Elliot
banner
elliotcsmith.com
Elliot
@elliotcsmith.com
Making stuff. Mostly with computers. Run a lot. Play some video games.

Write some things:
misplacedpixels.com - video game stuff
elliotcsmith.com - product/dev/broader stuff

Currently Director of Growth at Explorate working on supply chain software.
And once again Valve finds a way to take more of my money.
November 12, 2025 at 9:54 PM
Reposted by Elliot
I can’t help but feel this whole humanoid robot thing is folks trying to inflate a new bubble before the AI one pops
November 7, 2025 at 10:11 PM
"It is very much a game about pretending that you’re getting better by getting better stats" - This captures how I felt about this game. Once you worked out something that kind of worked you kind of became an XP roomba. It was fun and all but lacked something to make it really stick for me.
NEW: Fun chat with Ball x Pit mastermind Kenny Sun

- Does it matter how I play? Sun: "It is very much a game about pretending that you’re getting better by getting better stats"
- Max no. of balls on screen: 1,000; (slightly less on Switch; alert Digital Foundry!)

www.gamefile.news/p/ball-x-pit...
Ball x Pit x Interview
GTA VI was delayed today, but that's not as important as a Ball x Pit interview, right?
www.gamefile.news
November 6, 2025 at 11:58 PM
Reposted by Elliot
Nintendo says 84% of the 10.36 million Switch 2 owners out there transitioned from Switch 1.
November 5, 2025 at 3:39 AM
Reposted by Elliot
Ball x Pit is pretty good.
There's a formula that seems to work pretty well, combine a simple and familiar game mechanic with rouge like/lite elements and you've got a good chance at a winner. Balatro was poker, Rent is a Landlord was slots, Ball x Pit is Breakout. As is the trend, these games are simple to pick up and win based on how well they can stack systems and unlocks. Ball x Pit does this pretty well. I played through to the credits and enjoyed my time with it but overall it's a mild recommend. Now that I am done with the levels my desire to keep playing has gone but it was good while it lasted. At the start you have one player and some initial items unlocked. Enemies fall in from the top of the screen and you damage them by hitting them with balls. If they get too close for too long or reach the bottom of the screen you take damage. Your goal is to reach the end of the level, beat the mini-bosses and final boss and repeat. To get the job done you'll pick up experience and level up. You can add multiple special balls to your inventory, each of which has different effects. You can also add items which unlock some passive and active side effects. There's also items called 'reactors' that drop and either give you some free levels or let you combine items. Some items 'evolve' unlocking new skills, others can just be fused together to add their effects and free up room for more items. As you play through more you'll unlock more characters, each with their own quirks. One shoots balls from the back of the field rather than the front, another changes the game to be turn based. There are a couple of duds but overall they're each interesting and keep the game feeling fresh. Characters themselves level up as you use them, increasing their stats and unlocking some side effects for the 'town' that you manage in between levels. The town lets you place buildings that collect resources and provide some general meta progression across characters. I didn't really like this part of the game. I can see what they were going for with it but I found the placement of buildings and the interaction with them more frustrating than interesting. If you beat a level for the first time with a character, you unlock 'gears' which are used to open up future levels. Beat them all and you unlock a new game plus. For the most part, new levels are a ramp up in difficulty and not much else. Some have unique mechanics but for the most part the same strategies work across the board. This is where I think the game falls down a little. Once you've found a combo that works you can replicate it fairly easily with all the characters. There's very little randomness at play so most runs looks fairly similar. It may have been the way I was playing but playing run after run got a little stale. I pushed through to the credits and to hit about 65% of the achievements on Steam. At the time of writing about 9% of players have hit the credits so either I am particularly quick to play through or I'm not the only one feeling the repetitiveness of the game. All this being said, the game is pretty cheap, pretty easy to pick up and fun enough that I was happy I gave it a go. It didn't quite live up to the level I was hoping but there's plenty of praise around the web for it so it's worth giving a shot for yourself.
misplacedpixels.com
October 26, 2025 at 9:43 AM
Reposted by Elliot
Edinburgh castle failing to render, likely thanks to the AWS outage
October 20, 2025 at 8:35 AM
Ball X Pit is good but felt like it only really opened up after unlocking the third map. The first couple of hours felt flat to me and it wasn't until I got a few more characters and unlocks that it started to come together.
October 20, 2025 at 9:32 PM
Reposted by Elliot
Some games are a guilty pleasure
Somehow, without quite meaning to, I have managed to spend around forty hours playing Megabonk. There's a little part of me that's ashamed to admit that. It's full of flashy lights, terrible humor and a fair share of mindless grind. Ultimately though, if its fun, does any of that matter? Megabonk on SteamSmash your way through endless waves of enemies and grow absurdly powerful! Grab loot, level up, unlock characters and upgrade to create unique and crazy builds as you fend off hordes of creatures!STEAM I played and liked Vampire Survivors and a little bit of Risk of Rain 2, this game describes itself a combination of the two. That means a lot of levels, items, enemies and failing a lot. Alongside the gameplay there's a clear effort put into the regular dosage of dopamine. Shiny lights, flashy sounds and enough randomization to keep me coming back for more. I saw somewhere online that the game was akin to Coco Melon for adults and in some ways that feels about right. At least on the surface. Initially I was quite bad at playing. I never quite got the hang of Risk of Rain 2 and in a similar way I struggled to get past the first stage. I'd misjudge timing, pick the wrong combination of upgrades and become overwhelmed. At this point I kept playing out of a desire to get better. Eventually I started to get the hang of things. I unlocked a few items and weapons and started to learn about the game's many stats and systems. It was at this point that the first hints started to show that while there were flashy lights all over the place there had been effort and attention to making the game good. A few more hours and I could regularly make it from zone one to zone two. I knew what I needed to do in order to stay ahead of the difficulty scaling. I started to unlock more weapons and tomes and spent time testing different combinations. At some point I'd hit around 80% of the steam achievements. I am not one to hunt for 100% completion but there were a few left that I could see a path towards. I setup runs with very specific builds and learned more about the game through these constraints. The lights kept flashing but underneath it all there was challenge. In amongst all of this grinding I found myself at the final boss. I was once again promptly killed. I quickly learned that what gets you through the hoards of enemies doesn't always work for defeating the final boss. Each round became a new balancing act. I had to pick upgrades that would both get me to the boss but also help me win. I first saw the game on YouTube on Haelian's channel. Much like the game it's a channel I feel a little odd saying I watch. There's nothing odd about the content but watching someone else play games falls into the same odd space as Megabonk. I enjoy it but feel a little silly saying it out loud. HaelianHello friends, I go by Haelian and mostly play rogue-type games right now. I have a set streaming schedule on twitch, so come by to say hi some time http://twitch.tv/haelianYouTube Eventually I managed to make it through the final boss. Getting there was a struggle. Plenty of runs, plenty of failed attempts. I'd achieved something at that point. Sure you could say that was time wasted. Maybe I should have been watching art house films or playing the latest AAA release but I am okay leaving that to others. After beating the boss and unlocking a few more in game items and characters I reached a point where I felt the fun was waning. There was more to do but I could see the amount of time I would need to get there. At this point I'd had my fun and was happy to move onto something new. Looking through my Steam library there are plenty of games like Megabonk in there. I'll call them guilty pleasures but the amount of guilt I feel isn't really very high. They're not pushing the boundaries of the genre but for the most part I won't play something for that long if it's lacking in depth. On the surface Megabonk looks like a low effort slot machine of fun. Underneath there's clearly been care an attention put into getting it to where it is. It's authentic to itself and that should warrant giving it a chance. Some of this you can feel if you watch the short form content on the developer's youtube channel. It has the same level of slightly cringe humor but there's evidence there of consistent effort, attention and an understanding of how to build interest over time. Vedinadyo! i am a gamedev working on a game called Megabonk. it’s a 3rd person roguelike with a bunch of upgrades, characters, randomly generated maps & more! It’s inspired by games like Vampire Survivors and Risk of Rain 2, so if you like those, you will probably (maybe) like my game Megabonk! also make sure to wishlist it on Steam. or else mr. bones will be sad. you don’t wanna make mr. bones sad. im hugely inspired by other creators on Youtube Dani (rip in piece dani my beloved), Blargis, Barji, and more!YouTube I say all of this knowing I'm not alone. With over 30k reviews on steam and a 92% positive rating this seems to be the guilty pleasure of a lot of people.
misplacedpixels.com
October 19, 2025 at 6:29 AM
I feel like all the talks I go to now about AI for devs are just the same half dozen points over and over. Nobody is really being creative with tools, they're just waiting for tools to get better. I'm half convinced people are just using AI dev tools to say that they are and look like they're hip.
October 13, 2025 at 10:48 PM
I vote for `git comity` which instead of authoring a commit, books a meeting with various people to determine if the code aligns with the strategic direction of the company and current OKRs.

All things are made better which committees right?
concept: git commit wrapper called ‘giacometti’
October 13, 2025 at 12:52 AM
Reposted by Elliot
We are thrilled to announce that our NEW Large Language Model will be released on 11.18.25.
October 1, 2025 at 2:38 PM
Reposted by Elliot
Benjamin Button Reviews macOS

rakhim.exotext.com/benjamin-but...
September 17, 2025 at 3:59 PM
Personally, I think the first 25 OS Tahoes were enough.
September 15, 2025 at 10:48 PM
Other than with Django, I feel like every single time I think "I'll use an ORM" I end up regretting it and wanting to just write SQL.

I spent the weekend battling some subquery syntax with Diesel in Rust with what could have been half a dozen lines of SQL. I probably should have just used sqlx.
September 14, 2025 at 9:54 AM
Reposted by Elliot
Get Good - Is Silksong too hard?
I'll admin, I never finished Hollow Knight. It wasn't because it was hard but I reached a point where the fun waned. Despite that I enjoyed it. Enough that I was always destined to play Silksong all these years later. I'm about fourteen hours into Silksong now and I've reached the early parts of Act 2. It is not an easy game, that's a given. The question though is whether the game is good hard or bad hard. Like the first game, Silksong is about exploration, hard bosses and dying a lot. There are some crazy people finishing the game without dying but for me its just part of the process. I find a new enemy, misjudge its move set and I'm dead. I try run past too many enemies, mess up a platforming section and once again, I'm dead. Most games these days have some sort of failure state but one area where Silksong does feel a little punishing is the benches. Benches provide a pit stop, sit at one and you recover health and set a respawn point. Not an uncommon mechanic but Silksong does two things that are a little rough. Benches are miles from bosses and often cost rosaries, an in game currency, to access. The low frequency of benches and their distance from bosses means that I've spent a lot of time running between a bench and boss fight. On it's own, running back to a boss isn't the worst thing in the world. Combine it with platforming and plenty of enemies and I'm often entering a boss room with most of my health missing. I've seen plenty of complains about this online. The general sentiment seems to be that the world has moved on from run backs and Team Cherry should have as well. Personally I like the challenge. Running back through the levels and enemies in them is a good way to get in some much needed practice and helps to provide a little bit of a break between getting my ass handed to me by the same boss. One thing I would change is putting more benches right after boss fights. I've had a couple of instances where I finally beat a boss only to die to a basic enemy in the next room. That run back is one I could do without. The other mildly annoying thing is having to pay to access many of the benches. I've found myself on one health with just a few missing rosaries too many times, combine that with the fact that the nearest previous bench was probably a while ago and I often end up miles away now with zero rosaries unless I make it back. I still feel like I am pretty bad at Silksong, not as bad as I was at the start but still not great. Some of the bossess have taken dozens of attempts and for a couple I felt more lucky than skilled. I find that those moments are part of the fun. I swing and dash around the fights and sometimes string together the right set of moves to win the day. Other times I don't and that's fine. The more I play the more I do feel like I am getting better, that progression is rewarding. More rewarding than any in game score or hitting the end credits. It takes work to get good and you can feel when it slowly starts to happen. I knew the game would be hard going in, I've spent enough time playing punishing games that I can happily play until it stops being fun and still be happy with the purchase.
misplacedpixels.com
September 14, 2025 at 9:02 AM
From the company that brought you everyone's least favorite issue tracking tool, comes an AI browser I guess. Surely this is just a play for talent.

In any case, if you liked Arc, Zen is pretty good: zen-browser.app
September 4, 2025 at 10:20 PM
The math also works with a combo of delusion folks and folks willingness to ride the wave of BS for their own gain.
they’re so close
September 2, 2025 at 10:44 PM
Assuming this doesn't crash and burn like the AI pins that paved it's path I guarantee this will just be the millionth time we reinvent basic analytics and A/B testing but this time 100x more expensive because now its an LLM that says 'when a user clicks purchase, revenue seems to increase'
Raising VC funding by referencing an imaginary technology in (Psychohistory) Isaac Asimov novel (Foundation) was not on my bingo list for 2025

A bit of a buzz kill they are using this "Psychohistory" to "give you customer insights" instead of predicting the future of mankind
August 27, 2025 at 9:16 PM
In a way this is kind of a bummer. I hope its something really strange and not a sequel/prequel. The game was wonderful but it felt well contained.
Clair Obscur: Expedition 33 is "not the end" of the franchise, director confirms www.eurogamer.net/clair-obscur...
August 27, 2025 at 10:12 AM
*Laughs in Australian*
In 1962, my parents bought a home in the Washington Park West neighborhood of Denver for $25,000.

That is $267,423 in 2025 dollars.

Zillow estimates that house to be worth over $1.5 Million.

This is indicative that housing costs have far exceeded inflation.
August 23, 2025 at 3:49 AM
Reposted by Elliot
you’re so close
August 19, 2025 at 11:38 AM
Reposted by Elliot
Donkey Kong Bananza Review
Donkey Kong Bananza is the best Donkey Kong in a very long time and a strong contender to be one of the best games Nintendo has ever made. The game has been built to provide a pathway to a fun experience for just about anyone. It’s approachable for younger or more casual players while being packed with challenges and collectables to keep just about anyone happy. While Mario Kart made sense as a launch title, Donkey Kong Bananza is good enough on its own to justify the console purchase. At its core (pun intended) Donkey Kong Bananza is a 3D platformer. Players move through levels to collect bananas, beat enemies and make your way to the next level. What sets the game apart is that unlike a traditional platformer, the vast majority of the level is destructible. Walls, floors and platforms can be broken through, picked up, thrown and rearranged. Every piece of terrain you’ve punched, thrown or moved around stays that way while remaining on that stage. Players can reset the environment from the menu or by revisiting the stage later. This destruction gives the game a new canvas to work with. Platformers traditionally have very intentionally placed and spaced environments designed to challenge players with their traversal. When you can punch your way through a mountain, hiding an entrance to a cave loses its meaning. While the game does have some non-destructible terrain it is only a small fraction. This means that the designers of DK Bananza had to build a platforming experience that accounted for the ability to smash your way through the floor. The game knows players can destroy things and has built platforming around that in a way that truly compliments the destruction system. In some cases there are bananas hidden behind indestructible terrain, forcing you to find the right way to get to them. Similarly, there are a few bananas that you’ll stumble upon while smashing through mountains that were clearly designed to be found some other way. Those are the minority. Across the nearly 800 bananas waiting to be collected, most are hidden behind just enough of a challenge that heading “Oooh, Banana” for the hundredth time feel like an achievement. Some bananas are in the overworld, leaning on its terrain or unique gimmicks to make accessing them a challenge. Others are found in small isolated scenarios accessed via the main level. These might be a timed room full of enemies or a unique use of the current layer’s terrain in the form of a puzzle. As you move further through the game, more of the bananas become harder to find and access, ramping up the challenge for those looking for more while keeping enough bananas close to the golden path that you never feel like you need to grind to progress through the main story. Stages in DK Bananza are called layers. To keep the summary of the story spoiler free I’ll keep it simple. Your goal is to move through each of the layers towards the core of the planet. Each layer is its own, isolated design full of unique systems and NPCs. Players can move back to any layers they’ve previously visited quite easily allowing them to move on with the story without worrying about leaving each layer 100% completed. Early layers help introduce the player to the terrain system and basic controls of the game. The design of those layers is simple and more linear which helps when getting used to moving through walls and the floor rather than along a preset pathway. Later layers start to introduce their own unique mechanics. This might be icy terrain that causes you to slide as you run on it or rainbow colored rocks that jet you into the sky when picked up. Challenges in a layer do a great job of utilizing these mechanics and teaching you about them without feeling like you’re stuck in a tutorial. Across the layers there are NPCs. Some provide a shop interface or similar but most are there to bring a little life to the layer. Not a whole lot of the game happens via dialog other than the occasional cut scene. DK Bananza is about exploration and leaning into the fun of breaking the world apart and the lack of forced dialog helps to keep that experience running. While players can skip most of the bananas in the game, they do provide skill points used to unlock and improve DKs abilities. Some of these are nice-to-have upgrades like holding more consumables while others provide things like doing more damage to terrain which would be difficult to skip at deeper layers when terrain becomes more difficult to break apart. The skill points also add new moves and perks to DK’s various Bananza states. These are unlocked as you progress through the story and let DK turn into one of five different animals, each with their own skills and powers. Both of the skill point system and the bananza modes bring some extra flavor to the game. Some challenges, especially when you’re on a layer with a bananza unlock, require you to be in a certain bananza mode but for the most part you’re free to switch between them as you see fit. While these modes unlock things like flying, none of them feel overpowered enough to take away from the fun. Each transformation, as well as the skills they have are limited just enough to keep things in check. Again, the game feels like it has been very delicately balanced to provide something fun in amongst a set of systems that could easily turn a less polished game into a boring mess. These days, well balanced games are few and far between. So much so that during the first half of my playthrough I’d convinced myself that I had to sweep each level for all the bananas and fossils before moving forward. Somewhere around half way I gave up playing that way. I realised that this wasn’t a game I needed to blast through and put away forever. If I wanted to come back and collect things I could. That process might happen over month of casually playing for half an hours, it might happen in a second pass of all the levels. Ultimately though Bananza is a game that is there to maximise fun. There’s no grind unless you want one. You can blast through levels as fast as possible or run around and find all the secrets. Either way there’s a path there that is engaging and fun. Donkey Kong Bananza was made by some subset of the team that made Super Mario Odyssey for the switch. You can somewhat feel that as you play and I think the time they spent building Odyssey helped shape Bananza in a positive way. That’s not to say that the games feel similar but the same careful planning of environments, challenges and balance are present here as well. Looking back it’s been a long time since we’ve had a proper 3D Donkey Kong game. Tropical freeze was a good game but it was side scrolling Donkey Kong for a modern system. The last 3D platforming Donkey Kong was Donkey Kong 64 almost 25 years ago. It’s nice to have a DK game back in this format and lets hope it’s not another 25 years before the next one.
misplacedpixels.com
August 17, 2025 at 5:13 AM
It's interesting to see OpenAI treat 'training new models' as R&D. I would be willing to bet that if they stopped training and releasing new models they wouldn't last long, even if they were profitable in that time.

Training and releasing new models every few months is the business model.
ngl this sounds like a silly thing to say, but the point is, they’re default alive

if they stopped training models today, all VC funding was immediately cut, they would have a sustainable profitable business
August 17, 2025 at 12:02 AM
Reposted by Elliot
This evening, I joined a small group of reporters for a wide-ranging, on-the-record dinner with Sam Altman and some of his top lieutenants. And yeah, Altman said — we're in an AI bubble www.platformer.news/sam-altman-g...
August 15, 2025 at 6:23 AM