The Tao of Gaming

Tom just left a comment that he has written a quite long response about Goa dealing with luck, auctions, and the value of different paths. It's well worth a look. (Scroll down).

A local gamer strongly resisted Air Baron, claiming it took forever. But we played 6 players in the typical 90 minutes. (w/ advanced rules). Most players followed what I call the 'typical' strategy.

1. Get a single hub in several spokes (to earn the $3 bonus when that spoke is drawn). Usually get ones that are valuable (to have more of your money invested), but save a few bucks for next turn.
2. Once you've got a reasonable base, try to control a hub. Some prepatory moves (jumbos, foreign spokes) may help if you plan on doing this via a fare war.
3. Now that you have market share (and income) start attacking adjacent hubs.

Let's call this strategy "Expansive" since it spreads out quickly.

The old group would start attacking players who split up in phase 1. (They call it wimpy play). So they attack! I'm not sure on the details ... but this strikes me as a non-stable strategy, on the other hand, it poisons the environment.

If we assume five opponents who will attack people who split up ("Punishers") then an Expansive player is doomed. He expands, and then punishers attack. Even if they fail often, and cost themselves dearly, odds are that one of them will knock me out. And while they incur a large cost (paying twice as much for ~60% shot at kicking out the expander), the expander losses the whole investment as well. So in a spoke valued Y, several punishers lose 2Y (paying and failing), one looses 2Y (paying and succeeding) but has potential future income. And the expander loses all that future income. [I'm making a lot of simplifications].

Now, this clearly does suck the amount of money out of the system. In a game with all punishers, they'll start grouped out (for safety and too avoid triggering punishment) and so will earn money at a slower rate, and increase market share quickly. That will toss in the 'event' chits into the bag, and those suck money out of the game too. 9 Hours seems excessive, but I could imagine three.

But put one "Mellow Defender" into the mix. Mellow won't provoke an attack (like the expander), but won't attack him either (unless it's genuinely a good play). In essence, he'll let the punishers deal with him. Given that the punishers overpay (to punish), he should win more than his fair share. Other players notice, and switch from Punishers to Mellow. If there's only a single punisher, he'll rarely win.

This doesn't help the poor expanders, who were driven extinct (in that play group) long ago. Ah, group think and evolution. To complicate things, Punishers may retaliate against the 'parasitic' mellow player. Evolution worries about free-riders.

[Sick days -- when web content gets written!]

Multiple plays:

• Leonardo da Vinci (3)
• Air Baron (2)
• Ave Caesar (2)
• Stage II (2)

And the rest: 6 Nimmt!, Beyond Balderdash, Boomtown, Celebrities, Die Macher, Fast Food Franchise, Flaschenteufel , Frank's Zoo, Fussball-Taktik, Get the Goods, Goa, Igel Ärgern, Indonesia, Ra, Samurai, San Juan, Scepter of Zavandor, Schoko & Co., Smarty Party, Struggle of Empires.

Before I played lots of games, I played Chess. Through intense effort, I achieved ... tournament mediocrity. But I studied the game. I once spent dozens of hours (over the course of several weeks) looking into a one variant of the Sicilian Poisoned Pawn. In this variant, white checks with a bishop that can be captured by a pawn. If the bishop is taken, white gets a powerful attack. Quite complex. Beyond my capabilities to prove, in any case. But I knew a fair amount of details.

A few months later in a tournament I sprung this on an expert. After playing my move, I got a drink. He sat and stared. And stared. Then he stared some more.

When people noticed that the small-fish had unleashed a TN, lots of players came over to check it out. My opponent took about 75 minutes to make his next move. He declined the sacrifice. After another ten moves we were in a dead drawn position, and the expert had 5 minutes left to make twenty moves.

So I played quickly (to keep him from thinking on my time) and missed a mate in one. Oops.

Why did I think of this just now? Why the recent match between world champion Kramnik and Deep Fritz.

After smoothly gaining the advantage Kramnik continued to play quickly, perhaps a little too quickly....

Kramnik played the move 34...Qe3 calmly, stood up, picked up his cup and was about to leave the stage to go to his rest room. At least one audio commentator also noticed nothing, while Fritz operator Mathias Feist kept glancing from the board to the screen and back, hardly able to believe that he had input the correct move. Fritz was displaying mate in one, and when Mathias executed it on the board Kramnik briefly grasped his forehead, took a seat to sign the score sheet and left for the press conference, which he dutifully attended.

One of the most amazing blunders ever. We've all been there.

34...Kg8 leads to a draw. In fact, the chessbase article has a screen capture of (Fritz's) analysis during the game where it's just waiting to force the draw after Kg8.

An interesting match, if only from a rules perspective. For fairness, while Fritz is "in book" (working with opening positions it knows and has studied) Kramnik is allowed to look at Fritz's book tree during the game. So he knows all the possible replies to each of his moves. [Once Fritz is out of book, the screen is turned so that only the operator can see it.] I haven't seen the rationale, but suspect its related to the practice of giving the machine every tournament game its opponent has played.

So here I am at home sick about to write up my thoughts about my last Goa game when I see an article at BGG discussing it. So I wrote up my response there.

We played The One last night, and I'm reminded of the same thing I've felt every time I've played in the 2000s (or so) — Flawed classic. [We were playing 2nd edition; I don't believe the new version changed]. While second edition fixed some aspects of the game, it broke others.

I like Die Macher, I rate it a seven. But each game just convinces me that I should play with more variants. Flaws, in no particular order

• The positive feedback money system, while reflective of real life (at least in the US), doesn't help.
• The auction inflation in the game indicates to me that other prices should perhaps inflate as well.
• Capping votes at 50, and the weird player order effects it causes.
• The platform deal and change system that may horribly imbalance someone off the deal.
• No bonus for consistency in party platform.
• The luck in setup, where some people may match quite a bit (with regions and other players).
• The huge luck swings based on a single opinion poll.
• Winning the first election gives you initial money, ups your party-base since you'll usually match two positions, and will match them after the 2nd and 3rd rounds, giving you three turns of growth. And your media marker bonus is high.
• Winning the last election lets you flip flop national opinion, a possible 50 point swing.
• In contrast, winning the middle elections usually don't carry much weight, money is less valuable and your positions can be reversed.

I'm probably forgetting a few things, too. I've seen (and proposed) lots of variants to deal with these, but here's a formal proposal. You can probably mix and match these to a fair amount. Many of these variants have been proposed before. Phases not mentioned are unchanged.

Setup:

[I'd certainly argue for this setup even if not playing anything else].

1. Deal out seven states at random. [All seven elections are known at the beginning].
2. Order them by size, then re-arrange them by mandate so that the smallest is 1st, next smallest is last, next is 2nd, then 6th, 3rd, 5th and the largest at forth. This means that opening elections, while important from a snowball effect, don't actually have much mandate. And the last elections will also have a big effect, but small mandate. [The 3rd, 4th and 5th elections will usually have the bulk of the mandate ... creating a Super Tuesday effect. This may be too US-centric, but there you go].
3. Take one of each party platform card and deal one to each player (or let them pick). No two parties should have opposite platforms. After each party gets a core platform, take the deck of remainders and shuffle in two more sets of platforms, and deal out the rest normally (reshuffling as needed). [This prevents 4-1 splits on the deal]. If you are playing with the platform variant below then set the cards face up, if not then deal out hands normally.
4. For the people's opinion (cards that indicate what regions believe) take one of each and use them to seed the starting regions (1st election with 4, 2nd with 3, 3rd with 2, fourth with 1). If you get a opposites in the same region, reshuffle the conflicting one back in the deck. Take the remainder of the cards and shuffle them with a second set of opinions to seed the face down cards. [This keeps one opinion from being over-represented]. After that shuffle all the cards and flip up six to form the opinions that can be swapped out.
5. Conduct the rest of setup as per the rules (including the "chinese menu".
6. Each player secretly (on the back of their form) indicates which of their shadow cabinet is their "workhorse" (see below). They may choose not to have one, but should indicate that with "none."

Party Platform Phase:

All the unused cards are kept face up. If a player switches, he simply swaps one face up card for another, and pays the cost in party base:

1. Reversing a platform costs 5 party base.
2. Switching a platform costs 2 party base.
3. Changing a core platform costs double (10 for reversing, four otherwise)! After changing, the party picks any new platform as their core.

If any party takes position that uses up all of those positions, then the party that is 'locked out' gains party base. This represents the single issue voters flocking to them. The 'odd man out' gains 2 party base immediately, or 5 if they are explicitly against that position. Example, Alex, Brian and Chad are "Pro-Education Reform". Dennis takes the last "Pro-Education Reform" platform (paying party base as above). Edward gets five party base if he is Anti-Reform, or two if he is merely ambivalent).

Shadow Cabinet Phase:

As noted in setup, each player chooses one Cabinet member as a workhorse (or none). That cabinet member, when used costs double and is revealed when costs are paid. During that player's selection, the Cabinet card gets two actions. Both actions must be different. If a player cannot (or chooses not to) use both actions, the card still costs double. Instead of a second action, a workhorse may choose to not place a Coalition tile (if he should). [A workhorse cannot choose to add a coalition tile].

[Rationale — I liked the old system of having to discard two S-C cards at the beginning, it allowed players to differentiate. I choose to add a card because it tightens the money supply instead, and it lets each player "double up" in a region once, or avoid a coalition.]

Place Workers Phase:

More than four workers can be placed in a region. Each worker beyond the forth costs $4,000 instead of $1,000.

Opinion Polls Phase:

Polls are auctioned from 'furthest' region, concluding with the current region. Pick one variant:

• The winner may purchase multiple polls (paying auction price for each), but must decide before looking. Only one poll may be published (or dice rolled). Extra polls are discarded.
• The current poll is revealed before auctioning.
• Players may not move up/down more than one space in the electing region, and two spaces in the next region.

Converting Workers/Elections Phase:

Votes may go above fifty. Each vote above fifty does not increase mandate. Resolve votes normally, using the 'short head' version to break ties. If the winner (or winning Coalition) required the use of "votes above 50" to win the coalition, then they are considered to win a short head victory.

Example: The CDU & Green party have a coalition with 58 and 30 votes respectively. The SPD and PDS have a coalition with 27 and 55. CDU/Green win 88 to 82. However, without the 'extra effort', CDU/Green would only have 80 votes, so they are considered to have won a short head. [That SPD/PDS extra effort was used doesn't matter.]

National Opinion:

If an opinion is discarded, its space is emptied, then all opinions are shifted left. The new opinion enters in the right hand side (and shifts left as far as possible).

Example: The current opinions are pro-Euro, Anti-Genetik, and Pro-Healthcare reform, the other two slots are empty. The Pro-Genetik opinion is added. First, the Anti-Genetik opinion is discarded, then the Healthcare slides into the 2nd slot, then the Anti-Genetik card is placed in the fifth slot and 'slides' into the 3rd slot.

6th/7th Election:

The last turn resolves the 6th and 7th elections, there is no quick turn for the 7th election. [I've often thought the rules are unclear in 2nd edition, so this may just be a clarification].

Commentary

I deal with money indirectly ... the new Shadow Cabinet power, the ability to place extra workers, and removing party base for platform switch should tighten money. Also, the fact that some elections may go above fifty lets the players trailing in that election force the leader to expend extra effort.

Die Macher has too many moving parts to think any variant will be perfect. But I think this will help prevent a runaway leader. I can envision situations where these variants do protect the leader, but I suspect they will be rarer.