Odd Lots: A Billion-Dollar Game Is Affordable Only When You Already Know It Pays Back

Notes on the Odd Lots episode of 18 November 2026, where Tracy Alloway and Joe Weisenthal talk with Bloomberg games reporter Jason Schreier about how GTA VI became the most expensive video game ever made. The rumoured budget Alloway cites is $1bn to $1.5bn, never confirmed by Rockstar; Schreier's benchmark is GTA V, which has sold 235 million copies since 2013, took $800m on day one and $1bn in three days, while roughly 80% of Take-Two's revenue now comes from what players spend inside the games. He reads this one as an anomaly while the rest of the industry cuts budgets, a reading that holds while the installed base stays affordable — and consoles and graphics cards are moving the other way. Educational listening notes only — no investment advice, no price targets, and the judgements here may be wrong.
Contents

The fingers striking one after another do less than the closed fist landing once; ten thousand men advancing in turn do less than a hundred arriving together.
—— Huainanzi, “On Military Strategy” (Western Han; my own translation)
In the Odd Lots episode of 18 November 2026, Tracy Alloway and Joe Weisenthal talk with Jason Schreier, who covers video games for Bloomberg, about how GTA VI became the most expensive game ever made. The rumoured budget Alloway cites is $1bn to $1.5bn, a figure the developer Rockstar has never confirmed; Schreier’s benchmark is the previous game, which has sold 235 million copies since 2013, took $800m on its first day and $1bn in three days, while its parent Take-Two now earns roughly 80% of its revenue from what players spend inside the games. His conclusion is that this one is an anomaly — Rockstar is the only company that can spend like this, because it knows the money comes back. For that to hold, the machines people play on have to stay buyable, and console and graphics-card prices are moving the other way.
The strongest version of “only Rockstar could do this”
Schreier opens with a pop quiz: how many copies did GTA V sell? Weisenthal guesses 250 million. The answer is 235 million, and Schreier says most people he asks say 40 or 50 million and are floored by the real figure. That volume makes GTA V the best-selling game ever on anything other than phones, and it took $800m on launch day, $1bn within three days. His expectation is that GTA VI could double the day-one number, because the fervour this time is at a different level from 2013.
The evidence for that fervour gets absurd. PlayStation and Xbox both say GTA VI broke their digital storefront pre-order records, with outside estimates above 10 million copies. When Rockstar premiered its half-hour gameplay reveal on Netflix over the summer, Netflix said it was the single most-watched thing on the whole service that week, beating a brand-new Robert De Niro film.
So the billion-dollar logic works. Schreier puts it plainly: you can only spend more than a billion dollars on a game if you know you will make it back, and in this case they knew. He asked Take-Two CEO Strauss Zelnick whether Rockstar had a blank cheque, and Zelnick answered with his son’s wedding — I told him he could have an unlimited budget, and he surpassed it. The joke exposes the mechanism. Rockstar can spend this way because the Grand Theft Auto name is itself the payback guarantee, which leaves the budget number free to be whatever it turns out to be.
Why the money runs this large comes through in an analogy Schreier offers. If he were making a film about a guy with a guitar, he says, he would point a camera at the guitar behind him, and if he wanted a cinematic angle he would swing around the room for a few looks — a couple of hours of work. In a game he has to build the guitar as a model, then everything behind it, and if he wants to spin around the room he has to model the entire room, which could take weeks. Scale that to a whole city, add a player who controls the camera and can look wherever they like, and that, he says, is what makes this stuff take eight years.
He has been playing a review copy, and a few things he described stuck with me. Driving through the city when the light hits right is worth stopping for. The two leads, Jason and Lucia, are a Bonnie-and-Clyde couple; you control one and the other walks round to the passenger door, lets themselves in and starts chatting, or hops on the exercise bike beside you while you squat in the gym. The sharpest image was this: early on he pulled into a gas station and found a man filling a bathtub in the back of his truck with gasoline while the attendant screamed at him. He could intervene or stand and watch, so he watched for a while until the man escalated into something unhinged, then stepped in and did the attendant a favour. These stranger encounters were a hallmark of Red Dead Redemption 2.
Set that against the rest of the industry and the anomaly label gets sharper. Call of Duty — annual, normally the biggest release of any year — is struggling, partly, Schreier says, because players are saving their money for GTA. Most of the industry is hunting for ways to cut budgets, because $500m to $600m productions no longer work: sales have not risen alongside them. His phrase for that position is untenable.
Standing on the other side: five places the case gives way
Alloway asks the question that matters. Rockstar’s argument is that this spending is an investment in technology that will make future games faster — should we believe it? Schreier does not line up behind Rockstar. An expansion within a year or two he finds easy to imagine, because the world already exists and adding to it is tractable. Saying the next Red Dead will not take another seven or eight years, he cannot imagine. The one acceleration he will name is that the team is assembled and the pipelines, workflows and trust are built, and even that carries an assumption: that Rockstar will not double in size again. What the money bought is the right to keep building on one map. The schedule for the next new game was not part of the purchase.
The headcount speaks for itself. Schreier says Rockstar had about 2,400 employees when it shipped Red Dead Redemption 2 in 2018 and has 4,700 now. Games in the 2000s were made by teams under a hundred people, in two or three years, on budgets around $20m. Doubling your staff and spending eight years to produce one game is how Rockstar works, while its peers are moving the opposite way.
Part of the thirteen-year gap has nothing to do with technology. Schreier says you cannot discuss GTA V without GTA Online, the multiplayer side that launched a few weeks after the game, started out clunky and slow with everyone wondering why anyone would bother, and then grew and grew until Rockstar had to keep a dedicated team feeding it. On microtransactions alone — outfits, guns, assorted in-game things — it has been earning an estimated half a billion dollars a year. With a tail like that, the incentive to hurry a sequel thins out. So some of “most expensive ever” is a by-product of “longest delayed ever”, and the delay happened because the old game kept paying.
The schedule history also reads poorly as a repeatable template. GTA VI was formally announced at the end of 2023 with a trailer that said “coming 2025”; Schreier says that was a flag planted two years out, and even senior people at Rockstar knew in the back of their heads that 2025 was never going to be a date anyone kept. It slipped to May 2026, then to November. He describes game development as the hard parts of software and the hard parts of art stewed together: spreadsheet software can be broken into goals with time estimates, while “is this fun, does it feel good, is it paced well, is the story satisfying, does it land with playtesters” are squishier questions that will not go on a schedule. He wrote a book about Blizzard, Play Nice, and says one of the biggest issues Blizzard ran into over the years was its artistic sensibility butting heads with the appetite for predictability held by Bobby Kotick and by the market — games and public markets often do not see eye to eye, because games are unpredictable and markets demand predictability.
Finally, the inference that a billion dollars is what a hit now costs breaks inside the same conversation. Asked whether a brand-new franchise at scale is still possible, Schreier points to Clair Obscur: Expedition 33, made last year by a team of about thirty people in France, now near 10 million copies sold. Before that, Elden Ring was a new franchise, sold something like 30 million copies, and is becoming a film directed by Alex Garland. New franchises do emerge. The hard part is getting a company to put hundreds of millions behind one, because that usually works out badly.
Which side holds, and the condition it turns on
“Can I trust anything about something this hot?” — that was my first reaction to the pre-order records and the Netflix number. Schreier draws the line for me inside the episode. A pile of market research firms, he says, just want the headline “billions of dollars in lost productivity due to GTA 6”, and he does not buy that, calling it nonsense prognosticating. What he will report is that his high-school group chat, mostly non-gamers, says they are going to play, and that Bloomberg has stories planned about people taking a week of leave for it. One is a fabricated aggregate, the other is something he can go and ask. With a hot story I try to ask first who measured the number and whether they could.
“Does selling well mean earning well?” — this episode gives two separate lines. Roughly 80% of Take-Two’s revenue comes from what players spend inside the games. First-week units are the easiest figure to get and the first to be priced in, so following this business means following how in-game spending moves; whether the opening-week record falls decides one week.
“So what should I be watching?” — the third point is the one I care about most, because this industry’s bottleneck sits outside this industry. Schreier notes the game is over 210 GB and digital only, too large for a disc, and that PlayStation has announced it is phasing discs out starting in 2028. The sharper end is hardware. High-end graphics cards topped out around $700 or $800 and now cost double or triple that, often unavailable, sometimes thousands on the used market. The PS5 launched at $500 in 2020 and the baseline model is now $650 — the first time all three console makers have raised prices — and the fear inside the industry is that the next generation breaks $1,000. He puts the cause down to shortages and price gouging from the AI chip craze, and says nobody really knows when memory and GPUs come back to normal. A publisher holds no cards in that hand.
So my read is that “only Rockstar can spend this way, because the payback is certain” is true today, and that certainty comes from the name multiplied by a cheap installed base. Those 235 million copies accumulated while consoles cost $400 to $500. If the next generation lands above $1,000, the name survives, the reachable audience shrinks, and the arithmetic needs redoing — and the first failures would surface at the companies putting hundreds of millions behind games with no name to lean on, well before they surface at Rockstar.
How would I know I am wrong? Two markers. Whether an expansion ships within two years — Schreier says that one is easy to imagine, so if even that does not arrive, the “technology investment speeds up what follows” claim goes in the bin. And the price of the next console generation, which sets the multiplier above.
Further reading
- The episode itself: Odd Lots, 18 November 2026, Tracy Alloway and Joe Weisenthal in conversation with Jason Schreier.
- Schreier’s Bloomberg Businessweek international cover story on the making of GTA VI, which he mentions visiting Rockstar for.
- His review of GTA VI, which he says should be up on Bloomberg around the time this episode airs.
- Play Nice, his book on Blizzard, about creative instinct colliding with the predictability public markets want.
The one thing worth taking with you
What makes a large sum affordable is the certainty of getting it back. The size of the pile you spend from decides far less than it looks like it should — Zelnick’s son had no budget ceiling, overshot it anyway, and the wedding still happened. The billion dollars behind GTA VI works the same way: the budget figure stops mattering once the payback end has been settled.
Here is something I have tried, and you could do it today. Take the thing you have put the most time into this year — a course, a relationship, a side project all count — and write two sentences on paper. The first: what made me believe this would pay back. The second: if that reason fails, in which month and from which specific event will I notice. The first sentence tends to come easily. The second one took me a long time to write the first time.
This article is an educational discussion of investment method. It is not advice to buy or sell any individual security, offers no target prices, and does not analyze any current holding. Investing carries risk; make your own decisions or consult a qualified professional.
Comments
Loading comments…