The Tao of Gaming

I don't know if I've linked to Overcoming Bias before or not. It doesn't deal with gaming, but is interesting and sometimes deals with tangential items of interest (especially to game theory).

Today they are talking about Newcomb's Paradox (which I first encountered in a math class in middle-school. Thanks, Martin.)

And the following jumped out:

Nonetheless, I would like to present some of my motivations on Newcomb's Problem - the reasons I felt impelled to seek a new theory - because they illustrate my source-attitudes toward rationality. Even if I can't present the theory that these motivations motivate...

First, foremost, fundamentally, above all else:

Rational agents should WIN.

Don't mistake me, and think that I'm talking about the Hollywood Rationality stereotype that rationalists should be selfish or shortsighted. If your utility function has a term in it for others, then win their happiness. If your utility function has a term in it for a million years hence, then win the eon.

But at any rate, WIN. Don't lose reasonably, WIN.

{I'm adding them to the blogroll and cleaning up some old URLs).

Learning how to win games that you are losing has been in the forefront of my mind. Reading Wei-Hwa's latest comment, I realized that my thoughts aren't likely to evolve much in the near future, so I'd just throw this out.

These thoughts are definitely provisional.

The first step to come back from a losing game is to accept that you are losing. If you are losing, you are behind the curve (either on income or VPs, depending on the stage of the game). So, if you and your opponent stay on the same curve you can't win. You'll have to get lucky. At some point, you'll have to take a risk.

For example, if your natural play would be to settle a pretty good world, perhaps you can risk someone else calling settle and exploring for a better world. I imagine that most of the plays are variants on that theme.

• Calling your 'perfect' role hoping that your opponent(s) call another role you need.
• Exploring +1 instead of +5 for that one card, hoping to bag your card and a spare.
• Dumping your hand on an expensive card (instead of playing to keep your next obvious play) and hoping that your next big income will allow you to continue on your path.

As Wei-Hwa says, the "leader" should be minimizing exposure to luck. Just select your 'good' role, make the steady play.

When taken together, that means that the leader should be more predictable then the trailing player. [This assumes that the leader has set themselves along a curve and isn't staring at a sudden dip ... like a hand of useless cards].

As a trailing player, you should also look at ways to vary the timing. If you are trailing with a larger tableau, then you have to concentrate on making your builds count more (having a nice 'density') or turning on your consumption engine ASAP. The nice aspect is that you can end the game, or let it go on, according to how the VP are working out.

Standard thoughts about losing apply -- if you are only slightly behind then you may just need to hope for a bit of luck. Would one perfect card salvage your situation? If so, then perhaps an explore +5. If you are grossly behind, then you'll need to swing for the fences.

If this all sounds vague; well, to a certain extent it is. By the time you can really ascertain losing, there are enough cards down to make it a case by case basis.

I'm sure that I am forgetting good pieces of general advice. Perhaps I don't know them yet.

A mere two weeks after getting to Kaiju on Hard, I have now beaten my first song on that level. I rock and/or roll. Now come the harder songs in the set. (Granted, I'm only playing every 2nd or 3rd day, but sheesh).

I just (today) noticed Keldon's Blue Moon program w/AI (via this thread on BGG). I've just tried a game or two against the computer opponent, and it seems credible. (I'm no expert). Clearly if Neural Nets can handle Blue Moon, they should be able to handle Race...

My one game of Race was interesting enough. (Oh, Linnaeus posted his thoughts about Race. And Mikko posted his review). But (amazingly), games were played.

New to me was Escalation! -- a Knizia game that I guess is in the climbing family. I suppose. You have a hand of six cards, which range 1-14 and a few specials. One player leads, and then each player must meld a higher total. A mild is one (or more) cards of the same value, and the total is just that. When one player can't meld, they take all the cards. Each card taken is one point, points are bad. After playing you refill, and once the deck runs out you play until one player empties his hand ... everyone else drops their hand into cards taken.

Well ... it's fast. Not bad, actually (perhaps better with fewer players?). But this felt slapped together. Basic idea, a few card specials, and go. My "some cards way better than others" pet peeve was off-the-chart. I mentally rebuilt the deck to narrow the range of card powers in five minutes, I'm sure Knizia could have done it if he wanted to. And, lets face it, the "Card game where we keep playing until the last gamer arrives" genre is just stacked like Pittsburgh's steel curtain.

Got in another Power Grid game with the expansion deck. I like it.

Two more games of Kutschfahrt included one game nearly deadlock. Several players each discovered an opponent, but made a supporting play (to throw said opponent off-kilter). Well, it worked as everyone kept saying things like "At least two of you are idiots." Deception can work, but it can backfire. I think the game works a bit better if everyone plays it straight. I would be very interested to see how a group that played this seriously evolved ... eventually you'd get to a point where people would end the game when they were 'almost' certain they were right (because waiting for certainty risks your opponents ending the game). Eventually you'd see people end the game when they had a reasonable shot (say 2/3rds). A winning strategy would be to make any claim that had a greater than 50% shot. Would the game be interesting at this point? (As a game, and not a theoretic device)? This would make an interesting project -- write computer players and evolve them. (I nominate ya'll).

Celebrities is still a good game.

(Is it just me, or is the Geek really slow these last few days?)

So, uh, does anyone have any Pokemon decks they want to trade (cheaply?). Reply via email (or geekmail) to hide your shame.

Update: To reclaim any lost 'cred' I had, I made this geeklist.