TWIG #408: Your Favorite Mobile Game is Lying to You About Ads and Why Claude Won’t Clone Games

Hello everyone and welcome to Deconstructor Fun This Week in Games number 408. I'm Jen Donahoe. I play the marketing and publishing role on the podcast. Today we have in first chair
Eric Kress, strategy investing.
And I'm Laura Taranto. I'm joining not back from the dead, still alive. You can find me on Puzzle Monthly as the sitting Puzzle Monthly Queen.
And Josh Chandley, mobile operator and on DOF UA junkie.
UA junkie, I love it. I call you the UA guru, but you know, I like junkie. All right, so fresh off of last week's prediction that Roblox would crush earnings, well, let's unpack that cuz it didn't quite end up that way. We couldn't go another week without getting into all the decompiling news. We've been meaning to cover the decompiling AdMon video that the guy decompiling a Block Blast and going through all the AdMon stuff. So we found a friend, Josh and Laura here to break it down. Then, on

Monday, Anthropic said that hey, maybe there's an end to all of that because we're not going to let you do it anymore, but we'll get to that. At the end, we're going to sprinkle in some Xbox news because sadly, there are people that have been there for decades that keep leaving and then maybe we have a reason as to why one of the other ones left. And then, if you are an Elder Scrolls fan, you got a name. Congratulations for waiting 8 years. We'll talk about what that is a little bit later. But first, Mishka is doing very important game-related business things and Phil has life news he'll talk to us about next time. So we called in the lovely Laura and the lovely Josh. Laura, what what have you been up to? How are things?
Do you guys know pickleball?
Yes.
It's a very popular sport here. I've gotten really into pickleball. I have a coach.
What do you mean like yeah?
In fact,
Yeah, okay, it's it's that new app. I could use some new partners.
[laughter]
So that

my summer was spent I spent I have pickleballing. So, pickleballing and talking about puzzle and working.
All right. Well, you need to investigate paddle or I think the actual pronunciation is padel, which is pickleball meets racquetball. So, go look that one up and that's going to be your next addiction. That's what the new Everyone in the Deconstructor of Fun Event Universe is into. Mishka with his you know, laying on the field and all that stuff. Anyway, Josh, what are you up to?
Well, if you haven't heard me talking about ad tech enough already, I interviewed Unity's Chief AI Officer Felix Tay on Vector and how things have progressed over the last 6 to 12 months over on my podcast Mobile Signal by Josh Burnsley. But, I I can't compete with pickleball. I have uh I had a great Halloween. My son, who's four, he he broke my heart. Um, we got matching dino costumes for Halloween. And the whole week leading up to it, we were we were having these ridiculous like dinosaur parties where

we both put our costumes on every night and blasted like toddler dino music and just danced around the house roaring. But, on the last day of daycare, I found out that they banned inflatable costumes at daycare. So, he couldn't take it in. And so, I had to go and rush to Target and get him like this tiny Hulk costume. And he fell in love with it.
Aw.
And so, on actual Halloween day at the Halloween party, I was I was the only dinosaur.
Aw. Well, what if you were Spider-Man and you guys could have been a little like break new world duo or something?
[laughter]
It was still fun. It was still fun. But, um, we we didn't have matching costumes. So.
Well, in another kind of fun, uh, Eric doesn't believe in any of that and went to GamesBeat to hang out with the elder statesman. That was That was your own kind of Halloween, right?
Yeah, I mean, I don't want to throw too much shade, but I think GamesBeat's kind of lost a little bit of uh
Oh, no. I love the GamesBeats. It's like

a reunion every time I go.
Yeah, it's just like OG people talking about the old days, right? I think they had a panel of people talking about gaming that haven't been in gaming for over a decade, you know? Like it's like it's an odd thing. I'm not going to name names, but on the mobile panel it was like you know, all everyone was retired, right?
[laughter]
It's like I don't know what they do. But the bigger news is I got absolutely chewed out by Alex de Ky, well deserved for talking trash about General Intuition, um them creating technology for drones that kill people. She assured me that that is not what they are doing. The technology may be capable of that, but it is not what they are working on. There are several industries that are looking at their technology to enable movement in 3D space, and I am certain that they will not fund drones kill people in bunkers. Okay. But I knew it was coming. I I I knew

Alex was going to give me a hard time. I went there anticipating it. That's probably the only reason I went there was to get a hard time by Alex. But anyway, well deserved. Sorry Alex. Good luck with everything.
[laughter]
You're badass.
Yes, it is well deserved that she crushed you and also she deserves to yell at you cuz you said her name wrong. It's Alex de Ky. Whose uncle we need to put in bubble wrap because we are losing all of the Star Trek alumni cast left and right, and so
Oh, Sulu
Right. Please bubble wrap George Takei right now.
[laughter]
Oh my god, that was huge news.
Yeah, we lost Sir Patrick Stewart.
Oh, Picard.
I I I mean, some people debate which is better, Professor X or Picard. He Picard for me is number one. And then we also lost Walter Koenig, who is Chekhov. So we only have George Takei and William Shatner left from the OG cast. I'm a huge Star Trek fan. It was like I grew up watching it with my dad, so like that

whole franchise to me is like one of my you know, cornerstones of my geekdom. But anyway
Yeah, that's really really sad day.
Yeah. All right, let's get let's get into shows. All right, so uh Chris, I thought this was really interesting. Our interview with the Riot CFO Mark Sodosante was picked up by the Chinese game media. So it was like game Euro or something like this. We're famous in China. What was interesting, I didn't realize that um even when I was there that uh Arcane the TV show for us is called Battle of Two Cities in China. So I was reading the article. I'm like, "What is Battle of Two Cities?" And I'm like, "Oh my god, it's Arcane." I had a lot of people reach out just saying it was really great interview and
Yeah, it was
And the Rioters thought that Mark really represented Riot values really well and like spoke well on behalf of them as a you know, it's a Riot's a cult. I say this with all the love having been in the cult. He's just like an exemplary person to talk about Riot culture. All right, Josh.

good interview.
What else do you got? You had another podcast on Deconstructor of Fun.
I did. It was such a pleasure to finally host Louis Della Camera on the pod. He's the long-time VP of global marketing at Rovio. He's now CMO of Smadex, one of the more exciting networks right now. But we talked about what a franchise you know, actually buys you, headaches included. And so all the nuances there was was really interesting.
I started it. I haven't finished it because I was so sad that he left. And then he said uh he's at your ad network people. He left my world.
He's come to the dark side. He is. But he's he's a really fun guy to to listen to. So I recommend everyone just go check it out. Like it's just a pleasure just to listen to.
This episode is sponsored by Moloco. More than 2 billion daily active users, over 2.8 million independent apps globally. That's the reach Moloco ads unlock for performance marketers who don't want to stop at the usual walled

garden. It's all powered by Cora. Moloco's compound AI system continuously learning from billions of events and getting [music] smarter with every campaign. Hop over the walled garden. [music] Check out moloco.com. That's m o l o c o dot com.
Unless you've been living under a rock, you know this is Web Store Romp Springer. 70% of top apps are using a Web Store and the best ones, they're doing it with App Charge. Do you want a $20,000 in-app purchase?
[music]
Yeah, App Charge does that. Do you want to pretend that subscriptions are still a viable monetization strategy for games? Yeah, App Charge even lets you do that. They get what high average revenue per paying customer means
[music]
and they're already doing over half a billion dollars in transactions a year.
[music]
So, are you going to be the first to do a $20,000 IAP? Don't accept microtransactions, embrace macrotransactions. Learn more at appcharge.com.

Satori. It's the Japanese Zen Buddhist term for sudden enlightenment. That flash where everything clicks. We're told that Heroic Labs team had [music] theirs when they realized that most studios run live ops like a horoscope. Vague, kind of generic, [music] same advice for everyone. So, they built Satori. Real-time experimentation, player-level personalization, events that actually respond to player behavior. Say sayonara to your old live ops stack and konnichiwa to Satori from Heroic Labs.
All right, let's get into our first topic, which is Roblox. So, Roblox earning time. Roblox beat its Q3 bookings guidance at 1.7 billion. Its DAUs fall 9% and shares drop up to 18%. Um this image we're looking at of the

stock from a couple days ago shows the the poor the poor drop that these guys keep getting hit with. First, I in this look, I wanted to see how Q3 was doing against Q2. The Q2 stock drop was 27% last week was only a mere 18%. So, what has to hurt is that the drop that we just saw basically they gained all the way back and then went back down. So, that that kind of sucks. Q3 bookings fell 12% to 1.7 billion above guidance. Helpful to know that the Q3 guidance had been 1. 6 to 1.65 around there set in July. Along with Roblox dropping its full year guidance. So, they beat a number that was revised down recently. And so Chris, I I know when you talk about things, that's one of the things I I wanted people to realize that they're like, "Oh, we beat our number." Well, you revised it down not too long ago. So, Roblox guided Q4 bookings down uh 10 to 15%. So, it's supposed they're not They're

not expecting things to get a lot better in Q4. How are people doing playing the game? Well, DAUs were down 9% year-over-year to 138 million and hours fell 15% to nearly 34 billion. So, a couple bonus stats. We always talk about age and obviously this is self-reported. Today approximately 20% of US under 18 gaming market runs through Roblox, which we kind of knew that, right? Like this is a kids platform. About 30% of US 13 to 17-year-olds use Roblox daily. A third Almost a third of all kids use it daily. Well, and that's even 13 to 17, not kids. And then US 18+ DAU grew 18% year-over-year. So, a lot to digest here. One of the areas I wanted to break down and just take a look at is this uh a a slowdown or is this a tough comparison? And I'm just you know, leaning towards slowdown.

They're having a comp Q3, which you know, Chris is going to show a chart in a minute that just absolutely shows this. And you know, the Grow a Garden explosion last year is incredibly difficult to comp. It's everything we've been talking about in the last couple of weeks that that has been the growth of the business and to be able to comp that or come back and do it again has been you know, almost impossible with not as many steal a games, you know, coming back and making that possible. The slowdown really is pointing to is the DAU numbers and the engagement numbers. So, DAUs were coming in at 138 and this was against a forecast of 145. When the players aren't playing, play time is down, that's what's starting to signal that this might not be the place that kids are still going to or anyone is still going to, although they've reported a lot of 18+ focus. I think I'll we'll get into that maybe in future episodes cuz we have so much to break down, but Chris, I wanted you to repeat your broken record and I I love

the graph that you brought.
Yeah, I don't want to be a broken record, but you know, they basically hit their pinnacle in 2025 with the steal from kids games, you know, it's basically all downhill from here with the new strategy that's been stated, even without the new strategy. So, despite the tweaking of the algorithm which actually made them beat the quarter, which was what I was expecting, but even then, stealing from kids games are just not as good as they used to be, right? And so, they can't grow or get to the levels that they had last year with the two steal from kids games. And so, I I I put together this chart really quickly in
[laughter]
It's like this is what the stock looks like when you have games that steal from children up at $138. This is what it looks like without games that you steal from children, right? And this is what we're talking about, right? So, it's like $35 versus $138, right? So, if they want to change the algorithm to move people away, which is what Dave

wants, and I don't know why people can't uh comprehend this is that he does not like these steal from kids games. He that's he does not want that on his platform, right? And so, even though he changed the algorithm in this quarter in order to hit the numbers, you know, they still are not in the same space. So, we This is where we're headed. And so, this is where we will likely see it. And as a publicly traded company, you cannot pivot like this, right? This is not the way the investors are going to be looking at your stock. So, that is basically the simple math for me, anyway.
Eric, you've told me like a dozen times now that ad tech is a dirty business. Um is it fair to say that maybe games are a dirty business, too, looking at this data set and maybe white knights can get punished?
No, yeah, of course.
Cuz like a lot of their friction is coming from like them trying to do the quote-unquote right thing, right? But those things are antithetical to revenue and to user retention.

They would have been fine had the Steal a Game not come in and cause this comp issue, right?
But But here's my point. Like before the Steal Games came out, right? That genre of games, which is insanely predatory from a monetization perspective. And people may disagree with me on that point, right? And I have heard get feedback on this all the time. But again, he does not want these games. So, regardless of whether or not you believe that they're not predatory, whether Philip believes it or not, Dave does not want this game. And I've had this come come confirmed multiple times by different people. Then this is what the stock is worth, right? Because the $35 is before
Mhm.
you know, the these games came out. And so, I don't know I don't know what to tell you. Like I I don't know understand why this is such a hard thing for people to understand. Sure, there're great games out there that monetize that have been around for a long time, but that's not a valuable as having games that monetize at the level that these Steal Games do, you know?
No, I hear you. The other element that I

wanted to focus on coming out of this announcement was how much they focused on AI. And one of the things that they mentioned in the AI conversation was the new the Steal an Egg game was made by I think they said completely by AI. There's another game in the top 20 called I think it's Whole Valley that is also all made by AI. Anyway, I just wanted to pull some numbers around AI because it's something that we're going to really need to pay attention to. Well, overall, we've been talking about it, but on Roblox in particular, it's really exploded on the platform in the last 4 to 6 weeks. So, we're not going to be able to see all of the numbers we want to see yet in this Q3 report. Well, we'll probably get a lot more in Q4. Roblox reports AI costs in dollars and AI benefits and activity because dollars will show up a little bit later. And so,
Whoa, whoa, whoa, whoa, whoa, whoa, whoa.
[laughter]
Oh my god, is that the biggest spin ever, dude? Like, okay, go ahead.
Yeah, yes, yes. So, but but I don't

think the num- the dollars are coming. And so, let's let's take a look cuz they're kind of asking investors to like, "Hey, trust us for the payoff." By the way, this is how AI is working, you know, it overall. So, here are the benefits and activity that they talked about in their shareholder letter and all of their, you know, I went through all of their different filings. 80% of the top 100 games use AI. MCP daily users grew nearly 5x. Asset uploads tripled. Games published with AI grew 50%. So, Roblox build is their prompt-to-game kind of alpha beta program. It has 500,000 users, 70,000 games, rolls out to the US in December. It's only in like five countries, most of them like smaller countries. It's still in the test phase. Two AI built games built reach the top 20, which are the ones I just talked about, Steel and Egg and and Whole Valley. So, Chris, this is actually on the cost side. This was like hidden and no one's really talking about this as much is can they afford all of this AI?

When you think about all of the different things that they're hit with, infrastructure trust and safety was $258 million, up 24%. Q4 infrastructure spend goes up again for build, Roblox reality, and universal player. Remember, they're doing a whole thing where you can take the game and maybe move it outside of Roblox, too. CAPEX is 380 million for that. Q4 guidance, adjusted EBITDA of 10 to 70 million negative, a year earlier to 3 million positive, and negative free cash flow, and I know that there's always the stock buyback there. 2027, the CFO says infrastructure spend will grow again. I think that is the hidden story is yes, AI is leading to all of these quote you know, engagement benefits, but where's the money going? Their costs are going out. We haven't seen a revenue uptick. They better have some really fancy ass like charts and stats on how AI games are driving bookings if

investors are going to show up and vote yes.
Bookings are down, costs are up. Like, this is not good, right? And on this promise of delivering games that are retentive, but but they don't monetize necessarily, right? Particularly as well as these steel games. And so, they're in a really tough spot, right? And Wall Street hates this you know? Like, they hate this, right? And there's other places to go put money. And the big thing that I I want to be clear on is that without a good stock price and without a good stock-based comp, which is a huge part of how they compensate their employees, they're not going to be competitive out there. And these engineers that work at Roblox are going to go elsewhere. They faced this issue years ago when the stock got destroyed, too, you know? And so, it's a real challenge for them on where they are. And that's why going private, that you know, that's why that article makes sense, you know, from Convoy, is that going private may make sense in this case. Because they have flexibility to do more of a pivot and start to grow from here.
Yeah, moving on. I just thought that was

an I just wanted to do that kind of take because, you know, obviously everyone can read the 50 million takes by the time we do our podcast, so I wanted to pull that aspect out and we'll stay on the AI train because we've been meaning to get to this topic for a couple of weeks. I'm glad we have our puzzle and ad expert here. So, Laura, I want you to take us through the story of that There's a creator who's decompiling all the ad mon games, getting into the weeds on how linked the gameplay is to when the ads pop up. So, can you help me help me understand? Help me understand.
Yeah, we're still writing the AI trend. So, I think it's been a growing topic of concern that now we know with all this AI tooling you can use it to take games apart, which you could always do. It's just now you can do it faster and much more easily. And it's impacting the game community in ways where even though we we knew about that, the clones and fast follows and people trying to reverse engineer things and it's coming to a little bit of a head. So, this guy AO Ships, he's an influencer and allegedly

a data scientist. I did not go and vet this. He co-founded this company ao.dev, which is an AI app builder, which is the opposite of what he's doing. So, what he's doing is he's decompiling and deconstructing mobile games to reveal their secrets. Again, we've had ports, people have done this in the past. I cited like one example. I think Epic has filed notices about bytecode ports of Fortnite. Like this is not new. And I remember in my when I was still working on mobile games in my previous life, we worried about this, [clears throat] but it didn't necessarily impact our approach. Whether it be if we were deciding what feature to build or our go-to-market plans. If a feature was good for the game, even if it could be copied or just easily taken apart, we'd still release it. Like it wouldn't change what our plan was. Now, I've been seeing that games and game companies are scared to release demos because why risk it? Why risk someone deconstructing the demo and potentially rebuilding it faster than you can release it.

So, what did I do? I went through I was like, "Okay, what is this guy AO? What's he doing?" And I I stuck with the Block Blast article. That's the one I kind of focused on. And it was hard for me to understand the why. He's going through and showcasing his interpretation of code. And my read, and this has come up on the pod and we've Eric and I have debated this for years, is that he's suggesting it's malicious. Using it as a way to expose developers. And I felt that the article was a little one-sided that he's not considering the overall outcome of when these things are designed, why they're designed in such way, and what they want to evoke for players cuz it's not all It's not always Sometimes [laughter] it's good. Sometimes it's good. And the approach was interesting, but then I was like, all right, let's pretend he's doing this to my game. And there's a couple things I wanted to I was like pit I'm trying to poke holes in that. Unless it's commented in, like you don't know why something's built. So, it could be a feature or some sort of line of code. Could be The goal could be to lift your attention or protect players from frustration.

Or how it's tuned server-side. And that last point, he does acknowledge those models are invisible, so he's only interpreting half the system. And this is kind of where his whole malicious dark framing tends to fall apart. I go back to and uh hopping on my soapbox, I do still feel like it's design expertise plus looking at market data is still king when it comes to approach regardless of what of putting all of this together. So, the TLDR, he decompiled the Android iOS versions of Block Blast, he ran them in emulators trying to understand the game traits of the game rules, and then what the game was trying to draw for players, like what player patterns were learned. And I can't speak for Block Blast specifically cuz I haven't worked on that game, but some things to keep in mind is when I did work on games, they have such a long evolution that you might have a game trait you worked on initially and you called it let's say like show red pieces. Despite that, you might change that going forward based on what you learn so that it showing red pieces might turn to

green and then you may not rename it. He said he tested the logic as well, but I wanted to call at this point that if you decide to go down this path, just please keep that in mind because you need to do the testing aspect he did do to make sure that things were matching appropriately. Now, the dark pattern he's talking about via ads, this is specifically what he's digging into is called lethal hands. So, it's basically logic that determines when a when a lethal hand appears. And a lethal hand is basically something when you play the game, it reminded me a little bit of like weird freestyle Tetris. Um, you have three pieces that appear on the bottom and the three pieces need to be placed on a board with sometimes half placed pieces and then you're you're creating lines and rows. Sometimes you get three pieces that can't be placed. So basically, with the logic that in his version, because different players get different versions, if you had three rewarded ads, the ad that's now queued for you is worth 3.5 times the last one and your score is a thousand points higher than your last high score. It basically requests a hand from a

generator that says the three pieces you have cannot be placed and you'll fail the level. That one he called out specifically and he thankfully [clears throat] he does also say they're player friendly rules. So if you have you're losing a lot, then you can basically the game becomes easier. And we This actually Neil Long talks about in this article. This feels like dynamic difficulty. But what is really he's really pinpointing is the showing of ads of which Block Blast is earning substantial revenue. So if I play the game for a bit, it feels like an endless level. If I was going to be put my design hat on and I was going to approach this, there is an ideal time to lose and that should be calculated by a combination of whoever's designing it, their sense of fun, what the pacing should be and behavioral data. Games do need to generate revenue to run and combining them with an ad at a natural break where you feel like this would be a good time for the player to reset seems logical to me and is something that I would consider designing to. One endless level, it doesn't sound necessarily fun and I'm not sure that's the way games are expected to be played today.

I will give some credit to AO here. I think where what he's trying to show is that picking an unwinnable hand because of priceier ad loaded is harder to justify publicly and I understand from a developer's point of view, but like if you're if this article and his videos are on TikTok, so it's going to the masses, there's a lot that a non-gaming development audience may not understand. So Josh,
[laughter]
I feel like I'm treading in your territory a bit.
Yeah, I know. I I I I have big feelings here. I'll I'll try to contain it, but I do mostly agree with Laura. I think it's very important to manage when people struggle and when and how they win. You know, difficulty management isn't necessarily an evil monetization trick. This technique also boosts retention and engagements. Like a classic example, nobody wants to have a blowout loss in a match three game. We do generally want to make sure that if you lose, as often as you can, it should be a close loss so that it feels more fun. And so designers

taking the time to make sure that their levels adjust so that you you don't get your ass beat basically. It just feels better and that makes a better game. You know, that said, just because we can't assume their intent, doesn't mean it's not sinister. Some of these games are very very aggressive and the company culture is can be, you know, incredibly aggressive. I guess one disclaimer, I'm personally implicated here. I run an ad monetization studio, but I also designed a large part of our difficulty management system in our games. But the team is growing now, so I'm watching product and game design slowly dismantle all my stuff piece by piece.
[laughter]
Um yeah,
So you're stuck at it, I guess.
All right.
Hey, hey, hey, I I can't beat an army. I can't beat one man can't beat an army. So difficulty and monetization are deeply intertwined. The challenge here that Block Blast is trying to manage is that every ad,

you know, generates revenue, but it also increases churn. And so the more likely an ad is to make someone leave, the more you need to get paid for that ad to be willing to show it. And that's a super easy thing to manage with interstitials. Like a forced ad that shows on the screen as a full video, because it's just is it worth it? Will they leave or not? Rewarded video where players get to pick is much more complicated. And so what you see here with Block Blast is they found a moment where there are players that like to watch rewarded videos in a moment where it turns out, I'm guessing if you just hit a new high score and then you die, you want to run up that new high score. And so they found a placement where people want to watch these ads are more motivated and the CPM is high, the value of the ad. So now is the perfect time to just end the game, force the rewarded video and let it go. I can sympathize with the players

here cuz what I just said feels very much like saying Santa isn't real or your favorite reality show is scripted. Like it's sort of disappointing that the magic is gone, but it's too profitable not to. If you don't do that, your game doesn't get to exist at this point. Like you can't have enough LTV to run UA.
Welcome to the game.
the games that rise to the top are the games that do this. And so it's just one of the realities like Santa is not real in gaming.
Well, I think the other reality is I think we figured out where you need to go to school. You don't need to go to college, you need to just pull up these Tik Tok videos and get, you know, go to AdMon product management via cool decompiling videos. Is this the new required learning? Is this guy's channel is like here's how you do it. But I did look at the um just the number of likes. It was Tik Tok, right? So 1.5 million people uh liked the Block Blast one. All

the other ones he's got, I think he decompiles, I don't know, maybe 10 other games and they were only about 50,000 likes. So I think all of our game industry press actually pushed people to watch that one specific ad. It was pretty interesting and so he should be thanking the game industry for going and watching and taking notes on on what he's doing, but I feel like you guys cover it well. I think there's pros and cons on both sides to this. I think, you know, is the mobile game industry, is the game industry full of dark arts? Is it full of white hats to your kind of white knights, you know, thing earlier? Yes and yes. And by the way, this doesn't like mean that console and PC games are off the hook either. Like they have design and play patterns and this is all considered. Chris, any any any thoughts?
I got nothing on this stuff.
[laughter]
I thought you'd be like yeah, go go make money, of course. Like now we just can see it.
Yeah, I don't know. This stuff makes me throw up in my mouth

a little bit, but
But that's predatory marketing.
Yeah, you're totally right though. Like my brother till recently he worked at Sucker Punch. And these Sony studios when they build their trailers, they're building totally different versions of the game that are unrealistic just to maximize how like cinematically interesting it is. And so is that a fake ad? I don't know. These are really hard trade-offs to really evaluate.
Well, that's a whole other topic. We'll get into actually later I would cuz they were there was actually some fake ad news in the UK this morning where I think a a number of of game companies were asked to take down some other fake ads in mobile That's a whole other topic which I also want to throw up in my mouth sometimes about that. But um let's talk a little bit about what's going to make decompiling harder. And so on Monday of this week Anthropic said that it added two sentences to Claude code steering Opus 5.5 away from recreating existing video games. Original and genre

games can still work when you're creating them from scratch. If you're recreating a specific existing game, it's blocked even if it's renamed. Anthropic's Adi Osmany confirmed it on October 30th, 2026 citing the ban on using Claude to infringe IP. And so I think to what we've been saying the last few weeks, I think the lawyers are finally calling up people saying, all right, you're dealing with IP now. We got to we got to we got to cut this out. Developers report refusals on their own projects and on crash dump analysis. So it might be hard to kind of police this, but Josh, tell me a little bit more you wanted to go into this one.
Yeah, I mean, this is already happening at scale. People are using AI to do tons of things that are potentially in a gray area, like deconstructing games, cloning mechanics, recreating playable ads. You can even do things like watch revenue

movement in Sensor Tower based on app updates and be decompiling every app update and seeing what code may have caused revenue deltas, like very, very granularly. You know, I don't have an industry-wide number, but with the dominance of AI, this is not a hypothetical. And so, can Anthropic do anything about this? No. No, they can't do Why? Because people are are using Chinese AI to do this. And those things don't have morals. They don't even They don't even pay for the server cost. You can run that yourself. This practice is way too profitable. I think the West has been quite slow to adopt it, and so we're seeing this now way more fully formed. And this is a very real constraint that we have to work with. You know, and this like actually this connects directly back to Block Blast. You know, after this article went out, I heard from people outside of the industry or outside of that specific

niche of the industry disgusted, like this is disgusting, like this is a dark pattern, this is this is gross. Inside the more retention-focused casual ad-monetized industry, I heard oh, We need to do this right now to catch up, because otherwise we're they they have an advantage that we don't have. The same article that made players feel manipulated gave developers, you know, something to test. This applies to playable ads, too. You know, everyone's stealing everyone else's homework. Like whether you're talking about UA or product, we've always been copying competitor ads, competitor live ops features. This is now just faster and more automated. The takeaway is that the amount of time that you enjoy a benefit from innovation is collapsing down and the cost-benefit of being the fast follower keeps getting better and better because the the cost side the denominator just is going down as

well.
The challenge of this whole thing though is that regardless of whether or not you're copying someone else if you don't have the resources for user acquisition and ROAS campaigns you know, you're Anyway, right? So you can copy all you want.
Right? So it's like
huge joke.
The only people that are really at risk right now are are companies like Voodoo and you know, this hybrid hybrid casual nonsense because people can replicate those really quickly and and shovel them out there, you know?
The more you rely on new users and user acquisition the more at risk it is the more you have one huge user base that has been around for a while and is being retained the less you have to compete in that literal sense. And so if you're a business that has a a massive live game or games and you're mostly just running re-engagement you are more safe.
Yeah, existing games with existing audiences that are figuring out how to keep them you know, those make sense. IP always make sense.

Yeah, yeah, yeah.
Uh but
The the traditional the traditional kind of things that create motes um those are still those are still work.
But we've been wondering who the gatekeepers are going to be. Claude raised its hand on this one and you're right models are commodities. You can if Opus doesn't do it for you, all right, go to open AI, go to the Chinese models, who who cares? It it doesn't it doesn't really matter on that point but somebody else is going to have do the platform step up. We know that Steam isn't stepping up to really do anything to manage the copycats that are coming in. Apple is standing up right now. That's why review times are you know, three or five x longer because Apple does have you know, rules against copycat, but those rules are so kind of light. What what does that actually mean? The who is trying to police this is becoming a little bit problematic and and players I didn't you read that there's a itch.io article that they said they had, I don't

know, like 70,000, you know, AI slop games flood in, but no revenue growth from all of these AI games coming in. And so who is suffering? You know, it's kind of the consumer, the player who is like, "Ugh, I I just want to play a good game. I don't want to play this shit." But um there are good things coming. I mean, today I was reading about like, you know, double-A or single-A games that are popping up on Steam that are becoming hits from developers or creators who would never have been able to do that before. So again, like you know, we we just need a little bit of guardrails around where this stuff is going and you know, what is we're in the wild west of all of this right now.
Eiffel Tower, iconic. [music] Tower of London, historic. The two towers, simply epic. But none of them tell you anything about mobile app intelligence. Downloads, revenue, active users, and my [music] personal favorite, weekly cohort and retention. Get the real story, get the real data, get

Sensor Tower today.
In sickness and in health. In soft launch and in live ops. Through every meta shift and every platform policy change. That's the kind of loyalty vibes built between your game and your players. AI power targeting, [music] a real meta layer, ROAS optimization that actually compounds. [music] That's what makes vibes the fastest growing loyalty platform in mobile games. Till churn do us part. Vibes. v y b s .co
In today's hyper-competitive digital landscape, understanding your app's journey from discovery to download isn't just insightful, [music] it's the difference between scaling and stagnating. Enter AppsFlyer, the undisputed leader in mobile attribution and marketing analytics. They've mastered [music] the science of making every ad dollar work harder by revealing exactly where your players come from and how they interact with your app. Their

platform transform mountains of complex data into clear, actionable intelligence that separates winning campaigns from budget-draining experiments. And I know that many of the top game companies [music] leverage AppsFlyer and are a key partner in their success. AppsFlyer, where your app's potential meets performance. All right, let's get to a couple of Microsoft story. We just need Microsoft corner these days, Xbox corner. Laura, plug your ears for the rest of this.
Yeah, I mean, we I don't want to keep covering this over and over again, but the news keeps coming through, so it's we'll be as quick as possible. For So, the first thing was that the Forza co-creator Dan Greenawalt leaving Microsoft. So, he was at Turn 10 Studios most recently, but he was at 27 years at Microsoft. Um Now, this seems like a voluntary exit, but, you know, likely a bit of a push-out. Um But, to be clear, his studio really no

longer exists. He became a layer management after Michael Bastian was appointed COO. And his franchise Forza Motorsport has basically been shelved in favor of Horizon. So, in some senses, they're just consolidating on one Forza game, which is is pretty smart, really, but it just kind of pushes someone that's been doing this for a long time out and not as part of it. So, that that's one thing. But, the bigger news this week was this Helen Chang thing. And she wrote an article basically clarifying that she she basically denies any wrong doing related to her departure. And I quote, "There's been some speculation about my abrupt departure from Microsoft, including claims that I was fired for a policy violation. That's not accurate, and while I won't turn this into a play-by-play, silence should not be mistaken for agreement. What has been said about my departure and what has been left unsaid has created a picture of me that I hardly recognize. It's important me to set the record straight. So, she went on saying,

"First, I did not leak confident information to the press. Second, I did not leave Xbox to join a competitor. Third, I do not engage in discrimination, bullying, harassment, insider trading, embezzlement, or other fraudulent activities. And lastly, I did not leave because I was passed over for a bigger role." So, unfortunately, she really didn't elaborate about why she left, but it implies that she may have actually violated some type of policy, but it was applied unfairly in her firing. That's kind of my interpretation. Like, it feels like Sharma basically used this excuse to fire her and did it in kind of a vicious way from her reaction. For Sharma to offer only like a one-line summary after 21 years of service, and then afterwards leaking, likely leaking, likely leaking the policy infraction to The Verge, right? Because they picked this thing up from some leak within the company, which is very sus, in my opinion.
And was not independently verified by whoever

the original leak was not verified. And so, that's why only one outlet covered it.
Yeah, and that's
I just I wanted to look into this cuz it gets really like allegedly allegedly on this.
Yeah, and so, it's like you have to kind of read between the lines here, but like it it just felt like Sharma was being vicious to get rid of this woman who had been there for 21 years. That's what it felt like. And that she was getting a bit sidelined from her role, like we talked about last week. Again, what I had heard was that she was just not happy with the direction of the company and how what Sharma was doing. So, I don't think this is a good look for Sharma, regardless of what likely is coming. There's no indication that there's any type of lawsuit happening, but she's probably contemplating that right now. I don't want to beat a dead horse here, but losing studio leadership, like these people that were they're losing, is basically the sign of the apocalypse for a game studio, right? Like you may disagree with some of the direction that they've had over the years, but people managing teams that keep the ship moving are worth their

weight in gold. Like it's really hard to find people like that cuz the value of the studios is not the IP or the engineers or even the artists, right? But it's the people that act design and put these things together and manage creatives that is like the hardest hardest thing to do and those are the people that get the games out the door, right? And when you have less capability in this department, you get less results. And so Sharma, Ball, and all these tech bros, they don't understand this fact at all, I think, anyway, or it seems to be because I think they're basically ultimately sending Xbox further into the abyss by just like losing this type of leadership. That's kind of my interpretation of this is that, you know, they're getting rid of people they should not get rid of and they're doing it in ways that are that are in this case vicious and nasty.
Yeah, it feel it feels a little like the move fast and break things old Silicon Valley ethos is is happening here with org structure, with the org. And and I

agree with you having been, you know, a worker bee inside of so many of these bigger companies, when you start making so much shakeup with established culture and leaders and ways of doing things and institutional knowledge, I I think it's really difficult to rebuild. The other thing I would say here, too, is I always say with layoffs, like please just rip the band-aid off. Make all the changes you want to make and then like let's go and then let's start. This like trickling of like things keeping happening, you you cannot as a culture have people reduce their stress level and and then go cuz they're worried about like, okay, well, what's happening in 3 months? 3 months from now? So, I feel for all of the teams who probably just don't know like kind of like what's going on.
To be clear, I want everyone to understand it is absolute chaos at Microsoft. Like regardless of all the PR that you see out there, people don't know which way is up at all, right? And

in that kind of environment, you can't execute. It's impossible, right? And so like this needs to like get back to like steady state and they need to keep they need to keep pushing forward or spin out and let new leadership take over or something, you know, I don't know. It's it's a mess.
When I was uh uh at one of my past jobs, um I had this employee on my team and I called him um the the office politics emotional vampire because all he wanted to do was gossip about the like office politics and like what happens if this leader moved and he was like all interested in like the chess of the the office politics. And I'm like, "Do you work, dude? Like I don't want to sit here. I just wasted an hour gossiping with you about nothing we can do anything about. Can we just get back to work?" I can only imagine this is just a fraction of how people are feeling. And then they also have the 60th Remember they have the 60,000 person voluntary separation package

that's going on as well. And so how long does that take before those people raise their hand? And by the way, the people that do that are the veterans that know stuff, right? Like you
Yeah.
You're like, "Hey, I've been here forever. Stocks up. I'm going to go. Peace out." Like what a great time to do that. All right, let's get to some good news. We have a few minutes left. And so Elder Scrolls 6 is now officially Elder Scrolls 6 Sentinel. Uh that was revealed for Bethesda's 40th anniversary. It's set in Sentinel, a city in Hammerfell, southwest of Skyrim. So if you're a Skyrim fan, you are just thrilled to hear this news. What you're not thrilled to hear is it's still years away and a PlayStation version is uncertain. Is this one of the exclusives? They did not say it. They didn't say it in the press release. I was scoop looking everywhere to be like, "Is this one of the ones?" Cuz she said, "Hey, we want to do exclusives." So, are you a fan? I'm not I'm not a Skyrim player, I hate to say.

god, dude, that's dude, that was like digital crack, dude. That game was [laughter] amazing. Like they actually brought the other games I I could be do with or without, but that game was just insane. Yeah, I'm totally looking forward to it. But this feels again like the the Blizzard stuff where they are basically announcing things early. That doesn't mean they can get it out on time. Like given the chaos that's happening at these studios. But yeah, Elder Scrolls are beloved. It definitely will be cross-platform. Xbox won't even exist by then as a platform in my opinion. Maybe on PC, but I don't think the console will exist. So, I think they're just trying to justify their existence and and throw something on the gauntlet cuz they're being required to by by Sharma. Like announcing it is going to mean that it's going to come out type thing.
[laughter]
Well, it's just here's the reason, it's the 40th anniversary. They're like, we have to say something. What are you going to say? And they're like, we have a logo and a name.
Dude, they announced this this they announced this the game this existence of this game eight years ago.

Like yeah, and so like we're in year year eight and now we're like three or four years away, right? So, you know, I mean
No, I've been in this meeting. I've been in this this meeting is all right, 40th anniversary, what do you guys got? And then the game team looks at you and they're like
Yeah, it
Uh and you're like, all right, marketing, solve this for us. And so they get a logo and a name. And then there you go. Check. Box checked.
know. I'm looking forward to it, but I'm not expecting it anytime soon.
Yeah.
Anyway, my power is about ready to go out. So, before we get blow up this whole podcast, let's uh
Okay, well well one final thing was Ubisoft is reportedly um among the suitors uh preparing bids for the purchase of French developer Arkane Lyon, makers of Dishonored series and the upcoming Marvel Blade from
Why why would you want any more operations in France? Are they out of their goddamn minds? Okay? I don't know anything about Arkane Lyon, but all I know is that France is the last place that I would put bodies, you know? That, dude. Why are they doing this

stuff? You are just adding more more expense in the worst possible place on the earth to hire people. I mean, what is happening?
audience You're welcome.
It's culture, Eric.
It's culture.
Just give me soft, Eric. [laughter] All right. That was my That was my episode ending grenade that I threw at you. Thank you for responding to that. All right, everybody. Thank you so much. Laura, Josh, lovely to see you. See you again soon, and see you guys next week.
You did it. You made it to the [music] end of the episode. As a fan of the show, it would help us out if you'd subscribe and leave us a review on the podcast service of your [music] choice. Those stars and likes, while a small action on your part, mean a lot to us and help to grow the show. More importantly, are [music] you a member of the Deconstructor of Fun Slack Group? If you have 5 years or more of games industry experience,
[music]
go to deconstructoroffun.com/slack and apply to join.

Join the games industry best professional community filled with peers always willing to lend a hand, or subscribe to our newsletter to get all the latest insights from the Deconstructor of Fun content creators. Thanks for listening. [music]