The thing which makes 2048 moderately easy to reason about is that incoming pieces are always "1". However, this game appears to spawn in random pieces, with random point values. I tried to follow the to-and-fro of the design blog in order to find out why this is, and I get the impression it's for avoiding getting the player stuck, but it does have the disadvantage of making planning incredibly hard. The good news is that by reading the blog I finally understood what was going on with the screen shake
off-topic: that "Subscribe to me blog" modal is not only bad UX (imho, of course), but also doubles down on the UX tire-fire by resetting my reading position to the top of the page. It's almost enough to warrant a command-w right there. Just one datum, but it for sure made me sad
I was afraid people would abandon good runs if too long, so started spawning 4 and 8 sometimes to speed it up somewhat. (There's 180-odd entries participating in the compo, so even 10 minutes of play time per entry is a huge commitment for people looking to vote on most / all games.)
Re: Subscribe popup: this is the sole reason I'm looking to migrate from Hashnode! All of their UX "improvements" are egregious, but this one surely takes the cake. I very much agree with your sentiment.
When I joined, it was a nice technical writing-oriented site, but these days it has entirely succumbed to terrible product managers. E.g. their blog post editor's main function is seemingly to upsell their generative AI wrapper.
I'm seconding the "incredibly fun" comment, but I would keep the occasional big numbers. It probably took me an hour to finish the game and I wouldn't want to slow it down further. The little extra bit of randomness/luck it adds is nice too.
When it comes to the UI, I was a bit confused about how the gameover and piece removal works. I guess if there are no legal moves it removes a piece?
Also, the multiple move mechanic where the last piece to move is automatically preselected sometimes tripped me up a few times and caused me to accidentally downgrade a piece.
> a huge commitment for people looking to vote on most / all games.
I wanted to say that I went to vote for your game but they seem to only support GitHub login and their GH app does crazypants shit like "Act on your behalf" versus the much more sane "user:email"[1] or whatever which would prove that I'm a person without granting some rando website "sure, bro, take GitHub actions for me"
So, my HN upvote will have to suffice for Internet Points of Appreciation
Please consider crediting Threes, not 2048 as the "ancestor". The designers of threes have a well-written blog post on how they feel 2048 ripped them off. It's a good discussion on iteration in game design, among other topics.
The combo UI just doesn't work for me. I didn't understand what was happening until reading the blog post, the game just felt broken since it was inconsistent whether a piece maintained focus or not. And even after knowing about it, I was constantly making misplays due to clickin on a unit I wanted to move, and instead having the pre-selected unit capture the clicked unit.
It might be worth trying out a model where the combo piece is highlighted graphically, but you still need to first click on that unit first if you want to move it.
I won my second game :) Then I read the instructions :)
(I think it's worth reading the instructions / design discussion, even for someone that is not going to play the game.)
[spoiler alert?]
I didn't realize the "combo" mechanics, but I intuitively made many combos because I usually used the bigger piece to capture the smaller piece. So, for e it's a success in the design.
Is it possible to win only with knight? My strategy was to keep alive as many queens as possible.
Loved the little twist on the mechanics of 2048, instead of just mindlessly pushing down-left-down-left most of the time, you actually had to think about the proper moves of the pieces. Nice to add a second dimension onto the play, had my rusty wheels turning for a little bit before it bit me.
I'm not colourblind, but found the colours weren't useful at all (compared to 2048). They're just too similar and you need to rely on the numbers, which takes more thinking
Two things, one on-topic and the other off-topic:
The thing which makes 2048 moderately easy to reason about is that incoming pieces are always "1". However, this game appears to spawn in random pieces, with random point values. I tried to follow the to-and-fro of the design blog in order to find out why this is, and I get the impression it's for avoiding getting the player stuck, but it does have the disadvantage of making planning incredibly hard. The good news is that by reading the blog I finally understood what was going on with the screen shake
off-topic: that "Subscribe to me blog" modal is not only bad UX (imho, of course), but also doubles down on the UX tire-fire by resetting my reading position to the top of the page. It's almost enough to warrant a command-w right there. Just one datum, but it for sure made me sad
I was afraid people would abandon good runs if too long, so started spawning 4 and 8 sometimes to speed it up somewhat. (There's 180-odd entries participating in the compo, so even 10 minutes of play time per entry is a huge commitment for people looking to vote on most / all games.)
Re: Subscribe popup: this is the sole reason I'm looking to migrate from Hashnode! All of their UX "improvements" are egregious, but this one surely takes the cake. I very much agree with your sentiment.
When I joined, it was a nice technical writing-oriented site, but these days it has entirely succumbed to terrible product managers. E.g. their blog post editor's main function is seemingly to upsell their generative AI wrapper.
I'm seconding the "incredibly fun" comment, but I would keep the occasional big numbers. It probably took me an hour to finish the game and I wouldn't want to slow it down further. The little extra bit of randomness/luck it adds is nice too.
When it comes to the UI, I was a bit confused about how the gameover and piece removal works. I guess if there are no legal moves it removes a piece?
Also, the multiple move mechanic where the last piece to move is automatically preselected sometimes tripped me up a few times and caused me to accidentally downgrade a piece.
I wanted to say that I went to vote for your game but they seem to only support GitHub login and their GH app does crazypants shit like "Act on your behalf" versus the much more sane "user:email"[1] or whatever which would prove that I'm a person without granting some rando website "sure, bro, take GitHub actions for me"
So, my HN upvote will have to suffice for Internet Points of Appreciation
1: https://docs.github.com/en/apps/oauth-apps/building-oauth-ap...
Deleted Comment
https://asherv.com/threes/threemails/
The combo UI just doesn't work for me. I didn't understand what was happening until reading the blog post, the game just felt broken since it was inconsistent whether a piece maintained focus or not. And even after knowing about it, I was constantly making misplays due to clickin on a unit I wanted to move, and instead having the pre-selected unit capture the clicked unit.
It might be worth trying out a model where the combo piece is highlighted graphically, but you still need to first click on that unit first if you want to move it.
When the board is almost full, it's difficult to know which pieces can move. Perhaps add some green shade to the ones that can move.
Both are on the list of improvements for the final version — couldn't get the fixes in time for the compo deadline.
(I think it's worth reading the instructions / design discussion, even for someone that is not going to play the game.)
[spoiler alert?]
I didn't realize the "combo" mechanics, but I intuitively made many combos because I usually used the bigger piece to capture the smaller piece. So, for e it's a success in the design.
Is it possible to win only with knight? My strategy was to keep alive as many queens as possible.
It's on the page before you start the game. I also cannot find any instructions once the game is started.
It might as well not have any instructions because it really doesn't tell you anything about the rules or win conditions.