Skip to content
Culture & Community

Reading the charts

What a chart run says that a peak position does not

British magazine charts recorded monthly positions for a decade. Read as runs instead of peaks, they describe how games sold — and they contain traps that make confident nonsense of the careless.

sinclair-zx-spectrumcommodore-64pressukchartsmethod

CRASH and ZZAP!64 printed monthly software charts for most of the eighties. This library holds 1,354 chart rows across 876 titles, each with a date, a rank and usually a publisher — and read as trajectories instead of peak positions they say things no other source here says.

The shapes

Across chart runs of four or more months, seven recurring shapes account for roughly nine in ten:

Shape Example What it indicates
Enters high, falls steadily Cobra 3→14 A licence burning off launch attention
Enters low, climbs, returns Match Day 9→3, for a year Word of mouth
Slow climb to a December peak Enduro Racer 10→1 Word of mouth plus Christmas
Peak, drop, peak again Exolon 1, 3, 1 Sold on a named programmer
Flat for months Renegade #2 ×3 Steady demand, met steadily
High in genre, low in general Heavy on the Magick #1 vs #16 A devoted, small audience
Long low tail after a year Dun Darach 16–20 A catalogue title

Peaks hide all of this. Cobra and Match Day both peak at #3 and are nothing alike: one is a film licence decaying on schedule, the other is a game people recommended to each other for a year. A single number cannot tell them apart; six numbers can.

The traps

Every one of these was found the hard way, by writing something wrong first.

A chart is not always a games chart. CRASH printed a video rental Top 10 and ZZAP! ran readers’ polls headed “MY FAVOURITE ARCADE” and “MY FAVOURITE VIDEO”. A rank column looks identical whatever it ranks. This put Aliens at number one a year after release, above Blade Runner and Ferris Bueller’s Day Off — and produced a confidently argued, entirely wrong Vault entry before anyone checked the distributors.

A title is not an identifier. Uridium charts on the C64 in 1986 and again, much higher, in 1987 when the Spectrum version appears. Two Electric Dreams games were both called Aliens, scoring 84% and 45%. The chart row carries a name, not a platform, a version or a medium.

A publisher can belong to the neighbour. Where several games share a printed page, an attribution can bleed — the same hazard that puts Thrust under Elite when it was Firebird’s.

Absence is not disappearance. A game leaving the general chart often means it has moved to the Adventure chart, not the shops’ back room. Lords of Midnight, Fairlight and Shadowfire all make that move; the market’s idea of what they were changed while they did not.

Why the charts are worth the trouble

Reviews record what a magazine thought. Charts record what people bought, month after month, and they are the only serial quantitative source this library holds about the British software market.

They are also the only place the difference between selling and being liked becomes visible. A 96% game can sit at #12; a mediocre licence can take the top spot for a month and vanish. Both facts are in the data, and neither is in the reviews.

Who counted them

The numbers came from Gallup — the same firm that compiled the record charts. Big K introduced them to readers in March 1985 as “the Gallup (those lovely chart people) software chart”, and by that spring the trade was converging on Gallup’s as the single standard against the several competing charts then in print.

That is not incidental. The British software chart was built by the people who built the pop chart, during the years the record industry was itself moving into software — see When the Record Companies Sold Software. The conventions this entry reads for meaning were inherited, not invented.

See also

Not yet fact-checked. This entry was drafted by an AI and nobody has verified it. The dates, figures and technical details may be wrong. Use it to find your bearings, then confirm anything that matters against a primary source.