This’ll probably be a multi-part post, written over a few days, just becuase I realized that one of the things I deeply wished I’d done while Self Aware Games was in its heyday was evangelize how we did what we did.

And what we did was that we beat one of Zynga’s big cash cows (Zynga Poker) with a team about a tenth their size with a minuscule fraction of their budget. They had the hype, they had the resources, and we ate their lunch.

Not only that, we did it with almost no one ever working more than a 40 hour week. The sole exception was having a handful of the artists stay for an extra few hours to get some critical stuff done one night, and when that was done, we post-mortemed that as a failure of leadership.

Our game (Card Ace: Casino) went on to become #1 on the App Store in its most competitive category, our company ended up getting acquired, and then later, that game was the basis for multiple $1B+ acquisitions.

Again, with almost zero crunch.

So there are two questions I hope to answer:

1.) How?
2.) Why?

And I’ll get to those. They’ll be the next two posts. But first, the thing I want to address is that there is a huge difference between what you do as a founder for yourself, and what you do for the team.

But they’re also very intermingled, and it takes a lot of thought to make sure that your actions match your words, and even more, that your values match your actions.

The first thing to know as a founder is that no one will feel the same way you do about the team/product/company. And unless you are giving everyone a really unusual amount of equity in the company and they have the same decision making power as you, that disparity in commitment is correct. Being salty that people aren’t giving it their all the way you do is profoundly un-self-aware.

They aren’t compensated like you. They don’t have as much of an ownership stake in the company as you. They don’t set the direction like you. Their risk/reward profile is usually totally different than yours, and if you expect people to behave like you behave, when all those variables are different… you’re making a huge mistake.

This is why when I see founders talking about “leading from the front” like it’s some deep thing… fuck you – of course you should be leading the charge. You know what’s harder than living the 24/7 grindcore life?

Leaving the office at the end of an 8-hour day.

I was never the last person in the office. And yeah, what kind of example does that set? It sets the example that I’m leaving because I’m gonna go pick up my kids, have dinner with the family, and you should do that with your family or friends, too.

And then I’d answer customer support e-mails and deal with other stuff until 1 in the morning. Did I tell folks that was what I was doing? No. Did I spam them with e-mail at midnight? No.

My job wasn’t to make people sprint to the point of failure. My job was creating an environment where my team could run at speed forever.

There are a lot of reasons I despise crunch. But the big ones are a.) burnout and b.) (I should have used B for Burnout) single-mindedness.

Burnout’s downsides are obvious, and I’d be surprised if anyone reading this hasn’t experienced it in some form. I give young folks some conflicting advice at the start of their careers. Never work for a company that crunches intentionally, if you can avoid it. But also, work your ass off and put in as many hours as you can. The difference is the company expecting it as a baseline from you – to devote your whole life to it – don’t do that – and choosing to crush every task you’re given and show you can deliver above and beyond what’s expected. That’s you choosing to spend your time to build a reputation.

But the single-mindness part of it is important, too.

The problem with obsessively working is that all you get to experience is your obsessive work. When I was at EA, working on the Sims, I’d get in at 10, and go home at 10-midnight most working days. There was a stretch where I did that 54 days in a row. During that time, I did absolutely nothing else but work. I didn’t have any hobbies, any interests. I didn’t spend any time with friends. I didn’t learn anything, except stuff I explicitly needed to learn for work.

There was a period in 2008-9, where every game’s main character was a white, bald space marine. The baldness was probably because hair rendering wasn’t very good yet, but everyone was a space marine of some kind because every person leading a major game at the time consumed the same media as kids (Aliens, Warhammer, etc.), and spent every other waking moment making and playing games. So they all had the same influences, and they all made the same game. Minor differences, sure – but basically everything converged on this one kind of expression of their interests.

On the Sims, I ended up doing my first real “game design” work revamping the food system for Sims 2 Console from the ground up. I built a huge spreadsheet of ingredients and methods of cooking, and spent weeks coming up with plausible results for every combination under the sun, and a whole matrix of stats for various ingredients and ways they could be prepped. It all was based in real food, and how real food would be prepped, because I loved to cook.

That whole thing came from having the time and mental bandwidth to devote to developing an interest. And bringing that interest to work made the game better.

So much of what we’d call “innovation” is the merging of unexpected and disparate interests. Seeing something in one field you can apply to another – where no one else could come to that realization because you didn’t have the interests you had.

That’s the opportunity cost of crunch. It’s why every game had bald space marines. It’s why sometimes every AAA game under the sun feels the same. It’s because when all your love is games, and all you do is games, you can only make games that are like other games. You need to have a team that has a diverse set of interests and worldviews, and time to cultivate those interests.

End day 1
—–

Turns out, there’s a lot to talk about “how” when I think about how we developed games at Self Aware & my later teams. But let me try to break it down into some bigger chunks:

1.) Continuous Deployment, de-risking releases, how we communicated with users
2.) Clear direction & distributed decision-making
3.) Scrambling the status quo with focused chaos sprints

My favorites are 2 & 3, but by far the most important was #1. When I think “crunch”, it always comes down to deadlines and scope. With games, a lot of times the project ends up talking about launch dates long before the project is done. This used to be a necessity because of long lead times for manufacturing, the importance of holiday releases, and coordinating shelf space and marketing budget in retail stores. Most of that isn’t relevant anymore, and even when it is, it’s probably not all that relevant to *your* project these days.

I wish I could say this was some sort of stroke of bold personal genius, but as with most things, we happened upon this concept a.) as an accident of circumstance, and b.) out of necessity.

The circumstantial bit was simple – in 2009, you couldn’t count on a release date because you couldn’t figure out when Apple would approve your app and release it. Yes, if you were “big and important”, you could figure these things out in advance because Apple would talk to you. But if you were us, you got absolutely nothing from them most of the time. You could “hold back” an app for a release date, but if you tried to predict when your app would get through approval, you’d almost always be wrong. So you either predicted a very conservative date, or you just had to roll with the punches. We chose the latter.

Again, in 2009, getting your app through approval took a [shakes Magic 8 ball] amount of time. Maybe a day, maybe a month. Totally unpredictable, but usually for smaller devs, weeks. At one point, we ran a bunch of calls to third party things when the game launched. An iOS update borked one of those things, and black-screened the game on launch. So the only thing that was making us any money now died on startup 100% of the time. And because we are now also in the future, we learned to feature flag all those things. But at the time, this was absolutely fatal. The only option was to update the app, and hope that Apple would approve it quickly.

We happened to have a very, very big gun we could fire at the problem. Someone very high up at Apple who we had a unique relationship with, but we could only fire that bullet once in our entire existence. This was that time.

The app was approved and updated in two hours. But we knew we’d never, ever be able to use that option again (I can tell you the story IRL, but not here, sorry!) – so we realized as long as we were beholden to the App Store approval process, we were never safe.

This led to a radical restructuring of our game, and one that was extremely high risk at the time. We completely revamped the app so that any iOS native stuff was in the client, but then literally everything else got yoinked out of the app, and reimplemented as HTML5, which at the time, was getting to a point where it could do what we needed for the games we were building (both Fleck and Card Ace Casino).

This meant that everything in the game, from the UI to the tuning to the gameplay, could be updated with a server push. Since Self Aware’s core mission was multiplayer social gaming (at a time when “social” for us meant *actually something you wanted to play with your friends*, and not just spammy dark pattern manipulation thanks Zynga), everything was always online anyway.

Which meant, aside from maybe a twice-a-year client update to keep up with Apple’s updates & added functionality, we no longer had to go through the app approval process. We could launch anything, change anything, whenever we wanted.

It’s hard to overstate how scary this was at the time. We were sure that if Apple found out what we were doing, we’d be ejected from the App Store, and the business would be doomed.

But we had to do it, because we’d seen the unacceptable alternative. And as we began to experiment with what was possible, we realized that this structure was an unbelievable advantage.

The key was that rolling out an update was easy. The super-key was that rolling back an update was also (relatively) easy.

Think about every release that you’ve ever done, and how risky it is. The risk is “how much person-time worth of work is going into this release, and how much can go wrong?” The riskier your releases are, the more stuff that can go wrong, the more you have to play defense. The more conservative you have to be with your experiments.

When you get to a point where you can release any time you want, and you can roll back any time you want… you can do anything.

We did entire top-to-bottom revamps of the UI. We added new games. We changed how the mission-critical purchase flow worked. We’d hotfix problems. We’d roll back features. We’d launch something, find a problem with it, roll it back, and then re-launch it in a single day.

We made all kinds of mistakes, but we learned so much, so fast, that it was unlike any other kind of development I’ve ever experienced.

Compare and contrast:

1.) After we were acquired, we had a change to make to the purchase flow for Casino. At a time when it was making about $500K/day. The acquiring company wanted to make sure it’d be safe! So throughout the feature’s development, they had reviews with PM’s, the lead PM, the lead engineer, designer, team lead (no longer me). Weekly meetings for months. When it was ready for release, it was tested to the ends of the Earth. Finally, months and months later, it was released. Made no difference. No negatives, so in that sense, “success”. But also, no positives.

2.) When we were operating the way I operated, we’d design and implement the feature – probably would take a week or two. When it was done, it’d go through normal testing, and then release on a Tuesday. We’d watch the data, and if it dipped, we’d roll back. If we screwed up anyone’s experience or caused any downtime, we’d reward players who raised any issues, sometimes literally giving out free stuff or trigger sales for everyone. Usually this would cause an increase in revenue. But we’d know in an hour or two if it was working.

The thing you may not be thinking about is that option 1 did have a cost.

Five to ten people in a room for an hour every week for months adds up. Particularly when the people in that meeting are expensive. It costs in money, at that point to the tune of tens of thousands of dollars, and it costs in opportunity. Both for the people, and for the company.

Most of those people didn’t need to be in most of those meetings. The four to six months it took to roll out that feature was four to six months of not learning.

Before the acquisition, we could have iterated on that feature 20-30 times, each time learning something significant. After the acquisition, we learned once.

De-risking releases builds speed. It builds confidence in experimentation. It builds an environment where the question isn’t, “How safe is this?” It’s “How cool would this be?”

In the time between 2009 and 2010, Zynga released two major updates to Zynga Poker. By the end of that time, they went from #1 to #2. We release every week, often multiple times a week. We made massive changes to every part of the game, building not just poker, but the first unified casino app, many world’s first features, entirely new types of slot machines that no one had ever made before, and we went from #100+ to #1.

Being able to release continuously has two other things worth talking about, but this post is getting super long.

1.) No crunch. If a feature misses its expected date, there’s another potential release date next week. Something is gonna release this week regardless, so if you need more time, take more time. No pressure to finish if it’s not ready.

This isn’t the same as “no pressure slack off whatever.” If you were late because you were lazy, you’d know about it. But games are inherently unpredictable, and building something new cannot be scheduled definitively. Acknowledging that, and acting appropriately, created respect for the process. We didn’t ask for bullshit dates and we never held you to them if you were working diligently, which everyone did.

2.) You have to communicate carefully to your players. You cannot promise things. Early on, in my enthusiasm for what was coming, I’d often tease features that were in the works. This would invariably backfire, as things slid their schedules. I think one thing that was interesting was that because I was the one who teased stuff to the audience, and I was the one in charge, and I was the one who caught all the shit from the users when stuff didn’t go out…

..no one was more aware of how stupid this was than me.

Which meant that I learned quickly not to do it, and was able to push back any time someone asked/demanded foresight into what was being released. Causing the negative consequences, and having them land squarely in your lap? That’s the best kind of feedback. Very impactful.

So I learned quickly, we don’t tease features. We talk about them when they’re done. Not even hours before. There were so many times when something was set to go out, a total sure thing, and then it’d be pulled at the last second because something went wrong.

And every time someone got pissed about it, it was my fault because I was a blabbermouth. So I’d have to apologize. But I did my best to be the wall that stopped the pressure from the players from getting to the team. Me feeling bad was my fault, because of my actions. It wasn’t because the feature was pulled, because pulling the feature was the right call.

Continuous deployment’s benefits are huge. So huge I cannot imagine developing in any other way anymore. Honestly, even if I was making a boxed product, this is how I’d develop it regardless. Maybe will talk about this more in the future.

End Day 2
—–

Gonna jump around a little to write about one of my absolute favorite “tools” as a team lead – the chaos sprint.

I started doing chaos sprints when I was running the team at Self Aware. The goal of these sprints was to drop everything we were working on, and as a complete team, focus in an absolutely ludicrous way on a single problem for a very short, very intense period of time.

Sometimes it was an internal problem – that we were missing some tool, or that some particularly set of problems had backed up and gone unaddressed for too long. Sometimes it was that a specific metric was weak and we needed to fix it. But the point was that it was one simple, measurable problem that everyone could understand. Let’s say “D1 retention needs to be doubled for this game to survive.”

We’d completely halt the roadmap, and everyone would stop what they were doing. We’d get together, I’d talk through the specific issue, things we’d maybe tried in the past, why we needed to address this, what it meant to players, etc.

Then we’d reorganize into as many small, cross-disciplinary teams as possible, with a bias to grouping people who had never had the chance to work together before. You’d have the entire morning to brainstorm with the team, and figure out what you were going to work on, and we’d get together after lunch to share ideas.

After lunch, you’d just get to it. Some teams needed more thinking time, some teams would get straight to prototyping and building.

The goal was to have something playable by actual users by the deadline. Initially, we gave teams a week. So by the following content release, you’d have to have something you could ship to players.

This forced a really dramatic focus on scope. What can you do in a few days + QA that will move the number in a positive way? Working with different folks gave you more exposure to ideas you hadn’t heard before. Andrew Klose, for instance, would always have ideas that were shockingly evil, which was hilarious because was normally such a thoughtful, player-considerate person. But these sprints brought out a wild mercenary side of him that was delightfully strange. Seeing new groups of people get rattled by him was always hilarious.

So, super tight scope, new teammates, harsh deadline. The week would end with everyone sharing with the whole company what their team had done. (PS: cross-disciplinary didn’t just mean devs. It meant marketing, social media, etc. The works. Ideas from everywhere.)

And then at the end, we’d ship it, test the ideas, and see how it worked. Every single time something surprising happened. Usually something surprisingly effective, sometimes surprisingly educational, sometimes delightfully idiosyncratic that we’d never otherwise have done. Always excellent. Not everything worked, that’s fine.

“Failures” were never failures. That phrase, “You never lose – you win or you learn” was really applicable here. Everything was public, everything was supportive, because no one knew what would stick.

Stuff that didn’t move the needle usually got pulled, and sometimes reworked.

Over time, we’d shorten the sprint length. What started as a week eventually hit my favorite duration – one working day. This is such a bonkers time requirement to ship something to players that the ideas get really weird, which usually means something very interesting happens. The one-day sprints usually aren’t complete things – but they often became the starting point for incredibly creative things.

Needing to work with new people under ludicrous time constraints was both a bonding exercise and a way to break everyone’s habits. People who were new to the process often hated it – it was scary and unnerving – extremely uncomfortable. But that’s the point.

After going through the process a couple times, I think people generally came to really enjoy it. The positive benefits become obvious pretty quickly, and the new relationships and practice focusing on what’s important carried over into day-to-day work.

The general cadence for this became roughly twice a year – that meant the main roadmap didn’t get shaken up too often, it was a change – if everything is chaos all the time, injecting chaos sucks, and there were usually enough new people that exposing those folks to more people usually happened at the right time.

This was one of my favorite tools as a team lead. The combination of positives was huge, the results were delightful, and it was a way to break up problems that everyone knew existed but would never make it on the roadmap, which was also super satisfying to the team.

At my last job, where one of the things we did was create a lot of small experiences, these sprints led to a lot of the best things we ever made.

Super fun. If you want to try this, you have to start with a small problem everyone can understand. You have to not dictate solutions at all. You have to have a very clear and short deadline. You will have to help teams get through the brainstorming part, unless they’ve done it before. You will have to correct direction, and initially, help them manage scope, because this is usually shorter & more intense than anyone is used to.

So it’s a work, for sure. But well worth it.

Woo. Pretty late on this update. Better late than never, I guess.

After Japan, we went to Korea. Obvious from the title. But the fun thing with this leg of the trip was that we were going to meet up with even more people. While the Japan leg, we traveled with one family we’ve known for ages, in Korea we met up with a second. And also met up with another friend – she was traveling with another group of 8 people we knew, but the timing didn’t work out to meet up with the rest of ’em. But there were a lot of us in Seoul at the same time!

Though it was our second time in Seoul as a family, we ended up staying in a pretty similar place, since it was really convenient for a lot of sight-seeing. Not exactly the same place, but within walking distance.

Yeah. We were there with everyone who’s visible in this photo except the person in orange in the way back.

We went back to the “secret garden” in the big palace in Seoul, since it was a really memorable place.

We also went to an LG Twins vs. NC Dinos baseball game. Korean baseball is bananas. Intense crowd interaction, tons of cheering, unique chants and songs for nearly every situation & ever player. It was wild. Super fun. US baseball is really, really disappointing in comparison.

It was also really hot. Not “insane Japan summer at its worst” by a longshot, but it was definitely … hoooot.

While many folks went to a K-pop concert, the kids & I went to a place that we wanted to go to on our first trip, but never made it to: Heavy Steak. It was surprisingly good/cheap! I’d definitely go again. I don’t really get this ad, though:

We then went to Aquafield Goyang, which was probably the most memorable thing we did in Korea this trip, IMO – it was a pretty spectacular experience, and something I’d do again, and recommend to anyone. Basically, BIG SPA, tons of saunas, super pleasant place to hang out & relax, good food.

We then went to Jeju, and stayed by Hamdeok beach, which was really pleasant.

We went on a fishing trip with our whole gang. Most of us caught a bunch of fish, but the kids & I weren’t as lucky. I managed to snag one, right at the end, but that was it. I think everyone else was got multiples. 😀 Poor Ei-Nyung was sick through this whole section of the trip, and motion-sick on the boat to boot. Rough days for her. 🙁

Still, a little exercise helps.

Jeju is known for its black pork, and this restaurant was apex black pork. Best meal I had in Korea. We intended to go back the next day, but unfortunately, they were closed. 🙁

We also went to the Haneyo museum, which documented Jeju’s women divers, who harvested things like sea urchin – all free diving in the ocean. Apparently super dangerous, and a really tough life. There are still a few folks living that life, and we even saw some of them in the water after visiting the museum.

One of my favorite meals was at Mom’s Touch – a Korean fast-food joint, which had a partnership with Edward Lee, who’s risen to some level of fame through a show called Culinary Class Wars. The fried chicken was astounding.

I managed to catch whatever it was that Ei-Nyung had, and the flight home was honestly one of the most miserable 10 hour stretches of my life. Which was wild. The rest of the trip was incredible, so it’s fortunately easy to overlook this when thinking back on it.

I think overall, the thing we’ve learned from this trip is that for both Japan and Korea, we don’t need to stay in Seoul, Osaka, Tokyo, or Kyoto again any time soon. We’ve “had the experience”. So if we go back, the goal will be to stay in more rural or otherwise different places, and try to experience different parts of the country.

Wonderful to travel with friends, though, and share some new experiences with them. I hope we get a chance to do this more often in the future!

We went back to Japan. We wanted to go again in part because we had friends who wanted to go, and travel with friends is fun, but more, because we wanted the kids to not just see a place for the pure novelty of it, but “get used to it” a bit, so that they can be in foreign places and have it feel more natural. To make the things that are shocking the first time in a place feel like, “Oh, people live here, this is just a thing.” This time, we started in Tokyo, and stayed in Asakusa, near Senso-ji and Tokyo Skytree.

One of the first things we did was to swing by a Jins glasses shop, and order some progressive glasses. They take a week to make, but relative to the US price, even with insurance, they’re quite inexpensive. They also have an almost totally automated prescription check that’s done by machines, and it’s very good.

Had Monja-yaki for the first time (it was okay). It’s a weird dish. It’s like a mishmash of cabbage and instant noodles in a thin batter? A derivative of an Okonomiyaki, I suppose. Then it cooks on a griddle until it gets cooked – maybe crispy at the edges, but otherwise, it’s the food I’d most consider describing as “slop”. It’s tasty. I’d have it again. It’s a sort of neat social thing. But attractive? Nooooo.

We went to Tokyo Skytree (we didn’t go to the viewing deck, as it was pretty cloudy), saw a random fire show, didn’t buy $60 grapes, did buy some tanghulu (the candied strawberries), walked through Harajuku to Shibuya and ended up at a rooftop garden, which was really nice.

Saw a LaFerrari, which is I think is one of the most $$$ cars I’ve seen in person (maybe a McLaren P1 or a Ferrari F40 beat it, but it’s up there), and then walked through Akihabara, usually super crowded, but we happened to be there on a Sunday, and they close the main street. Walking through what is normally incredibly busy, incredibly crowded areas with the streets closed was GREAT. It felt very 28 Days Later for a bit, and for a place that’s normally utterly packed with people, casually walking down the middle of a huge street was a very disorienting and strange experience.

The next day, we went to Ueno Park, met up with a friend for Unagi Hitsumabushi, which was super delicious. The dish starts out as just eel over rice, with some shiso & seaweed, but it evolves over the course of the meal. You eat 1/4 of it, then add some wasabi & green onion and eat another 1/4, then pour dashi over another 1/4 and eat it like ochazuke, then eat the last 1/4 however you preferred. It was delicious.

We also met up with our friend/ex-housemate Brandon, who had an astonishing story about a trip to French Polynesia, where he met up with a one-eyed pirate on a 65-foot yacht owned by an old Jewish librarian. Great times. He seems to be doing well, and enjoying life. What a wonderful evening.

The next day, we went to the Imperial East Garden (raining, but still very pretty) and then Teamlab Borderless – an astonishing large art installation. On our previous trip to Japan, we went to Teamlab Planets, and it was as eye-opening as the Louvre was, to me. Totally changed how I perceived art, and was the standout highlight of the trip (aside from all the “significant family” stuff). I enjoyed Planets more than Borderless, but Borderless had a few really killer moments. One of the funnier exhibits had you draw fish, and then the fish would be projected on the walls swimming around. The “borderless” part of it is that the art from certain “contained” rooms would bleed out into the rest of the exhibit, so at some point the fish swam out & around the main halls, which was pretty rad. I liked the more linear/structured format of Planets – it lets each exhibit sing on its own, and most of the big projection stuff, while neat for a few moments, didn’t really do much for me. Still, that last pic is neat, because in the one where it looks like the name is written over J & K’s heads, the other view is what it really looks like when you’re not at the exact right angle.

Then we went to a touristy Samurai & Ninja museum, where they talked about the history of samurai & ninja – even though it was a very intro experience, we all learned stuff, and it was fun. Got to throw some shuriken at foam targets, and the next day, we went to go see a statue of Saigo Takamori in Ueno Park (we were there for the Tokyo National Museum, and so we swung by since we’d just heard of him).

Back to Senso-ji at night (very pleasant, very quiet, very beautiful! If you want to see the temple, don’t even bother going during the day, nighttime was where it was at. Then to the museum, then to the statue of Saigo. Then to Kappabashi kitchenware street, where we looked at sample foods, and I bought a knife.

Didn’t get a Japanese kitchen knife last time we were here. While the “traditional” thing to do would be to get something carbon steel, which holds a very sharp edge but is prone to rusting, I went with a stainless, because frankly, we’re not great at high-maintenance stuff. This is a blade made of powdered high speed steel, which (after looking it up) is a super bizarre way of making very, very strong steel that’ll hold an edge for years.

But because it’s incredibly hard, it also takes really long to sharpen when you need to sharpen it, and I’d be loathe to sharpen this thing on the knife sharpener we have, which is good enough for the knives that we have but probably not good enough for this. 

I haven’t cut anything significant with it yet, but I did cut open a plastic bag, and even that was like, “wait, did I miss the bag?” because it just didn’t feel like … anything. I wanted something different than we have, and so this is a different form factor (a bunka, which is like a santoku but where you have a very sharp pointed tip), and with a different steel (ZDP-189, I think most of the other knives we have are VG-10).

The next day, we left Tokyo and hopped on the shinkansen for Kyoto. Grabbed an ekiben (train station bento), which was shockingly delicious. That evening, we walked around town a bit on a scenic street (Sannenzaka) with a gajillion tourists. Kyoto is legitimately overrun in parts with tourists, and I felt bad at times contributing to the problem. I have no idea how you’d deal with it, but it feels unsustainable, and the Louvre closing this morning due to understaffing & overcrowding leading to an employee strike… it feels like there’s gonna be a lot more of that coming.

Tourism is clearly one way for Japan to economically survive its degrowth, but at the same time, it’s made certain parts of Kyoto basically uninhabitable for locals, which is nuts. I suppose folks have said the same kinds of things about Venice for years, but this is the first time I’ve experienced it, where I look around and think, “Wow, this is actually really destructive.” Nishiki Market, Fushimi-Inari – the popular places were … too much. Yeah, we were here, too, and we were part of the problem. I hope we acted respectfully enough, but it is what it is.

The next morning, we had a private tour of Nishi Hongan-ji temple. Our friends go to the SF branch of this particular Buddhist sect, and through some personal connections, one of the folks here said they’d give them (and us) a tour. It was AMAZING.

I think in large part because this space is so restricted, there are things in it (such as that gold room) that are original from the 1600s that still belong in the context they’ve always been in. We got to walk around some of the interior gardens, and there was no one else there. It was pretty wild. Particularly given the overcrowding everywhere else.

https://en.wikipedia.org/wiki/Nishi_Hongan-ji

So, yeah. Totally amazing. Saw things few people ever get to see with a guide explaining the history of it all, and how it’s been used for hundreds of years. Pretty strange to have walked, apparently, where Tokugawa Ieyasu walked. Got to take some incredible photos of historic stuff, but they asked us not to post them, so I’m not posting them. If you want to see them after the trip, I’ll have them in a photo book we get printed after we get back.

I think the kids are at the age where this doesn’t feel super significant to them, but they’ll definitely remember it, and maybe in a few dozen years, it’ll sink in how privileged a moment it was. 😀

Then we went to Nishiki market, which was jam-packed, and ate lunch. Walked around some more – went to a guitar store because I wanted to check out what Japanese-made guitars are like (nothing really leapt out at me as something I’d be willing to endure the slog to bring home, but there were a lot of cool things), then we walked back along the river to our hotel in Gion.

The following (very rainy) day, we went to Fushimi-Inari. I assumed that the rain would make it less crowded, and sure, it probably was, but the base area near the “start” was jam-packed, and very slow going. We ended up bypassing the entire lower 1/4 of it, which was excellent, because we got past all the influencer ding dongs stopping the flow of traffic to get their photos, and just got to walk up the path at a reasonable pace.

…and the last day, we went to the Kyoto Imperial Garden – underwhelming, since all the buildings are private, and wandered around. Walked around some of the shopping streets, went to the Kyoto Nintendo store (no Switch 2’s, but I got a set of hanafuda cards), got some delicious abura-soba (which I dropped my glasses into, and then the noodles I was slurping fell out of my mouth in shock, and I *buried* my glasses in greasy noodles. Fantastic.), and some pretty excellent frites.

Next morning, we packed up, hopped on to the express to Kansai airport, had a final meal there, and off to Seoul!

Man, I love CC Lemon.

The first two days at Dirtfish were a pendulum, swinging back and forth between success and struggle. I kept getting caught up in old habits, then getting lost in my head trying to analyze or understand them, and then because of that, being unable to really be in the moment.

All of that changed once we got into the woods.

I think that the earlier things that we did, they were exercises. Lift-turn-wait, lift-turn-brake, pendulum turns – all of these were done in isolation on a single corner, and through repetition, you’d improve on your technique. But it also meant that you could do it “wrong” and you could do it “right”. So you practiced to try to get it “right” every time. Or at least, that’s what it looked like it should be.

In reality, one of the things that the instructors try to get you to understand is that every corner is a new challenge that includes some element of the unknown. Cars push gravel out of the way, so every time you go around the corner, even if it’s the same corner, the surface is a little different. Sure, you get more consistent, and you get better able to understand & adapt to the traction. But you never really get it right.

So being trapped intellectualizing what I did wrong was a mess, and the wrong approach. Trying to understand what I did wrong, sure – but that the thing I needed to get out of it was the feel. And the other thing was that it’ll always have room for improvement, the main thing is, once you’re through the corner, there’s another one coming.

We took these exercises and linked them all together into a course called the Boneyard. It’s only 6-7 turns around a small pond, but it’s all the things you learned in practice. The problem for me is that the Boneyard is a course, but it’s also determined by the cones on the course – and so I kept “driving the cones” trying to look at them, but not at the road. So I’d use them as reference, as though I was on a track, but when you needed to do things kept shifting, so I’d do everything too late, and end up in the slop (which is very slow).

Instead, I should have been looking at the road, and braking earlier and (while firmly) less. Giving myself more time to react if things went unexpectedly – but always watching the road, not the cones. If I don’t end up directly next to the exit cone on exit, that’s alright. But still, for my time in the Boneyard (and the Link, which is basically Bonyard+), I thought of it as a series of exercises.

Then we got to courses that weren’t a series of exercises, but a whole lot of linked corners, and more “driving” not just “corner corner corner”. And there, things started to feel like something. You got to drift out of a slightly under-rotated corner on the gas, and it was fine. You got to correct oversteer, and maybe it wasn’t the fastest way out of the corner but it sure was dramatic, and it was also fine. This is where my brain started turning off, and where I’d just started to establish enough unconscious attachment to the new (not track) way to drive that I got some better results.

But my brain didn’t fully “turn off” until the last track – The Mill Run. 2.6 miles, no demo run, you only learn it through the instructors giving you rally-style pace notes. Because you don’t really know what’s coming, you get a note (Right 3), and I had to scan the road to figure out where the next corner was, knowing it was a 90 degree right. There was no time to dwell on mistakes. The next thing was coming! And because of that, I finally was able to just drive by feel. Better corners, more aggressive, earlier on the power.

In retrospect, there were a lot of corners where I didn’t get on the gas fast enough, or took the wrong line through the corner and ended up bogged down. But this was where it went from “challenging” to “mindbogglingly fun” and all the things that you’ve learned over the last few days come together into something that feels like magic.

Everyone in the class made this kind of progress. I think the non-track folks made steadier progress & didn’t have their confidence smashed the way we did. But whether a 17 or 67 year old, every single person was kicking up gravel sliding through corners at speed by the end. It was absolutely wild.

I’m going again. Probably the 3-day RWD course in 2026. If you’re interested, drop me a line.

One of the many things that everyone struggled with at DirtFish was looking where you were supposed to go. It’s really surprisingly difficult. On the track, you look at corner entry, then the apex, then the exit. At least at the level I’m at, you’re generally looking at the next thing you’re concerned about. So when you’re looking for your braking point, it’s in front of you. The apex? In front of you. Maybe a little to one side, as the car is turning. Sometimes dramatically so, but because it’s the track, and because cars can only rotate so fast, you look… more or less in front of you. Corner exit? Same thing. You know where you’re headed, you let the car glide out to the outside, and you’re looking forward to the next corner.

Compare that with the pendulum turn. Prior to getting to the corner, you’re looking at everything. Where am I going? Where is the corner apex? How wide is the road where I need to turn? It doesn’t seem like that much more stuff, but if you’re going left, you’re looking for the apex and exit of the corner to your left, but also the edge of the track to the right. Because you’re approaching the corner on the inside (on the left for a left turn), and since you’re going to turn right, first, and boy, you don’t want to drive off the road. So as your car starts to turn right, you naturally look right. I mean, who doesn’t watch where the car is going. You, that’s who.

Instead, you look left as you steer right. You feel the car start to turn, and once it’s “weighted up” on the left wheels (since it’s turning right and slowing), you steer hard to the left, and let go of the brake. Yeah, you’re steering in a direction you’re not looking, and as the car starts to turn you let go of the brake.

The car’s weight is totally upset, the load shifts from the compressed left, essentially hops over the center line of the car, and slams down on the right, suddenly loading the front right tire (which is turned left), and the car dramatically turns to the left. When the car is in the direction you want it to go, you give it a ton of gas, it squats, unloading the steering wheels, and shoots out of the corner in the direction you’re pointing.

It’s a really disconcerting experience. But as you improve, you realize that the edge of the track on the right isn’t going anywhere. So if I look at it as I approach, great – I don’t need to look at it again. By looking left, I let my body unconsciously give it the right amount of steering input, and I can focus on the things I was struggling with – brake pressure and a more “sturdy” application of throttle.

But since the whole of the DirtFish course for me was a pendulum, I kept “looking right”. I kept second guessing how much brake. How much throttle. The whole course was set up safely – if I “failed”, I wouldn’t hurt anything. But the combination of ingrained “smoothness” on the pedals kept coming back. And I kept falling back on this instinct – looking right – when I needed to look left.

Even in the last runs of the 3rd day, I was second-guessing brake application. Repeatedly, my instructor would say, “You had the right pressure on the first application, but then backed off!” I looked right.

But one of the really genius bits of the DirtFish curriculum is that they teach you a handful of tools. How to turn in a variety of conditions, in a few different styles. How to correct it if you’re too tight. Too far out. They drill that into you in the “exercises”. Then you start driving the short courses – the Boneyard or the Link, which force you to use these tools in applications, not isolation.

I still struggled with these. I’d hit a lovely turn correctly, and then my brain would back off, and I’d smoothly enter the next corner and slide wide.

Then I’d look back at the mistake I’d just made (look right) and get lost.

It didn’t happen to me, but one funny technique that the instructors had was that some of them would point where you should look – in the pendulum turn, they’d point to the left – right in front of your face, blocking your view completely if you were looking to the right. That is, you were approaching a corner at full speed, and they would put their hand in front of your eyes, pointing to where you should be looking. Completely blocking your view.

On day 3, one of the things they did forced me to entirely turn off my brain. It was the first time it all clicked.

The defining arc of my time at Dirtfish was the Pendulum.

The one skill I picked up almost instantly and intuitively was the “Pendulum turn”, or the “Scandiavian Flick”. I got it right the first time I tried it, and could do it repeatedly. Not perfectly, of course, but to within a pretty recognizable and repeatable margin.

The problem was that for not pendulum turns, my skills developed… like a pendulum.

Trying to learn to apply the solid brake pressure you needed to get the car to turn and then wait for it to turn – I’d think really hard about it, and get to a point where I’d be able to do it. I’d get up in my head trying to analyze what was working, what wasn’t, etc. and often “get lost” trying to figure it out. In the end, it was pretty simple.

When I was conscious about it, I’d realize that I needed deliberate brake pressure – when braking on gravel, out of 10, I’d need to go from 0 immediately to 6 if I wanted the car to turn. If I rolled on, track-style – 0/1/2/3/4/5/6, by the time it got to 6, the car would be settled, and wouldn’t turn. I’d also sometimes apply the right pressure, but then chicken out. So what I needed was 0>6. What I’d either do is 0123456, or 0>6>4. If I did either of the latter, the car doesn’t turn. But I’ve trained myself to do the former. And that if I did 0>6, that was wrong and I should back out of it, thus 0>6>4.

So. I’d consciously work my way to 0>6. And I’d get it. The car would turn. It’d be awesome. I’d do it a few times, and the session would end. And the next time, I’d subconsciously try to rely on my subconscious. Because hey, we learned a thing! Now we can just apply it! But the thing stored in my subconscious was 0123456.

Because of that, the quality of my sessions was almost always “exert a ton of mental effort to learn a thing, learn it, execute it alright, then come back to try again and have an absolutely terrible time.” This was mentally brutal. It’d shatter my confidence, it’d get me all up into my head, I’d spend so much time trying to figure out what I’d done wrong, because it felt like I was doing things the same (I wasn’t), and it wasn’t until the coach could figure out what I was doing and give me conscious instruction, and then I could get un-discombobulated, and start to properly execute this totally counter-intuitive feeling thing that the performance would come back.

This was much more difficult than it sounds, because you’re trying to judge correct brake pressure, correct steering input, timing of when to lift, steer, brake, and then transition back to the gas all while traveling in an unfamiliar car at full tilt (because if you’re not going full tilt, you’re not generating enough momentum/energy to make it work at all). So for me, by the end, I found my two bad habits were “too smooth/gentle on inputs”, and “often steering/braking too late”, but it took two full days to figure out that those were the issues among the whole sea of variables we were drowning in.

In retrospect, there are two issues. In the moment, there were infinite possibilities no answers, and what felt like a constant surge in progress then a discombobulating, confidence-shattering backslide for what appeared to be no discernible reason.

It was really interesting to me, because like I’d said in the other post, the other track guy not only had a similar issue, but as we talked about it, and a similar confidence response. We’d both swing wildly from “Holy shit that was amazing!” to “What the fuck am I doing why is none of this working?” Sessions would vary from fantastic to abhorrent (at least for me, from my perspective), and my confidence would swing dramatically from right to left. One of the harder things was seeing the non-track folks with less seat time start from the same place (WTF is this lift-turn-brake stuff?) and then grow linearly, while the two of us swung back and forth from good to bad and wonder if we were going to ever make any kind of repeatable progress.

The thing I needed to do, and was finally able to in the end, was to look – not where the car was headed, but where it needed to go.

A few years ago, a friend of mine went to Dirtfish, a rally driving school in Snoqualmie, WA. They raved about it. Then they went again. And again. Then another friend went, for the half-day course, and he raved about it, too. You can probably see from the photo above where this is going.

For my birthday this year, Ei-Nyung got me a gift certificate to go. Kevin and I had been talking for a while about going to the 3-day AWD course, but I kept waffling – it’d be time away, it was expensive, scheduling it when we were both free was never ideal, etc. A few other friends had expressed going as well, but Kevin & I had gotten the closest to making it work, and so we both pulled the trigger, knowing it’d never be perfect. That ended up being a theme of the week, but in a positive way. More on that later.

But here’s the spoiler: I loved it. I don’t think I’m exaggerating to say it’s the most fun I’ve ever had. And it’s $$$$ – no question – one of the things they do is to show you where that money goes in parts & tires & wheels & all the people who make this happen, because it’s not some huge-margin thing. It’s expensive because it’s expensive. So it’s not gonna be accessible to everyone. But if you’re interested, and if you can swing it, I cannot recommend it highly enough. I cannot tell you how blown away I was by how much we learned, how quickly our skills developed, and how what seemed impossible on day 1 was hilarious on day 3.

You Must Unlearn What You Have Learned

I’ve been to a handful of track days at a handful of tracks. I’m not gonna pretend to be some experienced and talented driver – I’m not. But I’m alright. And I’ve spent those trackdays practicing smooth inputs – rolling on and off the brakes, applying throttle smoothly. Steering smoothly. You practice so that your inputs don’t “upset the car”, so that all the tires have maximum grip for the specific thing they’re doing at the time you’re asking the car to do it. I’ve spent a few years working on this, and trying to make it automatic.

This is a huge problem.

On tarmac, you want to brake in a straight line, when you have all your grip dedicated to slowing the car. You then turn in, and want to use as much of the grip as possible to steer (though you’ll roll off the brakes while steering most of the time – as steering consumes grip, you basically hand over the grip you were using to brake). Then you use as much grip as possible to accelerate out of the corner. Lift off the throttle, brake hard, decrease brake pressure, steer.

One of the first things they teach you at Dirtfish is that this doesn’t work at all on gravel.

You lift off the throttle (smoothly, so as not to upset the car). Apply brakes hard to slow. Car pitches forward. Weight goes over the front wheels to provide maximum traction & slow the car. Roll off the brakes & steer. As you do this, the car “settles”, which pulls the weight off the front wheels and rebalances it, but it means now that your steering wheels don’t have enough grip to get the car to turn, and you slide right through the corner.

Instead, you lift off the gas decisively, which shifts the car’s weight forward. Then you turn the wheel, long before you’ve reached the corner. Because at first, nothing really happens. I think this is because as you turn the wheel, the sides of the tires dig into the gravel, and that takes a moment with the car weighted forward. But it’s a terrifying moment. You’ve gone into a corner full tilt, lifted off the gas, not used the brake, then turned. And the car is still going straight.

So the last part is to wait.

You wait for the tires to bite enough that the car then starts to rotate. If you’re used to driving on tarmac, where with a fairly stiff suspension a car will respond instantly this is an incredibly disconcerting feeling. But then it happens – a fraction of a second later, the car rotates. And it will, for some corners, complete its entire rotation without any additional input. You then hit the gas once the car is headed in the right direction. You can do this before unwinding the steering, because by hitting the gas, you transfer all the weight to the rear tires, so the car takes off in the direction it’s facing regardless of the direction you’re steering. Again, unsettling.

But “unsettling” is actually the point. Where on the track, you want a car to be stable – to slowly load a tire and unload it so that you don’t “shock” the tire into instability – the difference between someone gently pushing against your shoulder and someone shoving you – on gravel, because you don’t have reliable traction, you’re constantly shoving the car around to maximize the potential grip for the specific tires you need to “work”. Its constantly unstable, because you need the car’s weight to dramatically move around much faster.

The gravel surface is kind of a damper. It takes all these shock inputs and slows them down. If you did it on a track, there’s no damper, the shock input breaks traction, and you go sliding off. With gravel, without the shock input, you don’t get enough weight transfer to make your inputs effective, and you go sliding off.

This is such a mindfuck when you’ve practiced tarmac driving that it required constant, conscious effort to not fuck it up. And any time my attention wandered for whatever reason – often because I’d done something right and thought I could repeat it with less focus – this bit me in the ass.

The other track guy in the class had the same problem, so it was nice to know I wasn’t alone.

But so the biggest struggle of the entire thing wasn’t that I needed to learn a new skill. It was that I had to unlearn what I’d learned, and then literally learn to do the exact opposite. It wasn’t going from 0 to 100. It was going from -100 to 100.

(Continued in next post)

Recently spent a few days customizing my old track day helmet. Got a trip to DirtFish coming up, and figured it’d be nicer to use my own helmet if possible. And that the old decoration I had on it, which was mostly plastidip and electrical tape, was falling apart. So I basically stripped the whole thing, and did this style of doodling I used to do a lot in high school w/ sharpie. This is a WIP picture, but in the end, I sprayed it with matte clearcoat. It turned out way better than I would have expected. The other side is blue, and it’s got a half-Japanese half-Finnish vibe going on for, you know, reasons.

Other than that, most of my time has been rehabbing an injured shoulder. It’s been since the beginning of March, but finally things are starting to turn a corner. I’d been lax about doing a particular type of lateral raise because it was excruciating, but the PT clarified that the pain wasn’t causing any kind of damage, and until I worked my way up from a very light weight and pushed through some level of pain, I’d be stuck. So the last week, I’ve been diligent about doing even this particularly miserable exercise, and it has made the difference. Hopefully back to being able to swim & wingfoil soon. The weather’s been great, and it’s been torture to not be able to go make use of it.

The end of the school year’s rapidly approaching, which is pretty weird. Feels like only yesterday the kids started their new schools. Now it already feels old hat. Lots of testing going on, and J’s gonna take a geometry course over the summer to try to get to Algebra 2 next year, which will be an interesting challenge. I hope it’s a.) challenging and b.) he enjoys it. I think both of the kids have had a bit too easy of a time at school, so it’ll be interesting to see if we can get him in a place where he feels like he needs to work, but that it’s not crushing.

Still chugging along learning guitar. Still mostly terrible, but definitely getting to a point where I can play things that would have taken me a *lot* longer to pick up. It’s been a satisfying experience, and nice to get back to an analog instrument instead of just bleeps and blorps.

I’ve also been seeing a trainer sporadically – sometimes 2x/wk sometimes 1x – he’s mostly helping with mobility and shoulder stuff, but we’ve finally gotten to a point where I’m doing some stuff that’s *hard* rather than just stuff that is geared towards stability. It’s still comically low levels of weight/exertion compared to what I’m used to re: “working hard”, but I’ve gotta start somewhere and not jump into crazy shit. Also started riding my bike more – not just the eBike, but the regular analog bike bike. It’s strange, because after the eBike, it’s almost more of a mindset change – it’s less “go fast”, and more like “advantaged walking”. Which goes to show how long it’s been since I’ve done it. But it’s good. Weather’s been great, and it’s a way to get some effort, enjoying the outdoors, and not trashing a busted shoulder.

If you talk with me long enough, you’ll start to realize that I’m like a weird robot with only one directive. I can talk about it in different ways – the one-sentence pitch, your “Coke is Refreshing” resume thesis – but the underlying directive is always the same. Make your idea smaller. Smaller than you think. No, even smaller than that. Make it simpler. Way simpler. So simple that you don’t have to think *at all* to understand it.

There’s nuance in this, and it can be very difficult to do. Making something as simple and small as possible requires deeply understanding what you’re *actually trying to do* and *who you’re trying to do it for*, and for blue-ocean problems, those are never actually particularly clear at first.

But whether it’s games or startups, companies aren’t formed by people with small dreams. They’re started by ambitious risk-takers who have grandiose visions for how they’re going to change the world. And so sometimes hearing the message, “smaller, simpler” – in whatever form – lands like a lead balloon.

Thing is, “smaller, simpler” is how you take those first steps toward your grand vision. They’re not replacing the grand vision.

Sometimes new tech enables you to develop in new ways. Big ways. They let you take huge swings at complex things in new ways. Unreal let novice game devs have the power of huge studios. AI lets individual creators make things faster than ever before. But the problem isn’t in the tools, the problem is in the people.

People using a new thing for the first time – they’re still the same. They learn best in small, clear, well-defined steps. They rely on direct, well-timed, *overwhelming* feedback. Give a new user complexity and depth immediately, and they will be lost, often forever.

So smaller. Simpler. Build small things. Ask small questions. Introduce simple concepts, and make them so clear that users can’t not understand them. Sometimes make them a little unusual and wonky, because wonky is really sticky. (Think N*Sync’s “It’s gotta be MAY” – why is that an earworm, when “It’s gotta be ME” wouldn’t be?)

It can sound unambitious. But no one is born running. The best way to learn to run fast?

Sit up.

Start small. Start simple.

I’ve been a fan of Formula 1 for the majority of my life at this point – since reading about the Senna/Prost days in high school study hall to catching back up with it after a few years off in 2009, when I started watching again because J couldn’t sleep, and I’d catch random races in the middle of the night while rocking him (and then K) back to sleep.

The overarching thing about the last decade of F1 is somewhat unfortunate. The rules have been bad.

Just straight up bad.

The problem is that the way things were going, teams were spending so much that there was no sense of any kind of level playing field. Ferrari would run new engines for every session, test constantly at their private track, etc. and it led to a period of both Ferrari dominance, and small teams falling by the wayside, unable to compete. It’s more complex than that, sure, but not a lot more.

So the way that F1’s decided to fix that is to be more restrictive. More restrictive rules. Less testing. Cost cap. Engine freezes. The unintended consequence of all of this have been significant periods of stagnation and dominance. Why was Hamilton able to get a huge collection of WDCs? Because Mercedes had a really powerful engine advantage, and no one else could catch up. Why has Verstappen collected a bunch of WDCs? Because Newey nailed the ground effect era rules, and no one could catch up. Get a rule change right, and there’s so much that goes into building an integrated car around those rules that with no testing & limited cost and engine freezes, there’s no way to build a better car, and once a team has a lead, they run away with it.

On top of that, as teams optimize their designs around the new rules, the intended effects of the new rules fall by the wayside. Ground effect was supposed to help cars follow more closely. Now? They still generate enough turbulent air because clever folks got around the rear wing rules, and cars can’t follow closely anymore.

Point being, all the cost restrictions and restrictive rules mean that cars are boring, races are boring, and worse, they’re essentially static for years. Gone are the days when a backmarker team could come up with an insane concept and achieve some success.

So here’s my proposal:

• Teams are given a fixed budget of $100M for a season.
• Spec monocoque – same for all teams
• Spec engine – yeah, I know this is a bummer, but unless you can develop engines throughout the season, the only thing having different engines does is entrench advantages. So the base engine is a spec design.
• Maximum team size of 150 people across all disciplines
• No wind tunnels, only computational fluid dynamics (CFD)
• Car must fit in a box of X size, which is used to ship the cars from race to race

Okay, so those are the basics. Other than that, the idea is that you can do whatever you want. Engine manufacturers can be involved to modify the spec engines, but they all start from the same place, and have to work within the budget. If an engine manufacturer gets involved, for instance, they devote an employee to the project, which counts against the staff count, and any work the manufacturer does counts against the budget.

Spec monocoque ensures some level of minimum safety for drivers, and also controls certain hardpoints on the car, so everyone’s working with the same sort of basic LEGO connections to the monocoque.

At every race, two weeks before the race, a specific new rule is introduced for that track. These rules can include things like:

• No front wings
• No rear wings
• Grooved tires
• Minimum ride height of 5″
• Smooth underbody floors required

Stuff like that. Essentially, “Here’s a difficult problem to solve, make the car as driveable as you can.” And in some cases, you’ll end up with cars that are very difficult to drive. Unbalanced aero. Awful grip. Whatever. Drivers will have to… gasp… adapt. Those who can adapt the fastest will be extremely valuable. But they’ll still be restricted by the cost cap.

Every race, the cars will change. It’ll be like a new season reveal every race. But because the teams will be working with hugely restrictive budgets, they’re also going to have to be very cost-conscious. Teams will have to make compromises across the season to maximize total potential, not necessarily potential at a specific race.

Because of this, you won’t be able to optimize solutions. Building a car that works in Monaco on old-timey skinny tires, then going to Miami, but having to race with regular tires but no front or rear wings? Yeah, it’s a challenge. Some teams will get it very wrong and be many laps down. But every race will be exciting and weird, and as teams get better at this, you’ll see them coalesce around some basic performance benchmarks, but they’ll never be able to “solve the problem” the way they do with the current formula.

Chaos formula. Every race is exciting. Driver skill matters. It’s a constant engineering challenge. Every race is a new car reveal. Costs are kept under control by specifying the critical components, but otherwise, the rules are much less restrictive, leading to more interesting solutions. Teams can test as much as they want, as long as they can afford it.

I’d watch it.