GMPuzzles 2021 Objectives and Key Results: Mid-year check in
2021 continues to be a year of transition and growth for GMPuzzles. Having brought Serkan Yürekli on-board as Managing Editor last year, and getting more support from the rest of our team, we’ve had the opportunity this year to focus on several different parts of our publishing business. At the start of the year I set our some objectives and key results for GMPuzzles. While this was just a set of internal targets, I wanted to write today to share our goals with our audience, suggest how we think we are doing, and get your feedback on anything more we need to be doing to meet our objectives.
Overall Objective: Grow Grandmaster Puzzles to be the best publisher of logic puzzles in the world.
Objective 1 (Consistency): Establish Grandmaster Puzzles as a highly reliable puzzle publisher, for both puzzle solvers and puzzle constructors.
- Owner: Serkan
- Background: In past years, Thomas’ work schedule and other commitments made the release of books and web weeks less certain. Twice we’ve had to “pause” the site when Thomas has taken on a new job. Some effects of this are that there are some promised titles, like The Art of Sudoku 2, that spent years on the backlog to be released, and authors have in the past had to wait awhile for feedback. In 2020, bringing Serkan onboard as a managing editor helped a lot, and the goal for 2021 is to continue to make our publishing process as consistent and reliable as possible, with clear publication plans and enough editorial support to never have another unexpected delay again.
- Key Result 1a: Release 48+ weeks of puzzles and solutions online throughout the year.
- Midyear score: 1.0 — we are on track to release puzzles every single week this calendar year. This will be our first year with that much content since 2014.
- Key Result 1b: Release 12 new e-book collections over the year, one per month.
- This should include individual puzzle packs as well as the launch of a more regular quarterly publication (in style of The Art of Puzzles) that includes content from several of our most popular genres and from a lot of different contributors.
- Midyear score: 0.9 — we will release at least 12 books this year but just missed in May the “one per month” goal due to some delays on one ambitious title. But we have a plan for our next year of e-books, including our next releases in June and July of a TomTom collection and a Sudoku pack with Even/Odd Sudoku and Thermo-Sudoku. We’re planning for the first Grandmaster Puzzles Quarterly collection to come out after that.
- Key Result 1c: Complete implementation of a regular puzzle submission and request system for authors.
- This should be connected to our book plans and web releases, to have a more stable flow of accepted puzzles and published puzzles. We should be accepting about the same numbers of puzzles in a year that we have opportunities to publish.
- Midyear score: 0.5 — Serkan is doing a great job getting regular feedback to contributors who email our submissions address, but we still need to do better on a few aspects of this puzzle flow. First, while we have guidelines for how to submit puzzles to us, we should be clearer on “why to write for us” and “what we are looking for”. We get a lot of unsolicited submissions we have to reject because they do not meet our publication needs or our GMPuzzles style guidelines like being visually/esthetically interesting. While we’ve also planned for our next several books, we have not placed an open call for submissions towards these books yet and expect to do that by end of year.
- For any author with open questions, including status of puzzles or needing to send in an invoice, please email our submissions address to ask for an update. For anyone wanting to contribute puzzles, feel free to also email us to start a dialogue on how your puzzles might fit in our future plans.
- Stretch Goal: Release 3+ backlogged puzzle titles including a collection of all our past patreon bonus puzzles, a collection or our past giant puzzle rewards, and/or expanded versions of older out-of-print books where GMPuzzles has the right to republish (like TomTom Puzzles or Mutant Sudoku).
- No midyear update here yet, but one of my own goals for second half of year when I have time is to finish the Patreon book and then rerelease (with more content) my very first TomTom book from over a decade ago.
Objective 2 (Accessibility): Establish Grandmaster Puzzles as the best destination for logic-puzzle solvers of all levels, from beginners to experts.
- Owner: Thomas
- Background: While our puzzles are high-quality, they are usually on the more difficult end of the spectrum. And by releasing as a blog form, there is no easy way for a new solver to encounter our puzzles and have a good initial experience. Further, even for experienced puzzle solvers, our available solving options (i.e., printable PDFs) limit some solving options. Our goal in 2021 is to continue to improve on our accessibility so that we can grow the audience for GMPuzzles content and make the site an easier destination for new puzzles solvers to find and grow to love.
- Key Result 2a: Complete new web flow / introduction sections for at least 10 main genres at GMPuzzles.
- This should include beginner packs pulled from prior web content with their own solving flow, web or video tutorials to discuss the rules and info and how to get started on the puzzle type, and other appropriate content to improve the experience for beginning solvers.
- Midyear score: 0.0 — Due to later key results (including digital solution options), we have not focused on rebuilding our webflows yet.
- Key Result 2b: Add online solving for 100% of new web puzzles as possible.
- This will begin in “beta” form in 2021 using penpa-edit tool (originally from opt-pan and then extended by swaroopg92), with more improvements to be implemented throughout the year to make the solving tools easier to use and more blended with our overall web experience.
- Midyear score: 1.0 — every 2021 puzzle has been on Penpa-Edit and we’ve added several hundred from our archive to the tool as well, with a new stretch goal to finish all digitization by end of the year. We’ve also set ourselves up in working with Swaroop for active development and updates to our tools based on solver feedback.
- Key Result 2c: Release at least 2 YouTube solving videos per week for our new web content, with some of these videos being suited more for beginners (including filling some of the goals of key result 2a and 2b for more accessibility).
- Midyear score: 1.0 — I am making videos for almost every puzzle we post now, but am considering how to balance these kinds of videos versus some “intro to [puzzle style]” videos or other content to help with key result 2a on accessibility above.
Objective 3 (Scaling): Continue to release high-quality content while exploring new venues for our puzzles to reach a broader audience.
- Owner: Thomas
- Background: Our puzzle content has been top notch from the start, and we want to keep this intact as we scale our company and distribution into additional channels.
- Key Result 3a: Continue our long-standing streak of having no published “broken puzzles” with zero or multiple* solutions.
- (*except in very rare cases where puzzles are meant to have multiple solutions such as the few puzzles we’ve now posted to the site that can be solved as two different puzzle genres.)
- Midyear score: 1.0 — the quality streak continues with no broken puzzles in our history of publication. Yes, errors like small typos in setting up web posts or occasional server issues have happened, but our puzzles have never had an error in them to frustrate a solver.
- Key Result 3b: Contribute GMPuzzles content to select magazines and international puzzle competitions.
- Including at a minimum writing a Puzzle Grand Prix round in 2021 and setting up the team for the 2022 World Sudoku and Puzzle Championships sponsored by Grandmaster Puzzles in Toronto.
- Midyear score: 0.5 — We have completed a Puzzle Grand Prix round for the year which will be held next weekend, but the WSPC plans have been further delayed due to COVID. We have separately started conversations to license some of our puzzles to other magazines/app providers but don’t have anything material to share on this front yet.
- Key Result 3c: Fully support [REDACTED] needs, a project that will utilize our GMPuzzles content.
- Midyear score: [redacted] — can’t share more on this yet, but if our overall goal is to be the best publisher of logic puzzles in the world, this opportunity to bring more of our puzzles to a new audience may be a big step for us in the coming years.
So that is our set of Objectives and Key Results that I shared with our team in early January and that I’ve put some grades on here in June. I shared them here as I want you, our GMPuzzles community, to share your own feedback on how we are doing. And any other ideas of how to improve our site’s “consistency”, “accessibility”, and “scaling” or other key results we should be tracking are welcome too.
Hi Thomas,
Normally I’d be rattling this off as an e-mail, but seeing as you’ve made a very public post I thought it deserved some kind of public reply to help encourage any discussion – particularly as I speak as a member of the team (albeit a frequently errant and absent member).
On consistency, I can’t argue with posting schedules of either blog or e-book. I think 0.5 is a little generous on the puzzle submission front however. Perhaps part of it is that you have team members like me who aren’t particularly reliable in submitting puzzles, but in any case there is certainly a good proportion of guest contributions these days, and I think it’s fair to say that process has much room for improvement, as you say. Anecdotally speaking, I get the impression that lag time between submission and publication is putting people off – although as this is just anecdotal I’ll leave it for others to better describe their experiences. do try to encourage people to speak to you directly about this, although whether people feel emboldened to do so is something I don’t know about.
One point on consistency that none of the above covers is the overall style and difficulty of the puzzles. I think they have been getting harder and more WPC-like. Great for a certain section (the hardcore?) of the GM Puzzles audience, but I do wonder if there’s something to be said for keeping the levels of the puzzles more accessible – how would we know if newer solvers are tapping out by the time they get to Wednesday level for example? I do feel that a well crafted easier puzzle is something that passes by tha majority of the hardcore audience who tend to binge through dozens of such puzzles in one go without breaking sweat,
On Accessibility, there’s been great work done setting up penpa links, but I have to say that Penpa still leaves me cold. It’s a bit like using a 90s windows application – sure, the functionality and flexibility is all there, but usability seems to be there from a programmers point of view rather than how people actually use it, and it’s a steep learning curve picking up cold. The design of the user interface is probably the most 90s thing about it, and I should probably leave it there in case anyone gets the impression I have a grudge (I don’t – as a labour of love it’s clearly a highly impressive piece of work, but I think a commercial project has to hold itself to higher standards). Whilst obviously things like the pzv project don’t have nearly as much flexibility, what you can say is that as well-designed user experience it’s light years ahead of penpa at the moment.
The other point on accessibility is to do with the blog format itself. I’ve previously sketched out how I’d like to see the blog format completely left behind, and to instead have a proper database structure driving the content publication. Nikoli.com (RIP!) remains the benchmark here for me, and again I don’t think I’ve seen anything that has come remotely close.
One advantage to moving away from the blog format might also be to move away from a 1 puzzle per day schedule with greater flexibility – in turn this might help clear a backlog of puzzles and provide a better experience for guest authors. Improved accessibility for the users (solvers) of the site if done correctly will mean better accessibility for other users (authors) of the site.
On scaling: 3a sounds like it belongs as part of consistency to be honest, which we’ve already touched upon as being very good. Re the projects – I’m excited to hear more about this. I would hope that some of this scaling will either necessitate or facilitate (perhaps both!) some kind of a move away from the blog format. Perhaps that’s something very much for the longer term than the things you are talking about, but then again I think it’s worth mentioning regardless in case anyone else also thinks this would be a good idea in general.
Thanks Tom for your response — healthy to see your perspective here even if it is as a less regular member of our GMPuzzles team these past couple years. I hope it does spark some more discussion.
A lot of your points of feedback are part of my 2022-2025+ vision but we can’t do them all overnight. This is where some of the more meaningful measurement and part of my scoring is along the lines of “have we gotten better” versus “have we gotten to perfect”.
In reflection, I think I had let the strawman of a “perfect” digital solving tool keep us from exploring options for too long. I have a quite different perspective than yours about Penpa-edit. It started at a key point for us of being able to support every puzzle we publish, but has also continued to improve more in the last 3-6 months even as we’ve started to use it for more styles and had Swaroop start to act on solver feedback. There is not a “light years” gap in terms of where penpa is and pzv in several of the puzzle styles if you mean the distance left to catch up/pass. Some different things on the surface, a better “trial” mode, some “why is my answer wrong” type stuff, and most of the Penpa UI completely turned off. But without judging the front-end which we haven’t focused on in the beta, under the hood we are giving a nearly identical experience for solvers doing, say, Star Battle where I don’t have a preference of either tool any more. The harder comparison from your comment is between these puzzle hosting experiences and the whole digital solving/commenting/contest ecosystem that Nikoli had, and that is a longer term investment that digitizing our puzzles and putting a user-associated experience behind a puzzle database may eventually get us to in the future.
Key Result 2a is meant to be the first “de-blogging” step in terms of resetting the natural way people interact with our puzzles, and still meant to happen this year. That it scores a 0.0 is where we have more to do in the last six months to catch up. Getting a sense of our prior puzzle inventory and where Penpa-Edit is the closest fit for good solving experience after digitization will probably be how I select the ~10 genres for a new experience path. It will probably also be where we have a more limited set of options in our Penpa-Edit experience that would be the look of “Star Battle v1” and “Tapa v1” and “Sudoku v1” with user design starting from the solver side and not the constructor side and a very different front-end skin. And also without going through the WordPress blog to solve dozens of our puzzles and possibly purchase more. Where at the start of the year I thought I’d be doing more of Key Result 2a and waiting on some yield from Key Result 2b (Penpa), it has proven out in practice that 2b is the first thing and 2a comes after.
Maybe we’re looking at penpa from different angles. Another ability might be a 2000s comparison between iPhones and blackberries. Penpa here is like the blackberry – I’m not saying it doesn’t have missing capabilities, I’m saying that it’s clunky to find them sometimes, there’s a bewildering level of mostly irrelevant functionality for any one given style and generally (not all cases) it doesn’t pass the new user test for picking it up and it just working without any further fuss.
Other solutions are far more intuitive, especially for newer users. You can pick them up and they just work. That’s where I see the light years gap (whilst acknowledging that penpa has made incremental progress nevertheless)
Re: “I do wonder if there’s something to be said for keeping the levels of the puzzles more accessible – how would we know if newer solvers are tapping out by the time they get to Wednesday level for example?”
===============
For what it’s worth, I drifted onto this site via the content at the “Cracking the Cryptic” YouTube channel around the beginning of this year, as I realized that I enjoy solving pencil puzzles a whole lot more than I enjoy sudoku & variants. So I’m one of the “newer solvers” you’re asking about. I try to solve the puzzle here daily, and even if I do I generally watch the solve video the next day to clean up any of the (sometimes frequent) places I just “try stuff” instead of following a perfectly logical path.
Solves generally take me quite a bit longer than the “Expert” benchmark time presented for the hardest puzzles, and I approach this benchmark closely on the easiest if it’s a puzzle type I’ve thought about before. I only rarely tap out on the very hardest puzzles. This last Saturday’s “Plusses” Yajilin bested me fairly close to the finish.
For pencil puzzles I would say I finish 100% of puzzles through Friday and 80% of Saturdays. For number placement puzzles like Sudoku and Kakuru and TomTom, I sometimes lose interest. Although I look at them all, I’m not generally spending much time toughing out the hardest ones. Notable exception, I adored “Pluses and Minuses” on 3/27 and the novelty of that puzzle compelled me to finish it (in perhaps 2.5 hours).
I like the flexibility of the penpa interface. Most of the options are unnecessary, but being able to derive my own notations is sometimes useful. I might be in the minority in my tolerance for arcana, though. I haven’t encountered the other interfaces mentioned.
I would be buying your books if they were digitally solvable, but so far have not as PDFs.
Thanks Brian for sharing your experience of finding the site and solving the puzzles.
It is helpful to hear how you are finding the tools in adding extra options. We do default to a standard (best) solving mode for each puzzle and put anything else we expect is useful in the “tab” options with a lot of features not in the rotation at all. I wonder if you are even breaking outside of the mold in using additional features like shading cells in color for different notations in a sudoku which is a mode combination we’ve never put in the “tab” menu. Our future plan is likely to hide the toolbar options from view unless a solver opts into the full interface, and then properly describe the default interfaces like the composite modes for specific puzzle types better so that solvers know what left/right clicks and so forth do. Considering it rounding the corners of the interface for those who want the style of the iPhone.
I haven’t prioritized digitizing our books yet, as I think that requires a user-specific layer to interact with a puzzle database to be perfect (and not just ship out a set of links with no ability to track what you’ve solved). But if there is sufficient interest in this kind of support we can test it out for a few books. Some styles are far better for digital solving than others so we will probably start from those styles.
I doubt that I’m breaking any molds, it’s quite rare that I use anything more complicated than a variety of surface shadings (or something isomorphic to that — I used the “free line” tool on the Araf on January 29 to connect pairs as I discovered them (it ultimately bested me)).
For me, “a set of links” to penpa would be sufficient to justify the purchase, although there may be understandable access control issues with that.
Re: … and most of the Penpa UI completely turned off.
I agree with this. The experience could be dramatically improved by simply doing less (or sometimes simply reordering things).
In case it’s useful, here are some things that I think would improve the presentation and experience quite a bit. These are mostly surface-level and sorted to put higher bang-for-the-bug up first.
I recognize that you may already have a list with every item here on it. And I can understand if these things just haven’t been the focus so far. But these are my feelings on what would improve the experience the most with relatively little development effort.
1. Put the actual puzzle first on the page, (move any interface controls after the puzzle). It’s a really bad experience to have the actual puzzle partially or wholly hidden and require scrolling just to see it.
2. Drop words like “Title:” and “Author:”. The first things on the page should be the puzzles title, the author, and then the puzzle itself.
3. Eliminate all buttons from the interface that aren’t directly relevant to a designed solving experience for the specific genre and variation.
4. Implement better default scaling. All but the largest puzzles should be quite usaable with a total width that is not larger than the width of a typical smart-phone display. Puzzles should display by default with that correct width and not require scaling to be used. (This is a principle that could also be usefully applied to the gmpuzzles website itself.)
5. Restructure the tool to use very short, (shareable), URLs rather than the miles-long URLs used currently.
6. Present the interactive puzzle in the same place as the current puzzle image on the primary puzzle page, (rather than having two pages linking to each other, one without standard gmpuzzles web style, one without interface for “fave”/”solved”, etc.)
Note that none of the above has anything to do with the actual solving interface itself, simply how its integrated and presented.
I hope the above might be helpful.
-Carl
Thanks Carl — good suggestions here, and some (1, 2 for sure) that can be attempted easily soon when we expect to play with the css on Penpa to make it look more matched to our website, improve fonts, etc. Most of these should be adjustable and instantly seen across all puzzles.
On the blog posts I have gone to short urls (like https://gmpuzzles.com/s/180130GP for one of your Star Battles from 2018) for the Penpa links. But if someone is clicking into the Penpa-Edit pages you get only the long urls and the sharing is not connected to my short url server install. I personally do not like random short links like bit.ly’s, so prefer having a gmpuzzles short link for sharing anywhere the puzzle is viewed and we have part of that in place but should complete the rest.
View scaling is probably more a question for Swaroop, and integration to the other solving widget is probably the one I don’t see a path for except a new solving database built out to map to Penpa-Edit as the original widget is aging in ways that are harder to support.
Thanks for the discussion, Thomas.
I see that penpa-edit has source code available with MIT license, so I can experiment there with trying to improve the view scaling. And I can follow up with Swaroop.
One other comment I forgot to mention that should also be a trivial change: I find the current timer, (updating at tenths of a second), really distracting personally. I think it’d be nice to have it only display seconds, (but could still display the tenths of seconds when a puzzle is completed or the user stops the timer).
Also, the timer should use a fixed-width font so that the digits don’t shift back and forth as the times change.
Bu yeah, further discussion about details of penpa-edit are probably better suited for github rather than this blog comment thread. I’ll pick up the discussion over here for anything else I have:
https://github.com/swaroopg92/penpa-edit
And, big thanks to Swaroop for all of this code! It’s got an impressive amount of functionality.
-Carl
Thanks, Carl, any help is greatly appreciated.
Let’s connect and discuss further either on Github (you already have the link) or alternatively you can email me at penpaplus@gmail.com
Looking forward to the discussions.
Thanks.