Recent Comments

Tuesday, January 24, 2012

Brian

Without GF, the probability before any games are played of two losses is 25%. After the first loss, 50%. Assuming coin-toss-matched teams, blah blah. If you're taking the first game as already played and lost, then you're right. I wasn't, because you said "Belichick doesn't tend to lose twice to the same team", which isn't exactly the same in my head as "Belichick doesn't tend to lose a second time to a team that beats him", but that's pure semantics and I think we understand each other. You're saying that even if game 1 is a cointoss and the Pats lose, they get a bonus to the odds of game 2 due to BB being a smart coach and game 2 is therefore not a cointoss.

I want Bill James to tell me if you're right.

Posted at 9:20 PM in response to Super Bowl analysis

Dave

I guess I misunderstood. Bringing in the 25% probability of two losses is GF, since the first game is already in the books. Maintaining that there is some distortion in the probability when a second game is played is not, as long as it's based on some actual effect.

Posted at 12:05 PM in response to Super Bowl analysis

Brian

See, I'd say then that it's not GF -- that teams split a 2 game series on the season more often than would be expected by chance modified for relative skill.

I want more sabermetrics in football. I want someone to tell me whether this impression of BB is accurate or not without doing the work myself. :)

Posted at 8:18 AM in response to Super Bowl analysis

Dave

Gambler's fallacy. The first game is in the books, done. The probability of winning or losing the second game if the teams were evenly matched and there were no other factors would be 50%, regardless of the first outcome.

A good coach can gameplan for that second game, though, based on the outcome of the first game. I have tremendous faith in Belichick's ability to do this. I think the probability of a Pats win would be greater than 50%, regardless of how evenly-matched the teams are.

Posted at 8:05 AM in response to Super Bowl analysis

Monday, January 23, 2012

Brian

(a) Belichick doesn't tend to lose twice to the same team

I've been hearing variations on this (not just for BB, for other coaches) a lot this season. What does it ACTUALLY mean? If, hypothetically, the Pats and Giants are very well matched (50/50 per game), you'd expect the Pats to lose both games only 25% of the time. Do you mean that the probability is significantly less than that, or is there a little bit of Gambler's Fallacy going on?

Posted at 9:25 PM in response to Super Bowl analysis

Wednesday, January 18, 2012

Dave

I think there's a sort of Laffer Curve¹ effect here with piracy. At most reasonable price points, yes, the people who pirate are people who would not have bought anyway. However, are they people who would have bought at a lower price point, or people who would not have bought at all?

As the retail price of a product increases, the number of people who choose piracy over purchase necessarily also increases, to a point where there is a diminishing return on profit. Content providers have two options: spend money to prevent privacy (through DRM, copy protection, or lobbying for legislation) or reduce their price to the new equilibrium point that maximizes profit. The second option is the realistic one. The first is a losing battle. It doesn't matter that the piracy is illegal or immoral - it's a feature of the current market and is largely impossible to stop, especially at a cost comparable to the profit that could reasonably be recouped.

¹ To introduce another hobby horse of libertarian think tanks like Cato.

Posted at 12:09 PM in response to SOPA/PIPA

Brian

In the second model, if people take your product after the fact for zero dollars, it doesn't hurt you.

What you don't note here, and what is one of the ROOT fallacies at most guesses at economic losses to piracy, is that you ALSO don't get hurt by this in the first model when your product is taken by someone who wouldn't have otherwise paid for it. While they are most certainly doing something grossly unethical, they aren't harming you directly.

The "cost of piracy" is distorted in other ways (http://www.cato-at-liberty.org/how-copyright-industries-con-congress/), of course, but that's attempted delusion of others. It's common self-delusion that each (or maybe even most) instance of piracy equates to a lost sale.

Anyway. I'm not arguing against anything you said.

Posted at 10:53 AM in response to SOPA/PIPA

Tuesday, January 17, 2012

Dave

I've got it on my computer. You're welcome to read it when I'm at the store. I've also got Dungeon World (the D&D-style mod/add-on) and while it's slightly flawed (and still under development, really) it's still a pretty good game if you want a fantasy alternative.

Posted at 1:47 AM in response to What I learned from Apocalypse World

Yanni

The system does sound interesting... I'd love to take a look at it sometime.

Posted at 1:45 AM in response to What I learned from Apocalypse World

Tuesday, January 10, 2012

Dave

The interesting thing is that it's not immediately obvious to a new player that the correct strategy is getting as many victory points as soon as possible. Seeing the board, most people will be immediately reminded of Civilization or Masters of Orion, where the goal is to build up and beat/outclass your neighbors in the endgame.

That leads to a situation like we had, where the one or two players who realized the correct strategy accumulated victory points very fast and won really before there was any real conflict between players. Again, distant suns might have exacerbated that by slowing down player-player contact.

On the other hand, if all the players had gone for the imperial card right away, then the rate that the lead players approached 10 victory points would have been slower and there might have actually been more time for Civ-like play.

Posted at 2:04 PM in response to Evening Shorts

Brian

And yeah, the first player to pick usually "should" pick imperial, and the second "should" pick initiative to pick imperial next time. Someone getting Imperial "out of turn" has positioned themselves well to win, all else equal.

Posted at 1:34 PM in response to Evening Shorts

Brian

That's true. The Imperial Card acts as a clock on play, and if it is debuffed from being worth 2 points per turn, it's generally accepted that the goal score should be reduced to keep game lenght manageable.

That said, I've only played like 3 times and not in the last few years.

Posted at 1:33 PM in response to Evening Shorts

Dave

It was interesting - most of the length of play and the fact that many of us lost badly might have been due to us underestimating the necessity to take the Imperial card whenever possible. We were all building for a longer game but the correct strategy seems to go for victory points wherever you can (to the extent of picking Initiative so you can get Imperial on the next round).

Also, we did choose the "distant suns" option, which made the initial exploration phase much slower and more costly for some players.

Posted at 12:24 PM in response to Evening Shorts

Brian

TI3 doesn't have to take that long. With familiarity a 4 hour play is pretty reasonable.

There is some consensus that the Imperial strategy card is vastly overpowered and that the way it moves around the table drives more of the game than it should. Many variants exist on the 'geek, but I don't know whether there's a consensus fix.

Posted at 8:45 AM in response to Evening Shorts

Wednesday, December 14, 2011

Dave

There's also a story about a time in the '70s when a volcano erupted in Iceland. The lava was going to hit a major harbor, so the locals used a combination of barriers and water to divert the flow.

The thing was, they had to cross the lava flow in places in order to get the water they needed from the sea. So they ran across the cooler parts, melting their shoes but otherwise succeeding in diverting the flow.

Posted at 5:35 PM in response to Lava

Dave

SMBC is funny.

I'll add it to the rotation and see if it sticks.

Thanks!

Posted at 5:30 PM in response to Webcomics rotation: 2012 preview

Yanni

There is a rather famous lava flow in Hawaii that has footprints in it. It's been a couple decades since I had a history class in Hawaii, so I don't remember the name of it, but one tribe was fleeing from another and ran across a lava flow to get away. As long as rather severe damage to your feet is acceptable, and the lava has partially cooled it shouldn't be too impossible.

Posted at 1:24 PM in response to Lava

Brian

How about Saturday Morning Breakfast Cereal? The art is definitely neither sharp nor stick-figures, but it's squarely aimed in the same vicinity as XKCD in terms of content:

http://www.smbc-comics.com/

Posted at 10:46 AM in response to Webcomics rotation: 2012 preview

Dave

I'm a bit of an art snob. If it's not stick figures, it has to be really sharp. I don't think I could handle Spacetrawler.

Posted at 6:51 AM in response to Webcomics rotation: 2012 preview

Tuesday, December 13, 2011

Yanni

I like the HDMIs better.

Posted at 9:05 PM in response to I didn't watch the VGAs.