Chemistry is a diverse science. There are estimates of $10^{60}$ possible stable small molecules. A key aspect is often the "inverse design" - finding target molecules with the properties we want. This concept is often a core research topic -- how to find better antibiotics, pharmaceuticals, solar cells, etc.
The key is to propose a question that must be solved with a molecular structure, ideally in a few hours or days.
Objective functions need to be concrete computable, calculable, or researchable (e.g., in databases) chemical properties; for computational challenges this would entail accepting the inevitable tension that computed values do not always match well with experimental ones.
The concept of synthesis golf centers on the design for how to perform organic synthesis. Target molecules are posted and answers compete for the best synthesis (i.e., fewest steps, highest yield, etc.).
For this set of challenges, the question is more "what molecule is it" rather than "how do you make it."
In general, the questions will be focused on molecules that already exist, but it's possible that some may not (e.g., a new target for a synthesis golf challenge ;-)
The question itself should indicate requirements for answers. That said, there are a few general suggestions.
Answers should at least include:
In all likelihood, more suggestions will be added as these design challenges continue.
The aim isn't to try and increase our users' reputation. There will be at least one bounty during the one month period for which the challenge is "active." This will be derived from the reputation gained from answering the question. Other users may also choose to award bounties as they see fit.
Ultimately, the biggest incentive we can offer is that it should be fun and hopefully give you new insight into chemistry.
There may not be one correct answer, which is why this works, many people can provide what they subjectively think is a good answer based upon the chemistry they know.
At the end of the month for which a given challenge is 'active', the answer will be awarded to whichever answer provides the most convincing solution to the problem. (In many cases, this may be obvious - for example short bond lengths).
Bounties will also be awarded for creativity, etc.
Necessarily, molecular design challenges will be somewhat 'broad', as there are many possible answers rather than a single one. That is, however, the entire point - it requires more than one person to give a definition / copy some values, or perform a search or two.
Based on this, hopefully, no, the questions aren't too broad. So far the concept has been embraced enthusiastically, and questions should have specific constraints.
This should address fears about the overly broad nature of molecular design challenges.
The topic is new in 2018, but there are some questions which are similar in spirit.
2018
In order to avoid there being 'too much of a good thing'/ 'spamming the homepage', it's been agreed that it would be sensible to post these challenges at regular, monthly intervals, much like "synthesis golf."
If you have an idea for a target or challenge, first check that it falls within the scope of molecular design challenges as defined here, then suggest it in chat. If it seems like it would make a good synthesis golf challenge, then we'll use it for a future challenge.
Additionally, there is a related meta thread searching for new targets: Suggestions for molecular design golf problems.
The point of molecular design golf is to get as many people involved as possible, so yes. If you have a proposed design with a few gaps then feel free to post it anyway. It's quite likely that someone will be able to help you out and point you in the right direction to finish!
What should be avoided is the posting of heavily incomplete answers with the intention of editing them at some point in the future. Stack Exchange is clever enough to 'save' your draft answer so you can come back to it, avoiding the need to post and edit constantly.
The goal of these molecular design challenges is to be inclusive. Ideally, you'll be able to use free tools and databases to participate.
Open/Free Databases: