder, not helpful version. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make the square-ness or non-square-ness of the numbers easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it may become unclear whether the helpfulness is part of the Bongard Problem's solution.

Is the page BP384 really about "square number of dots vs. non-square number of dots," or is it "square number of dots that are arranged in a helpful way so as to communicate the square-ness vs. non-square number of dots likewise arranged helpfully"?

(It is pedantry, but the difference might affect how users on the OEBP want to sort that BP page within the meta-BPs.)

The answer is that a BP page on the OEBP can be what it wants. BP384 is called "square number of dots vs. non-square number of dots". The "help" keyword is just a fluffy administrative tag on top of that, which says we don't want users to submit any examples that are hard for humans to parse; that would clutter up the idea with irrelevant noise.

On the other hand, if a user insisted, nobody would stop them from creating another BP page for that slightly different solution: "square number of dots that are arranged in a helpful way so as to communicate the square-ness vs. non-square number of dots likewise arranged helpfully." This would be tagged with the keyword "overriddensolution" (left-BP568); the original solution is simpler. (If a person were given a picture of that BP and were told to sort a new, non-helpfully arranged picture of a square number of dots, it would still seem correct to sort it left.)

BPs should be marked as helpful when the OEBP wants most examples (on at least one side) to be helpful, rather than when just one or two uploaded examples are helpful.

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

There is a potential harder version of any helpful problem that is not helpful. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make square-ness or non-square-ness of the number easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it may become unclear whether the helpfulness is part of the Bongard Problem's solution.

Is the page BP384 really about "square number of dots vs. non-square number of dots," or is it "square number of dots that are arranged in a helpful way so as to communicate the square-ness vs. non-square number of dots likewise arranged helpfully"?

It is pedantry, but the difference might affect how users on the OEBP want to sort that BP page within the meta-BPs.

BP384 is called "square number of dots vs. non-square number of dots". The "help" keyword is just a fluffy administrative tag on top of that, which says we don't want users to submit any examples that are hard for humans to parse; that would clutter up the idea with irrelevant noise.

But if a user insisted, nobody would stop them from creating another BP page for that slightly different solution: "square number of dots that are arranged in a helpful way so as to communicate the square-ness vs. non-square number of dots likewise arranged helpfully." This would be tagged with the keyword "overriddensolution" (left-BP568); the original solution is simpler. (If a person were given that BP and were told to sort a new, non-helpfully arranged picture of a square number of dots, it would still seem correct to sort it left.)

BPs should be marked as helpful when the OEBP wants most examples (on at least one side) to be helpful, rather than when just one or two uploaded examples are helpful.

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

There is a potential harder version of any helpful problem that is not helpful. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make square-ness or non-square-ness of the number easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it may become unclear whether the helpfulness is part of the Bongard Problem's solution.

Is the page BP384 really about "square number of dots vs. non-square number of dots," or is it "square number of dots that are arranged in a helpful way so as to communicate the square-ness vs. non-square number of dots likewise arranged helpfully"?

It is pedantry, but the difference might affect how users on the OEBP want to sort that BP page within the meta-BPs.

The OEBP kicks aside the philosophical difficulty by titling BP384 "square number of dots vs. non-square number of dots". The "help" keyword is just a fluffy administrative tag on top of that, which says we don't want users to submit any examples that are hard for humans to parse; that would clutter up the idea with irrelevant noise.

But if a user insisted, nobody would stop them from creating another BP page for that slightly different solution: "square number of dots that are arranged in a helpful way so as to communicate the square-ness vs. non-square number of dots likewise arranged helpfully." This would be tagged with the keyword "overriddensolution" (left-BP568); the original solution is simpler. (If a person were given that BP and were told to sort a new, non-helpfully arranged picture of a square number of dots, it would still seem correct to sort it left.)

BPs should be marked as helpful when the OEBP wants most examples (on at least one side) to be helpful, rather than when just one or two uploaded examples are helpful.

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

There is a potential harder version of any helpful problem that is not helpful. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make square-ness or non-square-ness of the number easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it may become unclear whether the helpfulness is part of the Bongard Problem's solution.

Is the page BP384 really about "square number of dots vs. non-square number of dots," or is it "square number of dots that are arranged in a helpful way so as to communicate the square-ness vs. non-square number of dots likewise arranged helpfully"?

It is pedantry, but the difference might affect how users on the OEBP want to sort that BP page within meta-BPs.

The OEBP kicks aside the philosophical difficulty by titling BP384 "square number of dots vs. non-square number of dots". The "help" keyword is just a fluffy administrative tag on top of that, which says we don't want users to submit any examples that are hard for humans to parse; that would clutter up the idea with irrelevant noise.

But if a user insisted, nobody would stop them from creating another BP page for that slightly different solution: "square number of dots that are arranged in a helpful way so as to communicate the square-ness vs. non-square number of dots likewise arranged helpfully." This would be tagged with the keyword "overriddensolution" (left-BP568); the original solution is simpler. (If a person were given that BP and were told to sort a new, non-helpfully arranged picture of a square number of dots, it would still seem correct to sort it left.)

BPs should be marked as helpful when the OEBP wants most examples (on at least one side) to be helpful, rather than when just one or two uploaded examples are helpful.

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

There is a potential harder version of any helpful problem that is not helpful. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make square-ness or non-square-ness of the number easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it may become unclear whether the helpfulness is part of the Bongard Problem's solution.

Is the page BP384 really about "square number of dots vs. non-square number of dots," or is it "square number of dots that are arranged in a helpful way so as to communicate the square-ness vs. non-square number of dots likewise arranged helpfully"?

It is pedantry, but the difference might affect how users on the OEBP want to sort that BP page within meta-BPs.

The OEBP shoos aside this philosophical issue by titling BP384 "square number of dots vs. non-square number of dots". The "help" keyword is just a fluffy administrative tag on top of that, which says we don't want users to submit any examples that are hard for humans to parse, because that would clutter up the page with irrelevant noise.

If a user insisted, no one would stop them from creating another BP page for that slightly different solution: "square number of dots that are arranged in a helpful way so as to communicate the square-ness vs. non-square number of dots likewise arranged helpfully." This would be an example of an "overriddensolution" (left-BP568); the original solution is simpler. If a person were given that BP and were told to sort a new, non-helpfully arranged picture of a square number of dots, it would still seem correct to sort it left.

BPs should be marked as helpful when the OEBP wants most examples (on at least one side) to be helpful, rather than when just one or two uploaded examples are helpful.