How Many Units Were Sold?: Playing the Numbers Game with Video Games

The video game industry is unusual in that it has had no official channel for reporting sales numbers throughout it’s history. The other two major entertainment mediums – music and movies – have had consistent relationships with groups who report on sales numbers and box office revenues respectively. These may not always be directly available to the public, but they can matriculate into the world.

Image result for ps4 switch xbox one sales

Video games have never had such a thing and continue to fight against adopting the practice. Microsoft for instance has gone out of it’s way to not report official figures of sales for the Xbox series, leaving analysts to speculate on whether it is truly competing with the PlayStation consoles. Nintendo reports internal figures – which appear to be accurate as of the time of their showing – but it’s not done with official cooperation with an analyst group. Their highest sales figures appear on their website and others are shown in investor reports.

Almost all sales figures that you will here about games  – especially the old ones we talk about here – are what historians would call “anecdotal evidence”. This means that although they may have been said in a somewhat official capacity like an interview or a sales group meeting, the actual source they were drawing from could easily have been distorted through a number of factors. I’ll try to cover as many areas of fact, fiction, and somewhere in between in this post.


The first attempt at a comprehensive survey of video game sales numbers was undertaken by none other than Ralph Baer. Of course he wasn’t really trying to enhance the credibility of the industry or anything: He wanted to know how much money companies could potentially owe if they had infringed on the Sander’s Associates patents from the Odyssey. He went to the 1975 MOA show, surveyed all the manufacturers and games he could find, and came up with four pages of figures hand-written for potential use in the litigation. These are reproduced in the book Videogames: In the Beginning.

So where did Ralph Baer find these numbers? According to Baer, he got them from reading the pages of Play Meter magazine. This is unequivocally not true as there was no comprehensive survey chart used by Play Meter. It is possible that maybe these figures came from Play Meter directly as a publication, but it begs the question as to why they would give that information to somebody like Baer on a whim and not print such valuable information in the magazine.

Many of the figures listed by Baer are hard to qualify as many of the companies listed are obscure even when you dig into the coin-op trades. Some figures that stick out as inaccurate include Meadow Games’ Flim Flam at 500 units (Replay said it sold 12,000) and Wheels from Midway at 7,000 units (in 1986 Bally reported selling 2,400). However, other figures bare out, for example TV Pin Game is reported as 500 units in both Baer’s chart and the settlement agreement between Chicago Coin/Stern Electronics and Magnavox. Other figures from Chicago Coin are off, reading between the lines one can see that Baer – if he did had insider knowledge – tended to count several machines together as a single unit.

One area where we know Baer had insider knowledge was on that of the Odyssey, but we can’t take Baer uncritically when he reports figures. In this case it’s not so much the amount as it is the timeframe that he states. For the four years of the original Odyssey (1972-1975) Baer reports manufacturing figures of 140,000 in 1972, 23,000 in 1973, 170,000 in 1974, and a total to 1975 of 350,000 (which would be 17,000 extra units). The numbers themselves tell an interesting story, but the sales figures reported by Magnavox in both the publication Weekly Television Digest with Consumer Electronics and testimony by product planner Robert Fritsche give figures of 69,000 in 1972, 89,000 in 1973, and 129,000 in 1974 (total of 287,0000 with the last year unaccounted for).


On one level, Baer is reporting manufacturing numbers versus sell through so some discrepancy is to be expected. However he does also claim that 150,000 were sold in 1974 which contradicts internal figures. Presumably he’s basing these claims – which in total seem relatively accurate – on actual evidence so how could this be off? It’s possible that in Sander’s relationship with Magnavox knowledge of their manufacturing – the checks they received for the licensing – were not based on the same reporting figures as the sales from Magnavox. In other words: If the sales were marked for a year January to December and the royalties were scheduled from July to June, there would be a difference in numbers even when it’s twelve total months of product.


While secondary sources can’t always be accurate, there can be certain milestones which are more than likely as definitive as you can get to an accurate figure. For example in the book Atari Inc. Business is Fun, an article of unknown provenance shows the 10,000th Breakout PCB being produced and awarded to a distributor. These sorts of awards are always based on internal production numbers from the manufacturing people themselves, so while it’s not a document itself it’s just about as reliable as you can get for a secondary source. Plus, the internal Atari memo says a production run of 11,000 upright cabinets which at least partially corroborates this.


One can also put more weight behind secondary sources when several across time agree on a trajectory. Ms. Pac-Man for instance is often called the best-selling single unit arcade game of all time, with numbers ranging above 100,000 units. Among the independent sources which corroborate that are the Sanford Bernstein report in December 1982, Business Week in January 1983, and a memo from Mary Fujihara of Atari in July of 1983 looking at competitors games (which she tends to be rather accurate on). By 1988, the Toledo Blade newspaper reports 125,000, which can be taken with a grain of salt but doesn’t exist outside the realm of possibility.


Things get even sketchier when we move out of the arcade, especially in computer games. Because you were generally talking about sales an order of magnitude lower than even handheld systems, the figures probably weren’t generally seen as worth reporting publicly. However, one institution did get some direct sales figures as to allot awards to these games: The Software Publisher’s Association. Like the record industry, the SPA would award Gold, Platinum, and Diamond awards for achieving sales of 100,000, 250,000, and 500,000 units respectively. These aren’t direct sales figures, but they do give a date of certification as seen in the Computer Industry Almanac so one can estimate any further sales trajectory.

Starting in the mid-to-late 90s, some newspapers would actually get to publish sales figures on computer games as part of a wider interest in multi-media. These would include heavy hitters like Doom II (322,671 units in 1997) and Myst (853,765 copies in 1996). These rare precise figures reflect an increasing confidence of the computer game companies in their own accounting as well as showing their muscle up against the always encroaching realm of console games.


Console games are perhaps the most unreliable of the lot. Because the numbers being dealt with are massive, it’s extremely easy to over exaggerate a figure. Even a 1% increase on 1 million in sales is 10,000 after all, so handle such numbers with care. Generally relying purely on memory for these is completely fraught with peril, though at the same time it’s hard not to go with something when you’d otherwise have nothing. Several sales figures are quoted in the book Game Over for instance, and while some might be based on research or direct figures there’s no way to corroborate that.

Nintendo does generally publish the sales of their larger games within their annual reports, which can be a good source for sales up to that moment. There’s no sort of progression chart though so one has to dig further back to find how much it could have initially sold. One newspaper in 1988 for instance reports sales for the original Super Mario Bros. at 9 million in the US, which with the expanding NES market up to 1990 could lead towards the reported 40 million cartridge versions of the game before the discontinuation of the console.

What’s the take away from all this? Try to trust contemporaneous sources where possible, but understand the general limitations to any figures. You have to have a pretty good sense of numbers to make a call, looking at things beyond the sales numbers at things like general market projections (like if a market sector is projected at $2 million in sales for a year, one game probably didn’t sell 1 million copies right then). It’s guesswork, we can’t really claim to be better than analysts who were living in that period of time. Source acquisition is the main thing, and the historians are working on getting some of this information more widely disseminated where possible. There’s actually quite a lot of it out there, but the sources are definitely obscure.

If you have any questions about specific figures or more detail on how to parse them, feel free to get in touch or come by Gaming Alexandria’s Discord server!

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s