Saturday, June 30, 2007


This summer has had a lot of games crammed into a short time. Or maybe it just feels that way because we were so busy with NMS. Anyway, coed astronomy recently played in BATH3 on June 9-10 and the Google Puzzle Hunt on June 23-24. Here the write-up for BATH3.

BATH3 (Picasa Album)

BATH3 was a Pirate themed all-weekend game (yes, everyone said "Yarrrrrr"... a lot...). There were several major differences between BATH3 and other weekend games, however. First, the majority of the puzzles were written by other teams, rather than by GC. Secondly, we were asked to bring tents and sleeping bags, and we actually camped over night, rather than solving clues non-stop. And lastly, there was this weird point based side-economy for bartering for hints and help.

The Clues

Each team playing had the opportunity to write a clue. The benefits of this were that during the application process, you had priority over teams who didn't write a clue, and then during the game, you basically got to skip a clue for free. It was an interesting idea, though not one that coed astronomy would ever be likely to consider. It's all of the painful stuff (logistics and site scouting) with non of the fun stuff (clue writing)!

I'm curious to know how this worked from a GC perspective, since it meant less work, but I'm sure it was also much more difficult to ensure quality control on the clues. There was also an incentive for teams to make their clue as difficult/time consuming as GC would allow, and I'd be curious if GC had to deal with this issue at all.

I think my general reaction to the clues was mixed. There were some really awesome ones in there (Lowkey's crates clue was awesome, and puzzles by Ian never disappoint), but there were also some not-so-great or at least could-have-been-much-better puzzles. For example, I'm usually a fan of origami puzzles, but there's nothing like spending a ton of time trying to get the parrot to fold right, only to find out that you didn't need to have folded him at all. Also, the battleship/nautical flags clue (is that a French pinnace in your pants or are you just happy to see me?) was such a good idea, but it was so hard for more than 2 people to work on it at once.

There was one clue that we got skipped over, and while we were pissed at the time, I think it was probably good in the long run that we got skipped. From NMS, my elements clue may have been "[h]ands down, the worst puzzle in the event", but at least it didn't get peed on by not one, but two pirates! I would imagine that Stanford would be a little miffed by gamers peeing in the hallways of Terman.

Along with the regular puzzles, there were a set of mini puzzles. Since we weren't driving or navigating, Jan and I monopolized most of these early on. They were a nice diversion when we were stuck on the actual clues, or when we were doing something that not all of us could work on. Because they were a lot of fun, it was rather unfortunate that they really didn't matter much at all until the meta puzzle, which I think most teams didn't get to do.

The Camping

So, unlike the members of my team who are frightened of earwigs or don't like the outdoors or whatever, I was actually very excited at the prospect of camping. I haven't gone camping in a long long time, and this game was an excellent excuse, and I thought this was the coolest "innovation" that BATH3 brought to the game. This also meant I was tasked with solving the "tent" clue. Really, it's only like 5 steps to assemble the tent, but when Yar pulled out one of the poles and said "ok, so we put this down on the ground like the foundation, right?", I decided it might be best if I took the lead.

GC provided burgers and hotdogs, as well as soda and chips at the camp site. Given the $100/person entry fee and the likely low budget of most of the clues, we were hoping for something a little better. Having just run our game a few weeks prior, we were very familiar with the costs of various game expenses, and so I think we were a little more critical of things like this than we would have been otherwise.

After dinner, we hit up the mini-carnival for our various dubloons. After looking over the trading game and realizing that it was conceptually mostly the same as my BANG12 trading game (start with a random hand and a limited amount of information, trade with teams for rules or cards, various combinations of cards are worth points), Yar and I immediately sold our whole hand to Ian's team for skulls. This is not to say that my BANG12 game was in any way original, but more that I thought it was such an awesome idea at the time, but hated it afterwards and don't plan on doing anything similar in the future. Trading games are so cool in theory--optimization, limited information, interaction with other teams--but in practice, I simply don't like them.

On the plus side, though we kicked ass on all of the other carnival dubloons. I solidly bested Captain Bloodbath at liar's dice (the Captain really needs to learn to bluff better!), Jan turned in her pirate panda (sword, treasure map, eye patch, and all!), Justin did something involving a canon, and Yar and I wrote a limerick which I will attempt to reproduce here from memory:
Captain Bloodbath was a terrible menace,
He sailed the high seas on his pinnace,
But when he went ashore,
To consort with a whore,
She said "my that's an awful small....... nose"

The actual camping was what you might expect--everyone was smelly, the tent was a little cozy, there was some sort of pinecone under the tent right in the middle of my back, etc. At least there were warm showers in the morning. I thought it was fun enough that I'm trying to organize a coed astronomy camping trip, at least.

The Skull Economy

At the beginning of the game, we were given some number of skulls (50?), each of which were worth a single point. For comparison, a clue was worth 100 points if solved without hints. We were told we could trade the skulls to other teams for any form of assistance. Because the hint system was very restrictive, this generated an entire economy out of the skulls. GC had a bare-bones (har!) staff, and so all of the hints were given to us at the beginning on pages with that silver scratch-off stuff.

As with most pre-made, non-timed hint systems, if you were stuck on the very last step of the clue, you had to take all of the hints telling you everything you did over the past 2 hours just to get the hint that you need. And in the process, you would lose something like 55 points. So instead, teams would purchase hints from each other for 5-10 skulls. It caused unprecedented amounts of inter-team cooperation, to the point that the 3 lead teams solved the popsicle clue together.

While it was an odd diversion, I'm not entirely sure how I felt about the skull economy. It was interesting to see Darcy going car to car pimping out hints, and it did cause more non-contrived inter-team interaction than I've seen previously. However, when it comes down to it, I would have liked to have had a better hint system instead, I guess.

The Route

Oddly enough, we soon discovered while playing BATH3 that they had a very similar route as ours during NMS, for the first day, at least. Coming only 3 weeks after, it's not as if they had any time to change anything after playing our game. Humorously enough, I think we were actually quite happy with their route. Unlike the teams that had played in NMS, I was the only one on our team who had actually seen all of our Santa Cruz sites, and we actually kept hoping they would send us to more of the same sites. Sadly most of coed astronomy still has never seen Blue Ball Park!


All in all, BATH3 was, as most games are, a great way to spend the weekend. We came in 5th or 6th, depending on whether you saw the stats on the day of or what they posted on their website. We're not sure, but we think the difference may be some shenanigans involving the beach boardwalk clue that we were suppossed to have been given credit for. Maybe somewhere on one of Rich's spreadsheets, we have a little asterisk next to our placing, indicating that under more correct scoring systems, we actually did much better. =)


Darcy said...

I wasn't the one pimping hints, Ian was. I just came over to you guys because Ian was off, well, pimping hints. ;)

Jesse said...

It's a pretty consistent criticism that the the canned hints were frustrating, and I'm glad the skull economy at least ameliorated that somewhat.

I liked the live-human-hint-line you had in NMS, but a) we were really short on volunteers, and b) the route we had committed when we picked the campgrounds limited our ability to rely on cell phones.

Rich said...

I was wondering about the 5/6 ranking flip flop as well... I'm pretty sure I remember the scores they had on the day of, and I think our score remained the same in the posted results, but yours dropped by 100 in the posted results, causing us to switch positions. The only reasonable explanation I could think of that might make a clean 100 point difference would be that the posted results listed you guys as having scratched off a 100 pointer on Sunday, so my naive, uninformed guess was that originally they had forgotten to penalize you guys for that? I suppose only GC would know for sure what changed. They did mention that they intentionally gave no one credit for the beach boardwalk clue due to its brokenness, but it was my understanding they did that in both sets of results. But really I have no idea.

Anyway, if I do put together a custom results spreadsheet for this one, I'll be sure to include an asterisk for you guys. :)

CKL said...

Funny, I heard from a couple of BATH3 teams that GC actually asked them to make their clues harder. (Yeah, yeah, "that's what she said.")

Alexandra Dixon said...

Well, I can 'splain about the scoring! I’ll post a separate comment about what it’s like to tweak clues for difficulty level.

We learned late in the day on Saturday that clue 9, the Santa Cruz Boardwalk clue written by GC at the last minute, was broken. {Aside: I personally am sooo pissed off that we allowed a broken clue into the game. There’s an explanation of how it happened, but not really an excuse. Because – THERE IS NO EXCUSE FOR LETTING THAT HAPPEN!!! So, abject apologies to the teams.}

Anyway...there were five teams who went to the Boardwalk and solved the clue. “Solved” in the sense that they got the answer that the broken clue would give you if you did the data collection then had the requisite "aha" and worked out the answer. Then of course, Golden Booty went one better and actually figured out, by a series of correct assumptions, what the real answer would have been had the clue not been broken. Showoffs :-)

So, on Saturday Greg and I agreed to give those five teams credit for solving the clue.

But then! Later on Saturday night Rich said his team had bypassed clue 9 and gone to clue site 10, and had wanted to go back and solve clue 9 but were told by the site monitor at clue site 10 that it was broken, and not to bother.

I was a bit confused about why Rich’s team had apparently willingly bypassed clue 9 to go to clue 10 ... but to me, at the time, it sounded as though Rich's team had been gypped out of the chance to get credit for that clue, so I decided to throw it out. Greg didn’t disagree, though he suggested we just give Rich’s team 100 points. To me, that opened a can of worms – what if there had been other teams who’d been told not to go to clue site 9, teams we didn’t know about, who should also be given the 100 points? So, it seemed fairer to just throw it out, which is what we did.

I think the scores announced at the party on Sunday might have still included the 100 points for clue 9, which would have put Coed Astronomy in 5th place, not 6th. But the scores that were posted on the web site were based on this clue not counting at all, hence the 5/6 flop with Blood Pirates.

Theoretically we were going to give an hour time credit to each team who actually went to the Boardwalk, but we ended up giving back some teams their answer sheets at the party, which contained the only record of their arrival times, so we weren’t able calculate total time on the course for both days. In this game finishing time only broke ties, and there weren't any ties, so it didn’t really matter (except for bragging rights, I guess).

But...after posting the scores on the web site a few days after the game, I was curious, so I went back over the teams’ arrival times that the site monitors had recorded at each clue site, and compared them against the personalized skipping schedule we had generated for each team. Once I had done that, it became clear that in fact Blood Pirates was correctly skipped over clue 9, and they had arrived at clue site 10 too late to have any chance at all of solving the Tipover clue then backtracking to the boardwalk, solving clue 9 and getting to the campground by 8:30, which was our cutoff time for teams to check in on Saturday night. So, the “bad advice” they got at the tipover clue site did not really cost them 100 points after all.

Greg and I discussed whether to update the scores on the web site – on the one hand, we thought nobody would notice at this late date, or if they did, care about the difference between fifth and sixth place... on the other hand, technically the posted scores are wrong. We agreed to leave it alone, but then I read Dale’s blog and belatedly realize – (a) you noticed and (b) you do care (somewhat). So - we’ll update the web site to have clue 9 count again!

Alexandra Dixon said...

After running three BATHs, I can say with authority: having the teams write the clues is actually more work for GC, not less!

For one, we had to be diplomatic when giving feedback and asking for changes, and also consider (a) how much time each team would be willing to spend on revisions, (b) that teams like to “own” their clues, so we can’t just tell them “do it our way” and (c) how much physical production time was involved for each clue – either by GC or by the team itself. It’s definitely harder on GC when you don’t have 100% editorial control, and the timeline is not 100% within your control either because you have to work around the teams’ personal lives and schedules, and also get the latest version playtested by as many people as possible. I think in a game where the teams contribute the clues, you also tend to have more extensive modifications than in a game where GC writes them, so often it seemed as though we were playtesting a brand new clue, and not really getting the benefit of iterative tests on minor tweaks.

So – about difficulty.

When we laid out the course and each day’s timeline for the actual game, we came up with a budget of roughly one hour of solving time per clue. So, that was the target solve time that we sent to the teams.

We considered that some teams might try to create difficult or simply time-consuming clues in order to gain a competitive advantage. But in fact, I don’t think any team actually tried to do this, to their credit. It did turn out that in playtesting some clues took way too long to solve, so we worked with the teams to fix this. But I don’t think teams did this on purpose.

In other cases, the clues were quick solves. We agreed, though, that if a team wanted to submit a clue that took less than an hour to solve, that was on them, because they had to know this would give them less of an advantage in the game!

Any changes we suggested to teams were to improve the clues, not to arbitrarily lengthen them.

As an example, in the case of the “Pieces of Eight” coin fragments clue, the first version had words that were easy to put together, but the solution method was pretty basic and obvious, so we asked them to change it so that the “aha” would be more satisfying. They did, but in order to incorporate a different solution method (caesar shift in one column when words were laid out alphabetically), they had to work within some serious constraints. The list of words this could work with at all was quite short, and unfortunately, the final words in the clue were MUCH more difficult for teams to assemble. I think what we (GC) should have done, but didn’t, was to remove some of the other aspects of this clue that made it tougher by, perhaps, going back to having every word start at the edge of a piece, even the edge of a 3-letter piece, so the words would “pop” more quickly. Another problem, pointed out by Blood Pirates (but not playtest teams!) was that while the set of nine words as a whole might have had a unique could put together seven or eight words with most of the fragments, then find that the remaining fragments didn’t form real words...the problem at that point is that you’d then know that one (or more) of the words you had put together had to be wrong...but you didn’t know how many, or which one(s). That was pretty frustrating, I’m sure. If I had it to do over again, we’d have modified this clue yet again, even if we had to ask the team to go back to the drawing board on the solution method.

Alexandra said...


About the $100 per person entry fee - we actually priced BATH to break even!

We deliberately restricted team size to four people, partly due to having to reserve the campsites six months in advance, but also so that teams would not have the added expense of renting vans.

I haven't done the final accounting on the game yet, but I do believe we barely broke even. At least, just before the game we did a quick-and-dirty tally of what we had spent up to that point, then we budgeted the final expense items (food, etc) to break even, and we spent about what we budgeted.

In addition to the typical expenses GC might incur in a weekend-long game, we had some extra expenses.

First, we paid $600 for 50 flash drives because we were originally going to use an electronic hint/confirmation system. We only needed 30 plus 5 spares if the game sold out, which it didn't, but it ended up being just as cheap to get 50. Then we didn't use them. So that's $600 out the window!

Each campsite cost either $25, or $35 for a premium site (and we scooped up ALL the premium sites at New Brighton for June 9th) PLUS a $7.50 admin fee - per site! And the group campsite was $125. We had 16 campsites total, most of which cost $42.50 each with the admin fee. Plus, we paid for a bunch of sites for the playtest weekend May 5-6.

We paid for dinner Saturday night (about $500), and food for the party Sunday evening ($600), as well as $650 to rent the hall, and $600 for event insurance.

We also paid $400 to buy 40 t-shirts for GC, volunteers, and the winners. And they cost that little only because a friend of mine pretty much did them at cost as a favor to me.

Renting two vans for the weekend, to transport the tipover clue, the origami clue, and all the food for both dinners, cost about $550! Food, etc for several couch playtests was pretty expensive. We paid to produce most physical clues, including the metapuzzle. All the paper - I think with the minis, the skip instructions, envelopes, etc we probably paid $1000 just for paper and ink. The silver stickers were a last-minute purchase, and not cheap.

Even now, I look and think it couldn't have cost nearly $10,000 to produce that game - but it did!

So, bottom line - whatever money we had, we were willing to spend all of it to make the game a better experience for the players. If the food at the campout wasn't to your liking, I apologize - but it wasn't a question of money.

It was probably more a question of planning - despite Greg and I spending pretty much every spare moment of our personal lives on this game from early November on...and we micro-managed EVERY detail at first...I think perhaps it was a question of work expanding to fill the time allotted to it because toward the end, we found ourselves with too much to do and not enough time to do it, and some elements suffered as a result.