Visualizing the rise and fall (and rise?) of pinball

It’s not uncommon to hear people say that pinball is currently experiencing a resurgence. However, how big is this resurgence and how does stack up next to the previous glory days of pinball? I decided to try and answer these questions by making a data visualization of the plot using data from the internet pinball database. The plot below shows the number of unique titles produced, from 1960 to today.

The end result is striking. First, it’s just incredible how massive the pinball industry was in the late 70’s. At the peak of production, the industry was cranking out over 100 unique titles in a single year! I can’t even begin to imagine how exciting this would have been to experience. Shortly after this peak came the video game craze, which resulted in a dramatic decrease in production. This decrease would steadily continue until 2000, when pinball almost nearly died. Of course, there were detours along the way. In the mid 80’s, production briefly increased for several years before to downward trend continued again. During the 90’s production leveled out briefly again, before finally plunging to near 0.

This brings us to the current state of affairs. As 2017 comes to a close, there have been 20 unique titles have been released. Relative to the early 2000’s this is indeed a meaningful uptick. Current production, in terms of models released, are approaching those of the mid 1990’s.

While this is good reason to be excited, it is worth pointing out a few caveats. First, it is hard to directly compare modern pinball production numbers with those of previous generations. Modern pinball manufacturers have adopted a tiered production approach (e.g. pro, premium, LE) which was not used in previous generations. While these are considered distinct models, they are only slight variations on other models, which lowers overall production costs. One could reasonably argue that the modern numbers are artificially inflated. Second, the modern numbers also include re-releases of old titles (e.g., Attack from Mars), which do not impose the same development costs as entirely new titles. Finally, this analysis only includes the number of models released, not the number of units produced. Unfortunately, those data are not readily accessible in the IPDB without massive amounts of manual labor. If anyone has access to that data, please let me know!

In sum, while pinball is currently experiencing a resurgence, I think it’s fair to say that this resurgence is modest in size at best. We are a long way from the glory days of pinball. That said, any upward movement is good. Let’s hope that the trend continues through 2018 and beyond!


Speaking from personal experience, now is the best of times for organized pinball - - tournaments, leagues and shows. The 70s and 80s were far better for variety of games both as to frequent new titles and for number of different titles readily available. Back then, you could easily find several locations with 10+ games in most major cities and 100+ different titles in one metro area was common. A high percentage of bowling alleys had games, plus more freestanding arcades than there are now. Games were better maintained, on average. I doubt that location pinball will ever return to that level. (More details upon request.)


I wasn’t around in those times, but I definitely believe you–there’s been about a 9X increase in IFPA sanctioned tournaments from 2010-2017! 2017 IFPA Year in Review

Do you think you would draw a different conclusion if you had the actual production numbers of games to go by? Were there really 3-4x the number of games being produced in the late 70s than there were in the early 90s when you compare total numbers of games being produced? I wasn’t around for much of either of those eras so I don’t really know :smiley:

I don’t know what conclusions I’d draw with different data. I think the 3-4x difference in 90’s vs. 70’s is probably realistic though. A lot of those games in the 70’s had super high production runs (over 10K), whereas only a couple games passed 10k in the 90’s.

I really wish I had the production numbers available. The amount of work I’d need to do to collect them just isn’t worth it.

Production numbers are part of ipdb. As far as how easy the are to grab, I don’t know.

Right, but when I started this project, I couldn’t figure out a way to extract the production numbers in a way that would make them easy to process. Just last night I figured out a way to “hack” the IPDB search feature to get all the production numbers in one go. However, not all games have production numbers associated with them on IPDB, but are available on Pinside (Most Stern Pro models are this way). Seems like a lot of work to manually compile them together. I’m eager for someone to prove me wrong.

Stern has started to publish their production numbers? That’s news to me. Other than LE’s where they ballyhoo “only 400 made” etc.

What’s the source of the production numbers on pinside?

I take it back, “most” pro models don’t have production numbers on pinside, however some do. No idea where pinside gets those from. Nevertheless, the number of units produced for modern games seems unknown. Anyone know why stern keeps those numbers secret?

Is there any sort of equivalence between mid 70’s machines versus modern machines in terms of time to design and manufacture? My thinking is that while 100+ new titles in a year is great variety, that number of 70’s era machines wouldn’t compare at all to 100+ modern Sterns or JJ’s with their higher level of complexity.

Let’s also not forget that a lot of the games of the 70s were also retreads of each other (add-a-ball vs standard, 1/2/4 player versions, etc.)


I’d note that many modern games are terribly similar in layout, even if not EB/Std clones. I hear lots of “hey, doesn’t GotG look a lot like MET [or Kiss].” Of course, the extreme example is all the versions of El Dorado. But “back then,” it was normal to get a new title about once a month in most locations, and you wouldn’t see retreads in the same location since as a rule if free games were allowed, you’d only get that version, never the EB one, but if free games were illegal, you’d always get the EB game.

To follow up on this, just for fun, I went through 1976 to see how many games were repeats of each other or other types of edge cases.

IPDB lists 110 games produced in 1976. Of these I removed 15 because they either weren’t pinball or were one-offs/prototypes, and the Fireball home version. I also removed any game that didn’t have at least one photo, as it is hard for me to consider those as being produced in any meaningful sense. If we wanted to be more strict we might remove things that weren’t produced in any meaningful quantity or were just wacky foreign games but I was trying to get away from that kind of judgment.

Of the remaining 95 games, the lists drops to 67 when you take into account the copies/multiple versions. Additionally, another 8 were copies of games that came out in 1975 or earlier, so not “unique” to 1976, so I’ll pull those too. This leaves us with 59 games we can realistically say “got created” in 1976. That’s a far cry from the 110 we started with.

That said, the production numbers of those games, and remember this is just for the games that numbers are listed on, is a staggering 170,000. 9 games had 10K+ runs (when combining games with the same playfield into one production run number.) I mean, sure, the manufacturing for those games was easier and they were less complicated to design so you could crank them out faster. But damn.

By manufacturer:
Williams: 38,786
Gottlieb: 55,235
Bally: 76,169

By comparison in 1992 there were around 100,000 units produced of 15 titles, and almost half of them were Addams Family, Getaway, and Fish Tales.

I would love to see a way to compile this kind of data in a more meaningful way, but the way IPDB is currently put together I’m not sure how it would work without the kind of manual labor I did here (which took around an hour.) You would really need to re-conceive the database to formalize the relationships between games, which I don’t see happening anytime soon.


Production numbers don’t tell the whole story. Pinball has changed radically in the last 20 years or so. Prior to around 2000, there was no collector community. There were a few games in homes, but likely well less than 5% of the total games being built. These days a much bigger percentage of games are going into homes.

Yes, location numbers are also increasing, but they’re nowhere near what it used to be. In the 80’s, most every 7/11 store had two or three pins. It got so bad cities enacted rules to prevent small businesses from putting in pins and video games. If the pinball map was around back then, there would’ve been tens of thousands more games on the map.

What this means is that the average game is being played a lot less now than prior to 2000. Now we have pinball podcasts, tutorials, live broadcasts, pinball shows, video pinball games and forums like Tilt. All pinball related activity, but not actually playing pinball. Prior to around 2000, there was only playing pinball. That was the hobby.

I started playing in the early 70’s and it was nothing like today. While I’m happy to see the numbers continue to grow on the pinball map, I don’t see the hobby ever getting back to what it used to be. As long as the number of location games continues to grow, I’m fine with that, but there’s no going back at this point. For better or worse, pinball has changed in a very big way.


I think these two things are related.

In some ways, pinball has suffered the same fate as bowling and golf: excess technology. Both bowling and golf now rely more on how good is your equipment and how good is your “power game.” Skill is still important, but the most skilled players can now be defeated by “pretty good” players who have more power and optimized equipment. Analysis is getting to be a big thing, too, in all three. In the case of pinball, substitute “complex rule sets” for equipment and “time studying rule sets” for power. Being good at shot making, nudging and ball control means diddly on most Sterns if you don’t know all the nuances of which shots set up the really high value features. Compare that to any EM or most SS games where ball time was far more highly correlated with score than is true now. Step up to GotG and play for 10 minutes shooting for modes with one player [if your game lasts that long, which it probably won’t], vs. playing for 3 minutes shooting Groot multiball all day. Heck, close off the side drains on Star Wars and let some neophyte play for 30 minutes vs. you for 5 using the shot multipliers.

It’s a bit like society - - a widening distance between the “winners” and everybody else, and the learning curve to become a “winner” is steeper and longer for many people. You didn’t see factor of 100 differences in EM or SS scores; factor or 10 was rare in many cases. And people spending much more time talking about doing stuff than actually doing stuff.

Good or bad? Depends on your point of view. Different? Emphatically.


Great discussion. Maybe thinking about pinball’s current trajectory as a “resurgence” is a bit of a fallacy. It might be better to think of it as an evolution, and any comparisons to the past are just apples to oranges.

I think I get the technology argument for golf, but bowling? I admit to complete ignorance of the sport. But could a mid level amature really beat a pro using superior ball?

Also, let’s not forget the (now illegal) swimsuits that skewed the olympics… I assume biking is all about technology… I don’t think it’s unique to these sports I guess.

Okay, I had to get creative, but was able to get most of the production numbers out of IPDB. Here is a plot of overall production, expressed as the sum of all units produced that year.

Now, note that not all games have production numbers in IPDB, and I have no idea how accurate the numbers are. Also, ignore everything after 2000 because we only have data for LE and Premium models. So grain of salt and all that. Anyway, when looking at the data this way, the resurgence in the mid-80’s basically disappears. however, we get a nice peak in the mid 90’s.

Interesting things happen when we look at units produced per model across years. The data is expressed as a box plot here.

In terms of units produced per model, the bump in the mid 90s was very comparable to that of the mid 70s. They were producing about the some number of units per model, and there were several runaway hits. The biggest difference was just the number of models produced.


Bowling changed dramatically when balls got weights put in them and changed surface friction, especially in combination with the changed oil patterns on the lanes that alter how and when the balls hook. I see lots of young bowlers now who can’t shoot spares for sh*t, but still average 200 because their weighted, high-friction balls combined with throwing them to take advantage of those features and have the ball rotate sideways faster as it rolls forward down the lane, makes their strike percentage noticeably higher than far more accurate bowlers from 50 years ago. If you look at historical statistics on the number of 300 games bowled annually, which has gone way up, it’s quite clear. Oh, and that number went up as the number of bowlers went down vs. years ago - - the “number of 300 games per league game bowled” stat is even more skewed than the raw 300 count.

1 Like