Thursday 30 January 2014

Spoilers, Do Not Read

I think there are a couple of things in Sandcastle Builder that should be better explained to the player.  Perhaps through the use of the Free Advice boost.  If you've been playing for a month and feel like your game is at a standstill, here is a cheat sheet.

Wednesday 29 January 2014

So... Skyrim

It may be sufficient to simply say that I downloaded a mod to make a character marriable and then decided to not use that mod because the marriage dialog wasn't in character enough.

Video games that give you choices usually don't give you many choices.  When you sit across the table from people and play a roleplaying game you have a very wide variety of options, but video games are necessarily limited by what the designers imagined for you.  Sometimes a physics engine can allow novel solutions, but one area where you really don't find novel solutions is dialog.

So while Skyrim does often present me with choices that I don't like that much, I've actually been very impressed with how well they've allowed me to feel like I am making real decisions despite the fact that I am choosing one of two preset sentences from a list.

At any rate, I found myself really playing a character that I developed as I went through the game.  I started a second game and I'm alternating between the two different playthroughs.  This lets me play different combat styles, but it also lets me make different choices in how to approach the game.  I might start a third game soon just to do a stealth game.

I ended up paying $36 for Skyrim, but since it looks like I'll probably literally play it for 300 hours, I feel like that's probably a reasonable deal.  Steam really encourages me to wait for massive sales, but sometimes it's just better to get good things.

Friday 24 January 2014

Hardmode

There are generally two modes I like to play games on, hard and easy.  It's fun to be put to the test and it's fun to stomp, but medium modes where you know you are going to win but you can lose if you don't pay a little bit of attention are irritating.

Well parenting doesn't come with an easy mode, but you can certainly ratchet the difficulty up.  Maybe I was just discontent with hard and had to go up a notch.

Tuesday 14 January 2014

Efficiency

People will always do things the most efficient way in games, even if it is not the most fun way.

That is a hard statement to argue with because we know that people do behave that way.  But it seems like the problem is overstated, that people generalize from the some to the many, or even to all.  The problem is that we prove the statement true using a small set of people and then use it as if it was true of everyone.  The ambiguity of natural language lets us get away with that.

Still, the idea that people will act against their own interests to earn rewards in video games seems pervasive in the drive towards "balance."  If sorcerers beat the dragon faster than wizards then everyone will play a sorcerer, if wizards are faster then everyone will play a wizard.  If mining makes more stuff than gathering then people will mine, if gathering makes more stuff then people with gather.  If actually playing an engaging game makes fictional currency faster than dutifully pressing 'Q' every 15 seconds then people will play the game, but if pressing 'Q' is more efficient then undoubtedly that is what people will do, even if the gain is only marginal.

I haven't seen any evidence that any of that is true.  A very small number of people will radically alter how they play for marginal gains, and a slightly larger number of people will complain that it isn't fair to them because the way they prefer to play is no longer the best way.  A very disproportionate amount of attention is paid to the people who are on top and to the people who feel like they should be on top but that the system is rigged against them when more attention should be paid to giving everyone who plays the game a wide range of choices so they can play the game their own way and have fun doing it.

Monday 13 January 2014

Sandcastle Builder - New Logicat Tricks

The logicat revamp came out and essentially invalidated my old posts about logicats.  Now you can just solve the puzzles!

But solving the puzzles is one thing, solving them fast is another.  Despite the fact that you can now get one correct answer per statement in the puzzle, and even despite the fact that you can "wager" more than one puzzle at once, you are still going to be solving a lot of puzzles.

So here are my logicat tips:

1. No Contradiction Means You Are Right
The algorithm generates puzzles with a single correct solution.  That means if there is no contradiction in the answers you have given then that is the one and only correct soluation.

2. Just Start Filling Things In
This is probably the most important tip.  Don't think things out a lot, just take a statement, mark it true or false and start going with what that implies.  Each time you assign a truth value to a statement verify that you haven't hit a contradiction.  If you have, you know the first statement you marked is the opposite of how you marked it.

3. Self-Referential Statements
X: ~X or Y means that X is true and so is Y.
X: ~X and Y means that X is false and so is Y.
These are either great starting points to the whole puzzle or the solution to sub-puzzles.

But what about X: X or Y and X: X and Y?  A not-uncommon puzzle you will receive is one where there is only one statement of one of those forms and all the other statements make only a single claim.  When you see a puzzle like this, you know immediately that if X is an "or" statement then Y false and if it is an "and" statement then Y is true.  How do you know that?

Suppose I have a puzzle that says:
X: X or x(Y)
Y: y(Z)
Z: z(W)
W: w(X)

Suppose X was true.  That would mean that either W is true and w(X) = X or W is false and w(X) = ~X.  At any rate, w() will determine what the truth value of W is.  We can similarly go back through Z and Y and get a truth value for Y and hence x(Y).

If x(Y) is true, then we have a problem.  You see, you could change x(Y) to be false which would reverse the truth value of Y and that would reverse Z and W as well.  So it would finally reverse X.  But the problem is that set of statements would contain no contradiction.  If X is false then x(Y) is false and all the other statements line up correctly.  Thus, the puzzle would have two solutions, which is not allowed.  It's pretty easy to reverse this argument for and-statements.

So, if you see a puzzle like the one above, mark Y false and follow the conclusions through to determine if X is true or false.

4. If You Get Something Wrong, Give Up
This is not necessarily good advice.  But when your Automata Control is high enough and you have Zookeeper, by the time you've failed to solve a puzzle and been given a second chance, you can just hit clear answers and then submit and move on to the next one.  If puzzles are at a premium then obviously this isn't a good idea.  Also, clearly, if you got everything wrong it is pretty easy to solve.  But if I got two of seven statements wrong why would I read the same puzzle over again looking for a mistake I made when my mind is already clouded by a proposed solution.  I find it much better to start fresh.

5. If You are Getting Things Wrong, Go to Bed
The best of us are not above a misclick, but if at some point it's not just a misclick.  Go to bed already.

Friday 10 January 2014

Mad Myth Friday - Contemporary Brain Dump

Reposted from the Boing Boing BBS.  The indented bits are quotes from other people.




Yeah, as I implied, it's really okay with me that this is incomprehensible and appears to be self-contradictory. If I knew words that could explain it other than simply repeating, "Yeah, conscious might be an illusion" then I would certainly use them. Somewhere between radical materialism and being mindful of "the emptiness of phenomena" one day I said, "Oh, hey, that makes sense."

The problem I have with what I'll call "Socrates epistemology" is that, as Socrates said, "I know one thing: that I know nothing." And that's a pretty stupid thing to say. I know at least one thing: that Socrates has defined the word "know" to have nothing to do with what people actually mean when they talk about knowing things. Because people know things.

And if people know things, but it is also the case that according to Socrates-knowing you can't know anything, then knowing isn't about Socrates-certainty or Socrates-proof. It's about actual certainty and actual proof which are actual things that are made up of particles and occupy space and have mass. Direct perception is one form of experimental evidence about the content of underlying reality.

What we experience is a thing, and not a private one. Other people can stick their hand into it and can walk through it by accident. Zizek asked how, out of a dumb, flat reality that just is, could something like perception come to be. I found his 900 page answer unsatisfying. To me, the simplest answer is, "Maybe that's a silly question." I compare it to the at-the-time reaction to wave-particle duality: This was regarded as a paradox because of the conceit that light had to be like something that was already understood. The notion that light sometimes acts like a wave and sometimes acts like a particle is a metaphor for the fact that light acts the way light acts. When we talk about the very base components of reality, they are going to be non-apprehendable. The best we can possibly do when talking about fundamental reality is the equivalent of learning about elephants by bouncing hippopotamuses off of them and then having a soothsayer read the hippo's intestines.

It's funny that in the same thread that spawned this discussion, Medievalist said:



I absolutely agree with this statement in an extremely blunt and non mystical way. It is a brute and boring fact (but not a Socrates-fact) that the universe has eyes and that two of those eyes happen to be in the volume of space (though I'm not sold on the existence of space) that I call myself for social purposes.

I'm not sure what any of this has to do with the illusion of consciousness except to say that I always assume that however wrong I think I might be about anything, I might be even more wrong than that. No matter what word games I can play to convince myself that is not true, when it comes to knowing, it seems more correct to me to assume that the bottom might yet fall out.

So in one way, consciousness may be an illusion in that when I say, "I hear a ringing in my ear" I might be terribly, desperately wrong about what I am, what it is to hear, what a ringing is, what my ear is, and pretty much everything else to the point that my experience of hearing a ringing in my ear is a ridiculous fraud. I understand that that doesn't address what people like to think of as the immediate experience of things, but it might. The Dalai Lama said that no brain scan will ever apprehend the actual experience of seeing blue, but I'm not sure that this isn't just a barrier we have constructed for ourselves in our thinking.

Having to believe that we exist so make sense of everything else means we need an anchor. Dennett compared it to a boat driving in circles to identify itself on a radar map. We need a process that says, "This is me" if we are going to recognize "not me." But anti-evolutionists thought that the eye was too complex to ever have evolved and then people realized it evolved completely separately in seven different hereditary lines (the fact that this probably isn't true isn't important here). It's possible that evolving eyes isn't just possible but rather nearly inevitable.

Phlogiston was an attempt to explain heat. They were wrong, but they were still trying to talk about the same thing we talk about today when we talk about heat. We've been grasping at that straw for a long time and we have an understanding of it now that is a lot better than Phlogiston. What's more, we can recognize that our current understanding of heat is what we meant when we said 'heat' even though sometimes the definition doesn't match our intuitions that originally led us to talk about something we called heat.

We talk about something we call experience and obviously we've been grasping at that straw for a long time as well. When we figure out what it is, it may be that our intuition about it does not match the reality. It may be that we'll be able to say, "Hey, I actually didn't experience a ringing in my ear at all." Even that doesn't answer the real question because we aren't talking about the concept that we are grasping at when we speak of experience but instead of the experience we are having now which may or may not be included in that concept.

But there is no bottom.

Thursday 9 January 2014

Skyrim

Over the holidays Skyrim was on sale for $36.  I had planned on waiting for it to be like $10 but I don't exactly know why so I snapped and bought it.  I think the clincher was when I heard some guys on the bus talking:

"So what are you doing on Christmas?"

"I'm just going to stay home and play Skyrim."

This being two full years after the release, I was impressed.

Skyrim has a lot of stuff to do in it.  So much stuff that it feels like your quest log just keeps getting bigger and bigger.  I recall reading when it came out that it was much better than an MMO at giving you the feeling of a world inhabited by other people.  That's true if you want from other people is to play along with a grand story.  If what you want from other people is to make childish jokes in some kind of global chat.

It makes me really not interested in the Elder Scrolls Online that is coming out this year.  Why would I want to play an Elder Scrolls game with other people when I can play a perfectly good one without them?

Skyrim is a good.

Monday 6 January 2014

Awesome Games Done Quick 2014

This whole week is the Awesome Games Done Quick 2014 marathon hosted by Speed Demos Archive.  You can tune in and watch at gamesdonequick.com.

If you aren't familiar with speed running, it is a real treat to watch.  You'll see games that you played as a child destroyed by players who have mastered them to the extent that they appear to be psychic ninjas.  There are often lots of neat tricks from clever little moves to save a few seconds, to outrageous glitches that skip whole levels.

Last night, at the end of the first day, they had some tool assisted runs.  These runs were played on actual consoles by a robot programmed with a set of moves.  The runs included a Mario Kart game where each level was won in about 20 or 30 seconds by bouncing off of objects and falling on holes to trick the game into thinking you had crossed the finish line - neat stuff like that.

But most impressively there was a game of Super Mario World where they somehow, by sending inputs through all eight controller ports, modified the table where the game stores objects to program a playable Pong and Snake into the SNES's memory.  Watching the run I was completely baffled, it was so far beyond what should be possible.

The marathon goes until January 11 and is in support for the Prevent Cancer Foundation which may or may not be a great cancer charity in the vast sea of cancer charities.  I have a fair bit of faith in the guys that run the marathon that they picked a charity that is pretty good.  There are various incentives to donate towards, and if you type in a message they read it online at a slow part of whatever game they are playing.

You can check the schedule to see if there are games you are interested in, but I find that even games I don't know anything about can be fun to watch just because the level of play is so entertaining to watch.  There is a good chance someone other than me likes this, and who knows, you might be that someone.  Check it out.

Friday 3 January 2014

Less Reading, More Clicking

At some point, answering a logicat question becomes something you can do whenever you like.  There is only so fast you can move the mouse to click those buttons, especially given that the buttons aren't alway totally responsive, so when your caged logicat gets 18 extra questions every 8 mNP you can't actually run them out.  When you are at that point, you have to seriously reconsider the worth of getting the answers right.

Answering a question wrong costs you.  At a base getting a question right earns you 0.2 levels and getting one wrong costs you 0.1 levels.  As you scale up your getting-it-right value, your getting-it-wrong value scales up by the exact same amount.  A right-wrong pair is always worth 0.1 levels, regardless of how much a right question gets you.  I currently get 1.9 levels per correct answer.  So if I get a question right I get 1.9 levels and if I get one wrong I lose 1.8 levels.

Suppose instead of reading the question I just click the first answer and then the second if the first was wrong.  If I have two chances at a 50-50 guess then I have a 75% chance of getting the right answer.  But I don't have two chances at even odds.  Actually the numbers are more favourable.

First of all, there is no replacement.  If there are four responses, two of which are true and two of which are false, I have a 50% chance of getting it right immediately and if that fails I have a two in three chance of getting it right since I presumably will not choose the same incorrect answer.  That gives me a five in six chance of getting it overall.  If there are eight statements then my chance of getting it right is 50% followed by four out of seven, which is nearly 79%.

But it gets better than that too.  The game does something to gaurantee that you aren't given an impossible puzzle.  Given that you can have a 100% chance to get it right but you can't have a 0% chance, we might immediately suspect that as guess is better than 50%.  The mechanism for ensuring this determines how much more.

Rather than simply choosing whether you need to pick a true or a false response, it chooses a random response and assigns the truth values you need to pick to the truth value of that response.  So if there are two true and two false statements then there is an even chance of being asked to find a true statement versus a false one, and out first guess is a 50-50.  But if there are three true statements with only one false, then there is a 75% chance we are asked to find a true statement and a 25% chance being asked to find the false one.  What's more, if we are asked to find the true one we will always get the right answer because we have two guesses, only one of which can be wasted on the false statement.

In the end, the odds of guessing the right answer for a four question puzzle is actually 87.5%.  An eight question puzzle is worse, but still 81.25%..  Overall, guessing gives us very close to an 84% chance of a right answer.  The average value for a guess for me, then, is 1.3 levels.

Now I haven't actually timed myself answering questions but the idea that reading enough of a question to get it right is worth those extra 0.6 levels seems pretty far fetched.  I think I got pretty quick at answering questions but blind clicking has to be more than twice as fast.  Correct answers are more favourable at higher logicat levels, but the ratio between guessing and getting it right approaches an asymptote above two-thirds, so there is just no way that reading questions will ever be right when unlimited questions are available to be answered.

As an update, I plan to have infinite glass by the end of the weekend and I've already unlocked two things to spend it one, one of which appears to seriously invalidate the other.  Then again, I don't yet know the mechanism for generating another resource I will need, so the one might be useful in the short term.