I don’t consider myself a numbers guy, but there is something fascinating going on over at the Internet Pinball Serial Number Database (not to be confused by the equally useful Internet Pinball Database). IPSND webmaster Jess Askey has created a project that collects serial numbers from pinball machines and pinball parts, mostly through user submissions, and compiles them for public record. From the information collected, Mr. Askey presents his own analysis and identification of trends, and, since the data is open source, allows visitors to identify trends of their own. Personally, I have been a member of the IPSND for over two years and have recently volunteered my time to take care of some of the site’s administrative and moderation duties. Ever since I started doing interviews for Credit Dot, I’ve wanted to ask Mr. Askey a series of questions about the IPSND’s history and vision. In my opinion, the site is an underused resource in the community and deserves the attention. Please read on as Mr. Askey unravels the history behind the numbers.
Credit Dot: How did you originally become interested in pinball serial numbers? It’s kind of a strange fascination…
Jess Askey: Well, I suppose it all started sometime in 1999 when I decided to buy a Jungle Lord from eBay. Back then, it was very rare for auctions to have photos but I found a good deal on one and decided to buy it. It shipped the old fashioned way: Forward Air, no palette required! When it arrived, I was very upset because the owner had clearly painted the cabinet blue (with red and yellow stripes) and not disclosed this information in the listing. I shot him off an email and he said that all Jungle Lord games were blue. He said I was crazy to think that it should be red. However, all the Jungle Lords I’d seen in my hometown were in red cabinets when I was a kid. A quick posting to rec.games.pinball revealed that the majority were blue, only a few were red. It seemed that the running knowledge amongst the RGP community suggested that the sample games were red and the production games were blue. The guess at the time was that Williams was making about 100 or so sample games for each title. I then did what everyone else did at the time for the games that they were interested in–I made a game owners list and rule sheet for Jungle Lord (archived here). People of course e-mailed me their serial number and I asked if the cabinet was Red or Blue and I updated the owners list until, just like every other owners list out there, I stopped updating it due to lack of interest and lack of time.
CD: What renewed your interest?
JA: It sat for a few years until I decided that I would create the Internet Pinball Serial Number Database proper, and use it as a “practice site” for learning all the new programming technologies that interested me in my professional career. ASP.NET was the first challenge, Microsoft SQL Server was the second. I knew I didn’t want to conflict with the already established Internet Pinball Database, but I used their game identification numbers as my key for simplicity sake. I also had to get a listing of all the games on the IPDB along with pertinent info (number of players, manufacturer, release dates, etc). I wrote a little application that programmatically started at game number 1 and went all the way up to something like 4500 (which was the “newest” IPDB entry at the time) and screen scraped the data off the IPDB web server responses (Sorry Wolf!). Now I had a starting list of games. Next, I got the full listing of old pinball serial numbers from the deprecated and outdated Pinball Pasture/Internet Pinball Project, from Daina Petit (you can still find the original Internet Pinball Project on archive.org if you want to peek at it). From that, I had my first database of serial numbers.
CD: What was the original philosophy or goal of the serial number project?
JA: Circle back to the original Jungle Lord owner’s list project and its inherent problems–my goal was for the new IPSND to do three things:
1. Allow submissions of serial numbers to my database which was ‘in-sync’ with IPDB games.
3. Allow tracking of specific game traits (like cabinet color, or other minor changes that happen throughout the production run of a game) along with serial numbers, so collectors could see trends on those game traits and understand if they were specific to the production run. In the case of Jungle Lord, it also allowed us to know that there were more like 400 red cabinet Jungle Lords and not 100 as was originally suspected.
Points two and three were the big long term goals as it really allowed the IPSND to become a centralized storage mechanism for all these game serial registrations, but also allowed that centralized knowledge to be shared outwards to other sites and collectively freed the owners list community from needing to waste minutes a day to update their sites with submissions. Owner’s List site owners rejoice!
CD: How long has the IPSND been active?
JA: I first registered and put the site up in 2006 with about 7500 serial numbers from the Internet Pinball Project. As of February 2015, there are 32,917 serials registered by over 4900 different people.
CD: In what ways has the site changed from when you first started it?
JA: Well, it started simple. On launch day, it just allowed you to search games and look at serial numbers, plus submit a serial number of your own (without the option of submitting a photo). By 2008, we had added photo support, our “nudging” system, the SerialBot points rating system, an RSS feed, geolocation mapping using Google Maps, bulk uploading, mobile Android support and a host of other features. Late last year I added backglass images to help submitters get the correct game more often. To see a log of all the changes, updates and feature announcements over the years, readers can view them here.
CD: What are you attempting to achieve by collecting all this data?
JA: Well, outside of the 3 goals I started with, I just continue to enjoy seeing the data come in, and with the help of other members, come up with interesting questions and solutions to fixing them. My overarching goal is that eventually this whole thing will go into the cloud and will be improved and maintained by other collectors long past the time when I’m gone (maybe?) Since serial numbers get submitted multiple times throughout the course of a game’s existence (as a game is bought/sold, or shown at a pinball show), in like 50 years, not only should most circulating games be in the database, but you will be able to see how they have moved around the country, or even the world. Perhaps this becomes the main registry for finding a game you want to buy, or to sell?
CD: In what ways can this open-source information be used by collectors and pinball fans?
JA: That is a great question. I suppose I don’t have the answers for this but I definitely see that this should become something like the IPDB that is a public repository of all information. It makes sense for the IPSND and the IPDB to combine at some point, I would think, but I don’t think Wolf and I are up for that challenge yet. Maybe some awesome young pinball coder will come forth and take on that responsibility at some point…
CD: What is your day job? I’m guessing something to do with statistics?
JA: Ha, not so much, but I guess in a way. I currently am an I.T. Project manager for New Belgium Brewing in Fort Collins, Colorado. If you pinball collectors are ever in town for a Brewery tour, let me know, I will come have a beer with you and we can talk pinball. Beer + Pinball = Awesome!
CD: Say someone new to the site wants to submit the serial numbers for their games. Can you outline the steps one would take to submit that serial number into the database and maximize its usefulness? Is there any other information other than the serial that needs to be collected for submission?
JA: Sure, that is the point of the IPSND and it is amazing because I think this kind of stuff actually happens. I get lots of non-pinball people who submit serial numbers just because they found a game in their basement and somehow found my site. In general, to submit a serial number, it is very straightforward.
1. Search for the game name and make sure you find the correct one (we have added backglass images for the more popular games to help identify them better since some are very similar: for example, Bally’s Capt Fantastic & the Brown Dirt Cowboy and the Captain Fantastic Home Edition)
2. Look at the serial numbers for that game already submitted to make sure that your number is actually similar to the submitted ones (there are lots of non-serial numbers printed on games). Alternatively, you can click on the ‘finding the serial number’ tab for game listing to find user hints on the best way to accurately find the correct number to submit.
3. Click on the ‘Submit Serial Number’ button and fill out the form. Adding a photo helps immensely, so try to add a photo of the actual serial number if you can. That’s it. Done!
If a game has ‘traits’ (color variations, feature variations, etc.) then on the serial submission form you can select if your game falls into those categories, otherwise, you can simply leave them at ‘Unknown’ which is the default value. The submission form has several other pieces of information that you can add if you have it: date stamps on the game, whether you are physically looking at the serial number or perhaps you are just looking at a picture. There are little ‘?’ icons on the form that describe what each field is for.
CD: The database is “peer reviewed” by a nudging system that was briefly mentioned above. Can you explain how that works?
I added this feature because the one fatal flaw in the original system was that these numbers could be totally fictitious and made up unless there was some sort of physical proof of the serial number’s existence. So, I figured that it would be great for a submitter to upload a photo of the actual serial number. However, one of the first issues was that users were submitting photos of their machine, and not of the serial number! So, I figured that leveraging the IPSND points feature (members get points for doing various things on the site, submissions mostly) might make sense to allow the members to have a vote on the quality of the photo. So, basically that is what happens! When a submission comes in with a photo, it is “Open for Nudging”–meaning that another member can look at the photo, compare it to the submitted serial number and give it a ‘Thumbs Up’ or ‘Thumbs Down’ vote on the accuracy and quality of the photo.
The nudging process is only open until the voting swings +3 or -3 nudges (and has a minimum of 5 total nudges) and then it closes with the result of being ‘Nudged Up’ or ‘Nudged Down’. On the submission page for the individual serial number, this nudge result affects the points that the submission gets, and ends up being +3 points or -3 points accordingly. So, if a person submits a serial number with a poor photo that doesn’t show the serial number clearly or accurately, the submission quality will get dinged points. Additionally, members get a point for each nudge they complete. As of now, our highest “nudger” has examined and voted on over 7500 photos. Some members submit rarely, but nudge regularly. If there is good consensus on the nudging, the submission will close quickly, sometimes within minutes, but some submissions generate a discussion-heavy nudge history and some submissions have taken over 25 nudges simply because the voting result never swayed more than +3 or -3 over the opposing vote. You can see the Most Contentious Nudges and read the comments on the IPSND Statistics Page. I think the best part of this is that everyone judges a photo differently.
CD: You mention a point system that is in place. Do we win anything if we collect enough points??
JA: Well, you know how we humans work! Unless there is a way to create competition, some things can get boring pretty quickly…especially serial numbers. I figure that the points were created simply to let people get credit for their work and effort they put into the site. Originally points were only awarded for submissions. Points are now awarded for nudging, too, but the original motive remains: to try and get a high quality score through quality submissions. Right now, IPSND member John Vorwek has over 13,000 points on 2,270 unique serial number submissions. John has been supporting the site for years and has become very good at scouring eBay for good serial number photos and scrutinizing them to make sure that the game assignment is accurate. Some members have very few submissions, but mostly concentrate on nudging to get their points. On the topic of winning something: right now you just get name recognition on the site. I definitely owe a small handful of people an IPSND t-shirt, so I figured that I would get a bunch silk-screened and pay off my debts to supporters of the site. However, there is no formal time frame for that.
CD: The software you use to run the site and gather information is very intuitive–for example, your “Serial Bot” recognizes and alerts the submitter when a serial number doesn’t match the known format used on a particular game’s run. Can you talk a little about the software/code used and how it was developed?
JA: The IPSND site was actually created by me sort as hobby for the reasons I spoke to in the very first question. I have done enough software in my life that I didn’t want to make something that was just a database. The things that give systems and organisms something special are ‘closed loops’ which means systems collect, recycle and re-feed data in such a way that the whole system or organism evolves. While I’m no rocket scientist, the premise of the site evolving and potentially ‘self-governing’ itself seemed pretty cool to me: make some basic rules and let the site and members go at it and see where it goes, then modify the rules to keep it balanced. The site actually attempts to dynamically calculate everything. For example, the points are not scored once and then saved, but the software calculates the points on each load of the page–that keeps the math out of the database and makes sure that everything is interpreted at the time of viewing in case the rules (for accruing points, as an example) change.
CD: References to “The Internet Pinball Project” have come up rather frequently on the site and in this interview. Can you give a bit of history about them and how you came to inherit their collected information?
JA: The Internet Pinball Project was an old part of the original Pinball Pasture which started back in 1997. The Pinball Pasture was created by Dave Byers and included information such as the original Internet Pinball Database (before Jay and Wolf took it over to ipdb.org), playing tips and strategies, a Pinball Glossary and the original Serial Numbers Database. (check out the wayback machine here to browse it from 1999.)
Even in 2004, the original database was getting stale and the process of e-mailing in serial numbers and logging them manually was difficult. Additionally, there was no mechanism for submitting photos at that time (remember, this is back when digital cameras were not very common, much less phones with high-res cameras in them). When I started the IPSND, I posted on RGP about the old Pinball Database, and luckily Daina Pettit (of Mr. Pinball Classifieds) had saved a copy. They serials were only organized by ‘Game Name’ and manufacturer, so I had to do some cleansing of the data and also determine which exact ‘Game Number’ that each serial was for (for example, if a serial was for Williams 1960 Black Jack or Bally 1976 Black Jack). So, the database was born using that data. Thanks to Daina and David for all their work, because it certainly helped start with 7500 serial numbers rather than the 13 I had in my personal collection! One final note is that in talking to Daina about this, he requested that I make the site have a feature that would let people download the entire database for whatever purpose they chose at any time. I thought that was both a good idea and also a fun technical challenge. To this day, you can download the entire core database in a single, flat CSV format (.zipped and e-mailed to you) if you want to do any external analysis…or otherwise.
CD: What are some of the more interesting statistics you have extrapolated from the submitted information?
JA: Well, I have a handful of cool things that I reference…
1. Williams 1981 Jungle Lord – It was always guessed by the RPG community that there were approximately 100 ‘Red Cabinet’ sample games made. It is pretty clear now by looking at the data, that there were actually about 425 of them made (see here). You can see the early serial numbers are grouped by the ‘cluster tool’ on the left side of the table. Although we clearly don’t have all 425 accounted for and we probably never will, due to the habit of route operators taking games to the dump, it at least shows us the upper and lower bounds of the range for those particular games. I don’t think that the data could present itself any better in this particular situation. Luckily, this was the original question I had for building the database, so I seem to have found the answer.
2. Williams 1991 Black Knight 2000 – If you visit the page for BK2K here, and click on the ‘Game Traits’ tab, you will see something interesting. Summarized in the top table is a listing of two ‘properties’ for the game: the style of plastics on the game and the color of some lettering on the backglass. It was postulated on RGP, that both of these variations might have been related to the “sample game run” or early test versions of games. By looking at the gathered data, you can see that both the plastics artwork and the color seem to actually be dispersed across the entire production run, not just early run examples. So, more than likely, these batches were not made sequentially, but the assembly line were just delivered “either/or” versions of the plastics and backglasses, perhaps coming from different printers. As we get more data, we may start to see the smaller patterns of the game plastics flipping between the two styles, which may tell us how many plastics sets came on a palette from the supplier. Keep in mind that Williams also had two production lines, so maybe one line had the ‘futuristic artwork’ and the other had the ‘Stone Castle’ plastics. If I had to guess, this second scenario was more likely.
3. Williams used a single number sequence when making games, but as I mentioned before, they also had two production lines running at the same time. Imagine how games are created: they are done in “runs” or “batches”. If you take a look at this page or the graphic below, (it looks rough, apologies), you will see a listing of all 1980s Williams games with the highest and lowest number in a clustered range. You can see how most games are listed in this sequence, sometimes two to four times, as they were being built in “runs”. Let’s use Black Knight as an example (RED Color Block). You can see it first appears in the list around the time of Firepower/Blackout/Algar, which were the last of the System 6 games. Having Black Knight appear in that run shows that Williams actually made 42 Black Knight games at the same time as the System 6 games (and most likely, that is the version of the game with the System 6 hardware). Now, jump down to the next Black Knight (about 15 rows down): they made exactly 100 games in this run. Those must have been the sample games. You can see that they were finishing up with the Blackout and Alien Poker runs at this time too. I think that this view is pretty interesting because it really shows you how often they flipped games on and off of production, and this was probably very tightly related to pending orders from distributors. If you continue to follow Black Knight, you will next see the big run of 10,000 games and then a couple “clean-up” runs which probably took care of remaining parts inventory.
CD: Is there a socially acceptable way to collect the serial information when in a public setting like a show or a friend’s collection? Do some people view it as “stealing” their personal information?
JA: Yeah, that is a great question, too. I think it sorta depends on how you approach it. If you are at a show and are crawling inside of people’s games without permission then obviously, they might get a little miffed. The best approach is to always ask and take the opportunity to inform the owner about the serial’s value and how their data may help clarify some interesting things for the pinball collecting community.
Along those same lines, I certainly have received numerous e-mails from individuals saying that someone else had registered their serial number on the site and they would like it removed because they were not the owner of the game. I think in part that comes from the site allowing people to print out ‘Serial Number Certificates’ for any submission which look nice but really are meaningless. When this happens, I try to explain the bigger picture goal of the site and also the main point is that we don’t just want ONE submission for each serial number, but we want MULTIPLE submissions of a serial number over time. Because we also allow submitters to track the location of a game (in a general way, you probably don’t want to put your specific address in there), we can see how games have travelled around in their lifetime. This type of data is going to take lots of time to gather, but my best example right now is this particular serial number for Aquarius which seems to show up often at the Pacific Pinball Expo.
CD: Do you get the feeling there is a negative perception in the pinball community toward people who are focused on the collection numerical data? I’m reminded of the somewhat unflattering light that the filmmakers used when chronicling old-school data collector Sam Harvey in their Special When Lit documentary.
JA: Well, like I said above: if you approach people the right way, then I think that a lot of benefit can come from the conversation. The thing to understand about Sam is that back in the 80s and 90s when the pinball collecting community was much smaller, everyone knew Sam and they knew what he was about. They expected to see him coming up to their games, opening them up and digging around until he found every serial number in that game. As the community grew (and changed), Sam’s behaviour didn’t and maybe it rubbed people the wrong way. Some people could get really pissed if some dude with a big afro was digging around in a game that you just restored. I think that what Sam was doing was a remarkable undertaking and I wish that I could get access to his books upon books of serial numbers recorded numbers. However, I would also expect that there are a significant number of errors in his recording of serials just because they are so damn difficult to read sometimes.
CD: In your estimation, what percentage of all existing pinball machines has the site gathered the serial numbers for?
JA: Well, it is pretty small… I think that with the games we know of what the production run was, we get more than 3,300,462 pinball machines that have been produced. Since we don’t know production numbers for many of the ‘games of olde’ when pinball was really booming, I’ll bet that number is only 75% of the actual number. Right now there are 32,191 submissions for about 30,000 unique serial numbers. Thus, we only have .9% of all games registered. Looking at specific games that have a decent production run (reference the site statistics page: http://www.ipsnd.net/stats.aspx?id=2), we have 15.2% of all The Addams Family Gold games, 36.7% of Spacelab games and 8.4% of the Safecrackers. So, once you start looking at specific games, the numbers are a bit more impressive!
CD: How can the pinball community help grow the IPSND?
JA: Well, I think that it is sort of happening naturally. First and foremost, we can grow the site by talking about it and sharing the benefits. I think that conversation can ignite the curiosity that all of us pinball collectors have. Obviously, making sure that all your games are registered on the site, with a photo and a geo-location, helps move the site to its long term goals. Making sure that the data is correct is also important.
I have received help from many leaders in the pinball community in both organizing the site, submitting serials, and submitting traits…
• ‘Pistol’ Pete Haduch has been helping me on the site for years as far as fixing submission errors, following up with questions, approving tips and traits, creating game serial masks and overall keeping things in line.
• Convention Gatherers like Basil Leblanc, Mark Gibson, Dan Gutchess, Seal Clubber, KoP and yourself submit serials from games at the various shows around the country and can give us a snapshot of what is being shown, bought or sold without actually attending these shows.
• Service Gatherers like Daina Pettit, Ray Johnson, Antti Peltonen submit serials for games they service (these are the games often buried in people’s houses)
• Jon Vorwerk patrols eBay and doesn’t let a serial get past him. Of particular note here is that Jon is continuing to submit valuable ‘Game Part’ serial numbers, which, while are not associated directly with a game, helps us determine if a serial number exists for the future when we start to analyse bigger data trends.
• Pinball Eric helped out by giving all the serial numbers from the Pinball Price Guide (very nice of him)
• Jay Stafford and Wolf at IPDB.org have been very supportive as well, allowing me to integrate with their data and make our sites work in harmony. I think that if our sites were on the same technology (IPDB is PHP and IPSND is ASP.NET) it would make sense for them to merge someday… that is probably for the next generation of pinball collector software guys though!
These folks and many others are the real winners of the site.
CD: Can we help with a monetary donation to keep things running?
JA: I definitely get Paypal donations for the hardware of the server and the internet connection. Everyone that donates gets a little dollar sign next to their name on the site. I get about $100 per year in donations which helps a bit with costs (cost averages about $750/year). If you like the site, I would love to move it to an Amazon cloud server or something so that it isn’t sitting in my basement at risk of being knocked over and destroyed. Anyone own a hosting company? Ping me!
CD: Are there any concepts on the horizon to make the submission process easier? An “on-the-go” app for smartphones?
JA: Yes, I currently have an Android app made with Phonegap in the download section of the site. It works, but is a bit clunky. It certainly makes submitting games very easy with a smartphone, though. Anyone have any Phonegap skillz that could help out? Geez, I’m really starting to beg here!
CD: What challenges lie ahead for the IPSND?
JA: I think that with any big project like this, the guarantee that the data is secure and lives on is the biggest long term challenge. I will continue to make the site better in small ways, but that is time reliant, and my children really suck my time away from coding on the site (as they should). Luckily as I mentioned above, many people in the community are actively helping on the site which keeps it moving in the right direction. The thing I want this site to transcend is the 20 year lifespan of many sites. I want to eventually get the code for the entire site checked into GIT and made public so it can be maintained by multiple people and put into the cloud. Then barring some sort of apocalypse, the site should run for a very long time and this information can stay growing and informing into the future. With companies like Jersey Jack and Stern Pinball, luckily we still have new pinball machines coming off the production lines and keeping our passion alive for future generations.
CD: Finally, what games are in your collection? What era of games tend to be your favourite? What is your favourite game?
JA: I have a smaller collection than I used to, but I’ve whittled it down to my core games. I started collecting when I was 13 when my dad bought me a Williams San Francisco. I still have that, plus some video games, which are also my passion. Other than the San Francisco, I have a Tales of the Arabian Nights, Hyperball, Time Fantasy, Cosmic Gunfight, Solar Fire, Jungle Lord, Warlock, Joust (the pinball machine), Centaur, Big Chief, and a 1934 Pacific Novelty Contact. On the video game side, I have Robotron 2084, Mystic Marathon, Xybots, Alpha One (Major Havoc Prototype) and a Williams Spellbinder which was never released and is currently being rebuilt. Obviously, I love those early 80’s Williams games. But, Centaur might be my favorite!
The Internet Pinball Serial Number database can be found at http://www.ipsnd.net, and relies heavily on the support and submissions of the pinball community. If you have not submitted the serial numbers of the games and parts in your collection ,or the routed games on locations near you, I strongly urge you to do so. A greater understanding of the pinball business and trends within the collecting community await discovery…using the information gathered by pinheads around the globe.