CORDIAL MINUET ENSEMBLE

Yeah, it's fine to talk about it publicly or show videos of it.  Just don't give away the secret words!

Thanks for sharing your thoughts!

The penny buy-in is supposed to help brand new players.  $2 can get you 165 games even if you lose every single one.  That's a lot of games to learn on, with only a penny lost each time.

I get the anxiety over other games showing up invisibly while I'm waiting for an opponent.  Like, creating a 50 cent game, but sure, I'd join a dollar game too.  Right now, it would make sense to list some of those games right on the waiting screen.  I guess it can list three games there (the three that are the closest to the amount you requested).

The other option would be to set a range that you're willing to join.  "Call my game 50 cents, but I'm willing to join anything less than 2 dollars."  The problem there is the UI for displaying whatever match comes along.  Surprise, you just auto-joined a $1.72 game.  I'd rather keep it cleaner than that.

So, I think the solution would be to list close-valued games on the waiting screen, but just list them, not make buttons out of them.  So you do have to cancel your game wait and go back to the menu to join another game.

I'll have to think about ways to only have friends join your games.... seems like a simple password on game creation, and then a password search field on the menu screen could work.  This would also work great for one-on-one setups at my PAX booth, too.  But for the booth, I'd really want to lock the pair of machines into using the same password... hmm...

In the mean time, here's a trick you can use:

When you "create" a game, you're actually just asking to join a game at that value.  So there will never be two 50 cent games listed on the menu, for example.  Even if a second player tries to create one, they will just end up joining the existing one.  You can use this to join a friend by picking a unique money value (like 38 cents) and the both creating that game a the same time.  You will very likely join each other before anyone else even gets a chance to even see that your game exists.

No plans for alias changes.  I know that some people are unhappy with theirs, but the two random words are really a nice touch.  I don't plan on having those show up in the game itself.  They're just for recognizing yourself on the leaderboard.

Graphics are pretty much done.  I really wanted a clean distinction between the "digital" parts of the UI (widgets and stuff) and the "handmade" occult artifact in the center.  It was important to always keep the handmade part consistent.  Paint can only be added, never removed, etc.  It always behaves like a real piece of paper.  I don't want parchment buttons and widgets.  I don't want a parchment credit card entry screen.

No plans for music in this one.

I intentionally designed this as a skill-based game, but it's a "soft" skill that is more of a hard-to-define art.  It doesn't need a ranking system in the same way that chess does.

Yeah, this stuff will be tweaked eventually, and new leaderboards can be added.  All the data is in a MySQL database, so basically any query we can dream up, we can make a leaderboard out of.

The point of getting those in place now was to give us something to start with and support time-limited tournaments (where the dollar balance leaderboard will actually be the winner's leaderboard).

Okay, I put a 4-coin rake cap in place per round.  So, the rake is still 10%, but the rake stops increasing when the pot is over 40 coins.

Pot Size vs Rake:

0-9  : rake = 0
10-19  : rake = 1
20-29  : rake = 2
30-39  : rake = 3
40-200  : rake = 4

So, the skill cap you're talking about comes from lack of granularity in betting as the small stack gets smaller, right?  At the start, we both have 100 gradations between the ante and all-in, so choosing the right bet is a subtle thing.

Toward the end, when one player is down to 10 chips, then there are only 10 betting gradations.

It's almost like, at this point, the problem could be solved by taking the smaller stack, dividing it into 100 parts, and issuing each player differently-valued chips.  That would restore granularity and thus push the skill cap back up, right?

I'm not saying that I would use this solution (for many obvious reasons).  It's just an example to help me understand what you're getting at.

I'm currently leaning away from the idea of a total table rake cap and leading toward the idea of a per-round rake cap.

Also, as some empirical data on the current rake:

$52.28 was the total buy-in so far today over 47 games.  The total rake has been $2.65.  Thus, even though the rake is 10%, the fact that it is rounded down and only applies to the pot makes it come out to a 5% tax on the buy-in.

This is EXACTLY the kind of feedback I'm looking for.

Yeah, I left the old "mildercaution" folder there by accident today (that was the old secret code word for my very early testers), and a few people typed "milder" and "caution" into the box when I tweeted about the DOOR today.  I had also created the new "originensemble" forum without thinking that the old key still worked.  Three people guessed the old key and go through before I closed it.  Doing it that way is even BETTER than being invited, so welcome! 

The rake is a tough one, because I'm trying to keep it super simple.  I agree that it's way too high right now.

What I want to avoid is raking fractions of a chip.  Right now, it rakes one chip if the pot is at least 10 chips, rakes 2 if it's 20, and so on.... but 20 chips if both players push.  Obviously, when both players push the pot way up, the rake really adds up after several rounds.

So, let's talk about caps in terms of whole chips.  Maybe it stays as 10% (never raking at all for pots that are 9 or fewer chips), but caps the rake at X chips.  What is the ideal X here?

How much back-and-forth do we want to allow on one-buy-in before the rake starts taking a huge chunk?  That could help set X.  What is a huge, noticeable chunk?  Like, if you slowly win everything over time and walk away with 180 chips, that seems okay, right?  Last night, I watched a few games in person where there were only 130 chips left at the end... OUCH.  So, the per-round cap could be 2 chips, which would allow the table to live for 10 rounds before the rake started getting too painful.  I guess I'd set it to 4 at first just to be safe and the lower it later if 4 feels too big.

Another point is that the house should never take more than the winner, even over the long run.  So, this could be handled with a global rake cap for the life of the table, where the rake would stop after it hit 20% of the table's chips (and then the players would play for free after that as long as the table lived).  This is interesting, though it feels a big more complicated and less visible (players see the rake stop mysteriously after a point).

As far as re-buys go... I intentionally avoided that for a few reasons.

--It complicates the game (when are they allowed to rebuy, how long do they have to do it, etc.)

--It makes each table less clear cut (we don't play until one is out, there are more that 200 chips at the table, it could go on forever)

--It encourages a kind of predatory and exploitative environment where you can really milk someone who is tilting (or just weak) for many buy-ins, encourages weak players to throw good money after bad---in general, it just feels more like "problem gambling" to me.  If the rebuy is capped, the other player always has chip advantage anyway, so you're buying back into a disadvantaged situation on many levels.  I'd say you only do that when you're tilting.  The game tells you to walk away and start a new game with a clean slate instead.

The reveal step at the end was added after the Kotaku article.  I felt like the final betting round was too much in the dark without it.  Without it, your opponent has 6 possible scores that are often in a huge range.  I had the feeling that in Holdem, you actually have MORE information by the end about what your opponent might have when making your final bet (you can see three of their cards, after all).

So, the reveal gives one more betting round, which is good (same number as Holdem).

But post-reveal, there are only four possible situations (ignoring ties):
--We both know who won
--One of us knows and the other doesn't (and both know who knows)
--Neither of us know.
--One of us knows and the other doesn't (and the non-knower isn't sure about whether the knower knows).

The last two cases look the same on the score graph.  (Actually, this is a bit more nuanced than I thought it was... I'll have to get out some paper and work out the possibilities later.)

I was also hoping to have an interesting choice with the reveal at the end, but in practice, it is usually a forced choice.  At least it's a much less interesting choice than the choices so far.

So, anyway, you're making one last bet, but the possibility space is pretty narrow at that point, so it's often pretty anticlimactic.  I'm not totally happy with it, but I also haven't worked out any alternatives.  The game without the reveal step felt too much like a shot in the dark, even at the final bet.

I sent out emails to something like 160 testers tonight.  Hopefully, many more of them will get the message tomorrow and take a look.  There are only 2 active players at the moment, and I myself am heading to bed.

Yes, that's the idea.

Of course, this software has not been widely tested and could have bugs (this is the first larger-scale test beyond a small group of my personal friends).  Those bugs could cause people to lose money or who knows what.

SO... please make sure that you explain this to your friends.  "This is a new game that we're helping to test."