Pinball! Pinball! Pinball! Tournament report and format discussion


Last weekend I had the chance to play in the most… unique… of the annual tournaments on offer in Japan: Pinball! Pinball! Pinball! at Neverland, Toyoda, Tokyo prefecture, Japan.

Grand prize: A pinball. I’m not sure why. There appears to be a funny story somewhere here, but it’s in Japanese.

Format Rules: Head-to-head match play on few enough machines that there’s at least a player or two waiting at all times: I think the ideal number is ceil(N/2)-1.

At the end of each game, the winner gets one point. The loser stays at the game unless they’ve lost twice in a row, in which case the winner stays. The guy staying on the game plays in the same position (1up/2up) as they were already. First guy from the queue comes and plays that player, the guy that left the game goes to the back of the line and records the win and loss on a master list.

At the end of 3h, no more games may start. Most wins wins, ties broken by fewest losses.

Format discussion: Super fun. Use short-playing games for maximum turnover; people will tend to not wait around for tilt bobs to settle and so on because they understand the opportunity cost of doing so. I played 29 games in 3h: I’m admittedly a bit crap, but still. Note that the format means that if you walk up to a machine and lose, you always get a chance at redemption against someone just walking up to the machine. You also tend to play games you’re bad at more often than ones you’re good at, which is interesting.

Wait times in the queue were generally about long enough to grab a drink from the vending machine, but not much longer than that.

You basically play 2p games nonstop for 3h, with just enough time off to have a brief chat with friends. Also, I think that (per a discussion with @pinwizj a while ago, this will grade out quite well under WPPRs: we got craptons of games in.

Tournament bank:
Bally Wizard!
Bally Old Chicago
Bally Mata Hari (EM)
Stern Catacomb
Bally Radical!

A good choice for this format: these games all play pretty fast. The EMs at Neverland are always on 3-ball. Catacomb, with the backbox game, had the lowest plays at the end of the day, at about 20. Old Chicago played tough, with about 40 games over 3 hours, closely followed by Wizard! Radical! of course suffers from lock-stealing, but people around here are savvy enough to tilt out the lock strategically which greatly cuts down on the total number of multiballs.

My tournament report: It turns out that I’m absolutely dreadful at the skillshot on Old Chicago, which on a tough-playing 3-ball EM is a big deal. I think I went a combined 8-13 on Old Chicago and Wizard: at least I lost fast. When I did get a chance to play Radical! I did much better: I put up a couple of 10M+ games on a tough-playing Radical!, and eked out a couple of scores in the lower-single-millions by being better at the top ramp than most.

By luck of the draw I never played Catacomb, and didn’t play Mata Hari until about 2-1/2 hours into the tournament. After blowing the game up a few times in warmups I was obviously getting tired, and put up about 15k on my first game. Given the chance at redemption I put up…40k. Sigh. Thankfully, my opponent put up 35k or so, so I’ll take it. (He was pissed.)

All told, I went 12-17 (lose fast!) good for 5th/12.

This is my 2nd time playing this tournament. Last year’s bank had a few longer-playing games that were a bit unfortunate (AC/DC, Star Trek, and BDK, I think), but it was still good fun. This year, I hated Old Chicago by about 2h in, but in the best possible way.

I would tell you how other people did, but I was too busy playing to notice. Horiguchi Masaya様 (“Harry”, the Japan country director) won on tiebreaks from Naruke Totsunori様, one of the two really good Japanese players (and @Coast2CoastPinball’s pick for IFPA worlds).

Edit: Wow, this is longer than I expected. I’m happy to answer questions about the format or the Japanese scene as far as I understand it.

Tournament formats for group play
Different competition formats

Sounds interesting and fun, and I might use it for a future event. Thanks for sharing!

I guess there is no “round” play format, and the player coming out of the queue plays the very next available match.

Not only should fast-playing games be picked, but the tournament bank should probably be as evenly matched as possible. I.e., a single overly short or overly long playing game could unbalance results. Right?

I also imagine that the more players involved, the shorter the wait times in the queue would be overall.


Sounds like my old console days of playing tekken with a ton of my friends. Loser passes the controller and gets in the back of the rotation. Having the loser stay is a great idea for balancing things a bit more. Definitely looking forward to trying this out as well… Unique, simple and gets everyone playing for the whole day.


This sounds like a fun tournament. A couple things occurred to me when reading through the description.

It seems to me that loser staying could potentially create a strategy of throwing a match (or maybe just giving up on purpose if it’s already not close) against a “tough” opponent in order to stay and possibly get a win against a weaker opponent that might be next in line. The very best players don’t have much incentive to do this but the middle of the pack players might want to employ such a strategy in order to collect more games played and more wins against lower level competition. Doing two wins in a row before the winner must move to another game (in which case loser stays) might give more incentive to winning but still keep the play—and players—moving along plenty.

I also see a huge benefit to only choosing games that won’t last very long. There’s no incentive to win on a game that may take 15-20 minutes for two players when there are games that over 6 total balls, could be done in 5-10 minutes. Going 2-2 in 30 minutes is twice as good as going 1-1 in 30 minutes, even though in both cases you were .500 for the games played. Perhaps having a total number of games played limit versus a time limit could prevent shorter game times having so much impact? Or even a games played or reached time. So say 20 games max or 3 hours per player… whichever comes first.


Correct. This is a lot of what cuts down on the waiting typical of match-play tournaments.

A single short-playing game is fine as long as a player doesn’t consistently go back in the queue and get to the front again and get put on that machine again: i.e. it shouldn’t be more than ~N times faster than the N slower playing games in the bank. A single longer playing game is worse, but as long as it doesn’t play for an hour it’s probably not the worst?

In our bank the slowest game was about twice as slow as the faster ones and it was hardly noticeable.

The only thing you get from throwing here is time savings: throwing away a win for the chance of a win against easier opposition isn’t productive. This isn’t negligible, but it’s not a lot: remember that this gets the other guy extra games too. The time-spent-in-queue is actually quite low in comparison to time spent playing (you often have to write fast in order to get your results down before you’re playing again).

Loser-stays-on is also good for balance in a couple ways: you get a warmup game before playing against new opposition: (lose-win) is a fairly common outcome of a visit to a table; the bad (and therefore faster-playing) players get to play slightly more games, against cold opposition; and there’s no incentive to throw games on slow-playing machines to avoid them.

And this is the most important part of loser-stays-on: if you’re stuck on the slowest playing machine, you have no perverse incentive to lose to get away from the machine. Winner gets the reward of going to play something else. Remember that at no time does a player get to pick a machine, so incentivising good play on what you’ve been drawn on is important. Once you’ve been assigned a machine there’s no incentive to throw: you’re stuck playing it again if you do.

I think the total variation in games played in this tournament was about 1/4: 22-29 games per player.

The one bit of time-related shenanigans that does happen is if player 1 is way ahead on ball 3 how long should they play against weaker opposition. In practice this doesn’t seem to be a big deal, however.

I’d recommend trying it before you fiddle with it too much :slight_smile:


We’ve had a handful of tournaments using this format in the Bay Area (all organized by @echa). I have mixed feelings about the format. I think it’s fun for a casual tournament because you get to play a lot, but on the flip side it’s very random and I think there’s too much luck involved to say the winner is the best player on that day (i.e. I think it’s a poor tournament format for IFPA purposes).

I also really don’t like that if you are placed on a game you are really poor at you have to play it again. Personally, I find it quite demotivating.


Randomness aside, I am intrigued by this format because it is totally different. It seems like a fun diversion to keep the more casual tournament players interested and engaged. Could you clarify a bit how many machines vs. players you feel would be ideal? Can you also clarify how this format addresses the potential for cheating?


How were next games selected? I thought from your description that players were picking the lines to go into after they won.


I think having 1-2 people not playing is ideal for smallish groups. I think if you’re running this for 40 people+ you’d probably need a longer line just so people have time to get in the line and record results: say 10% of people not playing? I know neverland had the scoresheets set up to handle up to 24 players on 6 machines, so they must think it scales up that far. (That’s 50% of the group in line.)

There is exactly one line. Whenever a game needs a player, the first guy off the line goes and plays. Ideally only ~10% of the players are in line at a time, everyone else is playing.

I posted about last year’s tournament on Pinside. I think a few US organisers picked it up from that post (?).

The tendency to play games you’re bad at more than the ones you’re good at is an odd quirk of the system. I hated Old Chicago by the end of the tournament.


Because I’m thinking about running an “official” (IFPA) tournament with this format in August, I gathered some local players for a mini-runthrough yesterday. Details:

12 players
1.5 hours
5 games in the bank (Paragon, Firepower, Whirlwind, Black Hole, Mata Hari)
2 players in the queue

I didn’t find that too much that needed tweaking compared to what I could learn by looking back through this thread. But here’s what I’d do better next time:

A) Labeled chairs for the queued players to sit in or stand next to. Because of the layout of my club, it wasn’t always easy for an exiting player to determine at a glance who his/replacement was. We ended up going with a system where the exiting player would raise a hand, and the queued player was responsible for reporting to the machine in question. I think a small dedicated queue space will work well.

B) Because Whirlwind played relatively long compared to the others, 1.5 hours is a bit short for using this format to decide the best player. The winner of our event won 10 games and lost 1, but ended up on Whirlwind a bunch more than the other players and therefore played 5 fewer games than the player with the next fewest game plays, and more like 7 fewer games than the average player. (Part of that, of course, is because a more skilled player is more likely to be involved in longer matches.) The winner had the same number of wins as two other players, but won via the loss number tiebreak. The point here is that the winner was basically lucky to have pulled out the win, despite losing only one game in the 1.5 hour period.

C) Related to B, my plan is to have two 2-hour sessions of this format for the August event with a food break in between. The only question is whether session 2 will start where session 1 left off, or if the games will be populated randomly. I guess the former would be more fair.

D) Also related to B, I actually think having 2 medium playing machines is better than having just 1. Ideally, the game bank would be exactly even in terms of expected game time. But having just 1 seems to put a smaller number of players at a disadvantage. Not sure though – maybe having 2 slower machines would make things very even except for the 1 or 2 players who play a disproportionately small number of matches on these machines.

On the huge plus side, we had 4 players participate who are newcomers with minimal league/tournament experience. They all had some success (wins) and everyone said they had a good time.


The solution here in Japan involves magnets with numbers on. Next time I’m up at Neverland I’ll ask Horiguchi-san (Harry) if I can take pictures. The other thing that held the queue together was the scoresheet: after your game you had to record the winner and loser of the game on a master scoresheet, which kept people in the same place.

Often the player coming off the machine would be the guy that poked at the people waiting to actually go play when he got to the queue.


Thanks for sharing your experience with this format. I’m going to run a tournament on 6/17 at FreeGoldWatch in San Francisco with a slight variation on this theme. The difference will be that after a player has lost their second consecutive match, both they and the player that won the match go to the back of the line. This is to avoid the “timely win” situation where you beat someone and you get to stay on the machine as opposed to going to the line again. I’m also curious to see how the self reporting of wins/losses will go.


Keep records in duplicate. One set at the machine for that machine’s results, and a master list at the queue.


I ran the tournament Wednesday night and I wanted to share some of my impressions (and two conclusion) with you.

The facts: 38 players competed head to head on 13 machines, while 12 players waited in line at all times. (See picture of the play board I used.)

(The Freezer is a place where a player’s chip would be put in case they weren’t there when their turn was up. This idea turned out to be a good one. After said player showed up again, they were put at the back of the line.)

  • The length of the line (12 players) was OK. On average, people only had to wait 4-5 minutes which allowed for a quick smoke/restroom break.
  • Since I love playing more than organizing, there was no assigned score keeper. This was a weakness since a few people forgot to record their score. I think we caught all of these errors, but having someone designated to just deal with scoring is preferable. The people waiting in line were a resource that could have recorded scores as well since they helped out by moving players’ chips on the board anyway. I did take your advice to use duplicate scoring just in case someone would have claimed that they didn’t get their win recorded. Good tip!
  • Stuck balls/malfunctioning games now is harder to deal with since our operators also were playing. For them to go work on a game during the tournament would mean that they would have to lose out on precious playing time. Same for officials making rulings. Luckily it didn’t come up many times, but it’s definitely a potential pitfall.
  • The feeling I have is that the format is great fun! Everyone gets to play a lot, and wait less. However, it may be better suited for non wppr tournaments imo. The player who lost the most got to play 14 matches during the 2 hours. Another player who’s record was 7-1 only played 8 games. I would like to tweak it so that everyone plays the same number of games and go by winning percentage instead. (For example, the Golden State Warriors’ regular season record was 67-15, making them (by far) the best team in basketball with a winning percentage of .817. Had say… the Cleveland Cavaliers had a record of 68-30, or .694, they would under the current format rules have ranked higher than the GSW since they had one more win. Clearly, as we all know, the Cavs aren’t as good as the Warriors. Just an example though…

If you played in said tournament, feel free to chime in with your opinions. Echa? Andreas?


The format is intense because it’s fast-moving and that’s a lot of fun. I mentioned above that there’s a lot of luck/randomness involved and I still think that’s the biggest detractor of this format. The biggest advantage to me is that at least you get to play a lot of pinball.

The freezer was a great invention. It kept things moving and surprisingly I heard very little whining from people who was moved there. I think it helped that you announced in no uncertain terms at the start of the tournament that you had to be present when your name was called, no exceptions.

I’m all in favor of TDs playing in their own tournaments. Especially for these small (okay38 players not actually small) weeknight tournaments. What other incentive is there to organizing a tournament?! Having the winner report scores for both the winner and the loser was hard for people to understand (I don’t know why, it just was). Maybe have both winner and loser go report scores together and have the loser walk back to the machine with their new opponent?

Using winning percentage is a great plus in my book. I played great, but I played 5 games on slower playing machines (TZ and ST:TNG). As a result I only had 9 games for the night. If wins were the only determining factor I would have been in a massive tiebreaker for a spot in the finals, but since my record was great (7 and 2) I was moved into the finals above people with a 7 and 4 record (the eventual tiebreaker).

On the “softer” side I didn’t enjoy how this format made what’s normally my favorite part of pinball into a disadvantage. I had two games that were high scoring and very close. Normally this is very exciting and fun. Here it was very frustrating because the games were taking longer than the dozens of Paragon games being played.

In the same vein I intentionally drained balls two and three on my finals game on TZ just to get back in the queue. That wasn’t much fun either :frowning:

If the format is tweaked to have everyone play the same amount of games I would honestly rather prefer to just play a set amount of rounds of head-to-head match play. There you can ensure that players aren’t just facing the same opponent on the same machine over and over and over. It would take longer of course, but it’d be a lot less random. :slight_smile:


Great input Andreas. I did a bit of counting this morning and here’s a compilation of how many games were played on each machine:

Centaur - 17
Dirty Harry/TZ - 11
TAF - 15
Cap. Fantastic/MM - 17
Dr Dude - 19
Demo Man - 16
Tron - 12
WCS - 16
Baywatch - 12
Jokerz - 16
Paragon - 22
STTNG - 14
Mars Trek - ?? Probably around 20 ime.

Note. Cap. Fantastic & Dirty Harry had to be taken out and were replaced with MM & TZ respectively.


We’re running an exploratory Pinball Pinball Pinball next Tuesday at Logan Arcade! I’ll post with a recap once it’s done. Thanks to everyone in this thread for their commentary, experiences, and ideas!


As mentioned above, we ran Pinball! Pinball! Pinball! last Tuesday, and it was lots of fun! We had 16 players, so went with 7 tables at a time so there’d be a two-person queue. One player had to leave due to an emergency, though, so the queue was only one person - next time this happens I’ll reduce the number of tables by 1, because a 1-person queue gets confusing. We went for two hours. People played between 10 and 15 games - next time I’ll eliminate Medieval Madness and Whitewater, because they tended to take much longer than the others. I was divided about whether to do “most wins” or “winning percentage”, but it didn’t make a difference among the top half-dozen or so finishers anyway. I really like how well this format adapts to tables going down, too, in contrast to pingolf.

Anyway, we’ll be running it every Tuesday at 8 for the next few weeks at Logan Arcade in Chicago!


This is certainly an interesting and different format. Just out of curiosity, how does the TGP grading work for this? Is it simply the maximum number of games that anyone played that counts?


TGP is based on how many total games the winner of the tournament ended up playing (including their losses).