Why simultaneous action selection is NOT random
Thi Nguyen
United States
Salt Lake City
Utah
flag msg tools
badge
Avatar
mb
Recommend
41 
 Thumb up
2.00
 tip
 Hide
So BGG's Mr. Ekted, in his blog, argued that simultaneous action selection (SAS) is random:

http://ekted.blogspot.com/2006/02/simultaneous-action-select...

He writes:

He wrote: "Given a finite set of choices, there are usually a small number that are directly beneficial. Given that, the choices are somewhat predictable. This means that you should not always be predictable. A certain amount of reverse psychology and reverse-reverse psychology ensues.

With multiple choices and multiple players, the final choices basically become unpredictable. Although they are being made intelligently, they are statistically random."

I disagree. I think there's almost nothing about SAS that is inherently random (though, like any mechanism, it can be used in certain ways to create randomness).

And that a lot of seeming random SAS-centric games aren't really that random.

I don't want to quibble too much about what "random" means. I think it's more important the effect - Mr. Ekted says that SAS games are essentially light, and that that's where they belong. This means, I guess, that there's no skilled, intelligent way to navigate SAS choice situations.
Your Tags: Add tags
Popular Tags: [View All]
  • [+] Dice rolls
1. Board Game: O Zoo le Mio [Average Rating:6.63 Overall Rank:1460]
Thi Nguyen
United States
Salt Lake City
Utah
flag msg tools
badge
Avatar
mb
Later, in the comments, somebody asks if blind bidding is a form of SAS, and Mr. Ekted has to say, yeah. I think this is a good place to start.

Blind bidding is a valudational decision like any form of other bidding, with one piece of information removed: what your opponent's maximum bids are this round.

O Zoo Le Mio is a game of all blind bidding. Each person has their little zoo setup, and you lay out the next five tiles. You can evaluate relatively how important each of the tiles is to you and each of the tiles is to your opponent, and then make an offer that reflects that valuation. You don't want to underbid and miss economic opportunities or overbid and waste your money.

You're offering a valuation of the item. What's random about that?

What blind bidding keeps you from doing is automatically paying the *minimum* amount that would beat your opponent. So you have to estimate what that minimum amount is. So blind bidding forces you to have a little finesse in estimating valuations. It deprives you of running data of what your opponent would pay. But that doesn't create randomness - it's just a bit of information that your opponent has hidden from you, that you're required to estimate.

So this basic form of SAS doesn't seem to have much randomness at all.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Loading... | Locked Hide Show Unlock Lock Comment     View Previous {{limitCount(numprevitems_calculated,commentParams.showcount)}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
{{error.message}}
{{comment.error.message}}
    View More Comments {{limitCount(numnextitems_calculated,commentParams.showcount)}} / {{numnextitems_calculated}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
2. Board Game: Poker [Average Rating:6.70 Overall Rank:885]
Thi Nguyen
United States
Salt Lake City
Utah
flag msg tools
badge
Avatar
mb
Try this degenerate example: imagine a game where we unveil a set of addition problems, and everybody has 30 seconds to do as many as they can. They write down the answers and reveal at the same time.

(This is dumb, but it's not as different as you'd think from some of the crappier analytico-mathematical games I've played.)

It's Simultaneous Action Selection (where the action is a proposed answer to a math problem.) And the SAS does *nothing* to increase randomness. Each person is being tested solely on the skill of the operation. If you removed the SAS and had people declare their answers in a row, you'd have a bit more data you could use, which is that you could try to copy the answers of somebody who declared before you, who you thought was better at arithmetic.

But SAS just deprives you of that data and forces you to choose without it - it doesn't add randomness.

Here's a example that's more real. Imagine we made poker SAS. Make it the most analytical form of poker - hold 'em, probably. (I don't know, I like goofy pokers.)

Imagine the two main actions were done simultaneously. We all simultaneously selected whether we wanted to drop out, ante, or ante up, revealed, and if somebody anted up, then we'd have to do it again.

Again, this deprives the game only of the data of what the people to your left would have done in this round of bidding. All the other data - what they bid just before this, and the statistical probabilities with the cards on display - is still available.

I don't think this adds any randomness to the system. It actually de-psychologizes it a little, maybe. But mostly it just deprives people of that one little bit of data, changing the game a little.

 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Loading... | Locked Hide Show Unlock Lock Comment     View Previous {{limitCount(numprevitems_calculated,commentParams.showcount)}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
{{error.message}}
{{comment.error.message}}
    View More Comments {{limitCount(numnextitems_calculated,commentParams.showcount)}} / {{numnextitems_calculated}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
3. Board Game: Modern Art [Average Rating:7.35 Overall Rank:223]
Thi Nguyen
United States
Salt Lake City
Utah
flag msg tools
badge
Avatar
mb
Think about the moment in Modern Art when you're doing blind bidding. It's not like all the other bidding processes are totally information-driven and analytic, and blind-bidding is random. You're focused on the same data as before - what value the artists could have, what cards have been played this round, what cards are likely to be played. In other forms of bidding you get a little more running data from the other players about what they expect the values to turn out, which might let you guess at what cards they're holding, but the change isn't huge. The process is just as information-driven and calculative as the other biddings.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Loading... | Locked Hide Show Unlock Lock Comment     View Previous {{limitCount(numprevitems_calculated,commentParams.showcount)}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
{{error.message}}
{{comment.error.message}}
    View More Comments {{limitCount(numnextitems_calculated,commentParams.showcount)}} / {{numnextitems_calculated}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
4. Board Game: Hoity Toity [Average Rating:6.53 Overall Rank:1253]
Thi Nguyen
United States
Salt Lake City
Utah
flag msg tools
badge
Avatar
mb
"Awright, Thi, you're totally missing the point," I can hear everybody saying. Blind bidding is not really what it's about. It's about games like Adel Verp, aka Hoity Toity, where everybody has to pick an action at the same time, not just bid.

But why is it so different? Take the first action selection - it's information driven. If you keep track of what people have in their hands, then you know who needs which items. You could, if you kept track of it all, figure out who really needs what items. The value of various sets is calculable in each turn. So there's plenty of data around. What keeps the game light is that we don't try to remember who has what, but that's more about the hidden nature of the cards then the blind selection of where to go, or how much to pay.

"What about the detective, thief bit?" you might protest. "That's pretty random." Well, still plenty of data to go on - who has plenty of thieves to risk, who has a thief about to come out, who needs money, who needs points. But here's where the core case comes in - this is the situation where we need to outguess our opponent in a situation where our actions directly cancel each other. So is this totally random?
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Loading... | Locked Hide Show Unlock Lock Comment     View Previous {{limitCount(numprevitems_calculated,commentParams.showcount)}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
{{error.message}}
{{comment.error.message}}
    View More Comments {{limitCount(numnextitems_calculated,commentParams.showcount)}} / {{numnextitems_calculated}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
5. Board Game: Rock Paper Scissors [Average Rating:4.30 Overall Rank:15361]
Thi Nguyen
United States
Salt Lake City
Utah
flag msg tools
badge
Avatar
mb
So: consider Rock Paper Scissors. This is about as close to the pure non-data-driven, actions-cancelling idea of SAS that makes you think it's random, right?

And I'll agree that a single round of RPS is random. There's no data at all. But play a sequence of RPS, and there's data to go on - what the people played in the past. And that's actually a surprisingly large amount of data. Because people normally don't play RPS randomly - they try to outthink their opponent, and you can figure out what they're thinking.

Try playing RPS with the following principle: imagine that your opponent, once he loses, is going to try to beat what you played in the last round of RPS.

It's scary. I did this in high school, and starting winning 80-90% of RPS games. What's more, some people figure it out, but you can watch them and see the moment the light dawns, and adjust your play accordingly.

So you have two bits of data: what they played previously, and any psychological hints you get from their face and manner.

However, if somebody figures all this out, and starts playing completely randomly, the strategy falls apart. And RPS goes back to being totally random.

The lesson? SAS results in random games if *there's nothing in the game to discourage random selection*, and the players start playing randomly. Which they will, if you're figuring out that their attempt to impose order on the randomness is making them lose to you.

I did almost get beaten up once over RPS. Guy flipped his lid. Funny story.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Loading... | Locked Hide Show Unlock Lock Comment     View Previous {{limitCount(numprevitems_calculated,commentParams.showcount)}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
{{error.message}}
{{comment.error.message}}
    View More Comments {{limitCount(numnextitems_calculated,commentParams.showcount)}} / {{numnextitems_calculated}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
6. Board Game: Edel, Stein & Reich [Average Rating:6.93 Overall Rank:1294]
Thi Nguyen
United States
Salt Lake City
Utah
flag msg tools
badge
Avatar
mb
Which leads me to the king of non-random RPS-type games, Edel Stein Reich. I love this game. As an interesting factoid, in a group of 4, I win about 70% of these games. And the game is, at its core, RPS. So that's gotta count for something, right?

Here's the interesting thing. RPS gives identical value to each of the decisions. Edel Stein Reich gives differential value to the choices (and, interestingly, a slightly different differntial value for each person). People have started playing randomly before, AND THEY LOSE. The system punishes it, because there's valuational data to be had, and the system rewards people for properly calculating through that valuational data. So people start playing according to certain principles. And that's the end of randomness. If you can suss those principles out, you can win.

There's valuational data, and then there's psychological data. Certain people tend towards conservative play. Certain people are more drawn towards the more valuable items, while other people are more drawn towards collecting less valuable items en masse. When there's a high-value item, certain people are more drawn towards it, and certain people are more worried about everybody being drawn towards it and cancelling each other out and drawn away. What's more, you can read people's faces - some people have tells for when they're being clever, or when they're doing something they think is risky. It's all data under the bridge.

So the game becomes one of figuring out how people operate, and outmaneuvering them. These are probablistic measures, but that's enough to game on and win on. And that's not random, any more than a statistician making bank at the horse rases is random (and I've known more than one.)

 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Loading... | Locked Hide Show Unlock Lock Comment     View Previous {{limitCount(numprevitems_calculated,commentParams.showcount)}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
{{error.message}}
{{comment.error.message}}
    View More Comments {{limitCount(numnextitems_calculated,commentParams.showcount)}} / {{numnextitems_calculated}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
7. Board Game: Through the Desert [Average Rating:7.06 Overall Rank:440]
Thi Nguyen
United States
Salt Lake City
Utah
flag msg tools
badge
Avatar
mb
Through the Desert gets praised for being, as Knizia intended it to be, "like life." It's not just about analysis and number crunching, but about your risk tolerance, your desire for conflict, your desire for peace. And winning involves assessing other people's risk tolerance, conflict avoidance, etc., to predict their behaviour.

Edel Stein Reich is the same thing. There's a minimal valuational system - enough for peoples tendencies and principles of operation to come into play. And once that's there, figuring out what those principles of operation are lets you win.

It's a skill. It's a much more psychological skill, compared to the skills of analysis you need for Age of Steam or Puerto Rico or whatever, but it's a very definite skill.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Loading... | Locked Hide Show Unlock Lock Comment     View Previous {{limitCount(numprevitems_calculated,commentParams.showcount)}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
{{error.message}}
{{comment.error.message}}
    View More Comments {{limitCount(numnextitems_calculated,commentParams.showcount)}} / {{numnextitems_calculated}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
8. Board Game: Chess [Average Rating:7.09 Overall Rank:416] [Average Rating:7.09 Unranked]
Thi Nguyen
United States
Salt Lake City
Utah
flag msg tools
badge
Avatar
mb
So SAS yields hidden information, not randomness.

I mean, there is a sense in which SAS detracts from pure analytical calulability in the way that *any* form of hidden information does (like card draw in bridge) - it makes the game not a purely chess-like, open-information thing. It hides a bit of data in your decision making process.

But that data isn't random. It's created by the valuational system, as seen through your opponent's particular psychology. And there's a skill to getting at it. It's not a sure thing in any single round of play, but over multiple rounds of play that skill tends to prevail.

I've known at least one unbeatable Adel player. Like somebody said, Adel V. isn't a light chaotic game, it's a master's degree in psychology. It's just a very different skill from the one you need for, say, chess.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Loading... | Locked Hide Show Unlock Lock Comment     View Previous {{limitCount(numprevitems_calculated,commentParams.showcount)}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
{{error.message}}
{{comment.error.message}}
    View More Comments {{limitCount(numnextitems_calculated,commentParams.showcount)}} / {{numnextitems_calculated}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
9. Board Game: Fruit Bandits [Average Rating:5.66 Overall Rank:12066]
David Brain
United Kingdom
London, UK
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
I'd like to contribute this little card game to the discussion. It's not quite RPS because the stakes are variable. It's not really blind-bidding because of the way the items may or may not be available depending upon each player's choice. It's certainly not random guessing as the extra decision element of "attempt to score yourself vs attempt to score off other players" means that the random player will lose out far too often.

I think the reason I like it is because it's got that fascinating mix of psychology vs analysis; sometimes you simply have to take the percentage play based on circumstances but sometimes you have to try and discern how other people will behave. And the game lasts just long enough for each of those aspects to make an appearance but not long enough for them to become overwhelming.

I guess what I'm saying is that in an SAS game, you need both of those elements present much of the time, but that it shouldn't be automatically obvious when you should be taking which line.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Loading... | Locked Hide Show Unlock Lock Comment     View Previous {{limitCount(numprevitems_calculated,commentParams.showcount)}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
{{error.message}}
{{comment.error.message}}
    View More Comments {{limitCount(numnextitems_calculated,commentParams.showcount)}} / {{numnextitems_calculated}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
10. Board Game: Rommel in the Desert [Average Rating:7.54 Overall Rank:1078]
Thi Nguyen
United States
Salt Lake City
Utah
flag msg tools
badge
Avatar
mb
The comment below about wargames and strategy got me thinking: you know what's SAS? Leadership in war. The earlier the war, the more you have to commit in SAS. Napoleon sending out messengers and Wellington sending out messengers is a form of SAS. Then I thought - wouldn't it be neat if wargames modelled this? And then I realized: they do.

Take Rommel in the Desert - here you have to simultaneously commit resources at the beginning of the game. It's not SAS in the sense of, say, RoboRally, because you don't have to specify what actions you're taking - you're just committing resources towards how many actions you can take. But it's a simultaneous commitment.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Loading... | Locked Hide Show Unlock Lock Comment     View Previous {{limitCount(numprevitems_calculated,commentParams.showcount)}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
{{error.message}}
{{comment.error.message}}
    View More Comments {{limitCount(numnextitems_calculated,commentParams.showcount)}} / {{numnextitems_calculated}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
11. Board Game: Hammer of the Scots [Average Rating:7.56 Overall Rank:318]
Thi Nguyen
United States
Salt Lake City
Utah
flag msg tools
badge
Avatar
mb
Same thing with Hammer of the Scots - simultaneous commitment of action levels. The reason it doesn't even feel like an SAS is because of how flexible the commitment is - you're not committing to moving any particular piece in any particular place.

Which gives me the following idea:
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Loading... | Locked Hide Show Unlock Lock Comment     View Previous {{limitCount(numprevitems_calculated,commentParams.showcount)}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
{{error.message}}
{{comment.error.message}}
    View More Comments {{limitCount(numnextitems_calculated,commentParams.showcount)}} / {{numnextitems_calculated}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
12. Board Game: Fist of Dragonstones [Average Rating:6.17 Overall Rank:2533]
Thi Nguyen
United States
Salt Lake City
Utah
flag msg tools
badge
Avatar
mb
The reason simultaneous blind bidding games so often feel chaotic is not, I don't think, the simultaneity - it's the way the game exaggeratedly punishes you for getting things wrong.

Take any game of blind bidding + you lose what you bid, no matter what. As somebody pointed out above, this punishes you a lot for a small change in your action. But it's the exaggerated punishment that causes the chaos, not the simultaneity. Compare this came to O Zoo Le Mio, fer example.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Loading... | Locked Hide Show Unlock Lock Comment     View Previous {{limitCount(numprevitems_calculated,commentParams.showcount)}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
{{error.message}}
{{comment.error.message}}
    View More Comments {{limitCount(numnextitems_calculated,commentParams.showcount)}} / {{numnextitems_calculated}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
13. Board Game: Maharaja: The Game of Palace Building in India [Average Rating:7.16 Overall Rank:529]
フィル
Australia
Ashfield
NSW
flag msg tools
designer
badge
I am the wasp / that burrows in! I am the shriek / of twilight din!
Avatar
mbmbmbmbmb
Maharaja is a great example of an SAS game in which a random strategy will get you nowhere. Since there is a penalty if your actions are blocked by the other players, the best strategy is often to choose a flexible selection over a riskier, more profitable course.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Loading... | Locked Hide Show Unlock Lock Comment     View Previous {{limitCount(numprevitems_calculated,commentParams.showcount)}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
{{error.message}}
{{comment.error.message}}
    View More Comments {{limitCount(numnextitems_calculated,commentParams.showcount)}} / {{numnextitems_calculated}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
14. Board Game: Pirate's Cove [Average Rating:6.58 Overall Rank:1061]
Tom Mackel
United States
Redmond
Washington
flag msg tools
mbmbmb
The SAS in this game is most definitely not random, at least not in the probability distribution sense, even from a statistical viewpoint. Of the 6 possible choices of destinations that a player makes on his/her turn, the probability of certain choices is greatly increased or decreased depending on the board state.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Loading... | Locked Hide Show Unlock Lock Comment     View Previous {{limitCount(numprevitems_calculated,commentParams.showcount)}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
{{error.message}}
{{comment.error.message}}
    View More Comments {{limitCount(numnextitems_calculated,commentParams.showcount)}} / {{numnextitems_calculated}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
15. Board Game: The Brotherhood [Average Rating:6.17 Overall Rank:10406]
Cody
United States
Placentia
Select state
flag msg tools
Avatar
mbmbmbmbmb
I think The Brotherhood is a great example of the depth you get when you combine multiple players and evolving payouts. The limited resources available to each player (represented by counter mix and income) also become a factor. The game certainly isn't heavy but employing a random strategy is a pretty surefire way to get wiped out especially with a full compliment of players.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Loading... | Locked Hide Show Unlock Lock Comment     View Previous {{limitCount(numprevitems_calculated,commentParams.showcount)}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
{{error.message}}
{{comment.error.message}}
    View More Comments {{limitCount(numnextitems_calculated,commentParams.showcount)}} / {{numnextitems_calculated}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
16. Board Game: Saratoga [Average Rating:7.12 Overall Rank:2761]
Olivier Clementin
France
Paris
Paris
flag msg tools
badge
Avatar
mbmbmbmbmb
In some wargames, players must simultaneously pick a Tactical Chit to determine the combat outcome.

Most player try to psychologically outguess their opponent, but in Saratoga the game-theory optimum (Minimax) is pretty easy to calculate explicitly , and I have done it here:

http://www.boardgamegeek.com/thread/24381

If both players use the optimal probabilistic strategy, the outcome of the chitpick becomes a purely random distribution (which I have also shown in the link above) . Chit picking is in fact unnecessary, and should be replaced by a random table.

Empires in Arms uses a similar system but the table is more complicated and some subjective evaluation of results is needed (do you prefer morale loss or manpower loss ?).
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Loading... | Locked Hide Show Unlock Lock Comment     View Previous {{limitCount(numprevitems_calculated,commentParams.showcount)}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
{{error.message}}
{{comment.error.message}}
    View More Comments {{limitCount(numnextitems_calculated,commentParams.showcount)}} / {{numnextitems_calculated}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
17. Board Game: 55Stones [Average Rating:5.55 Unranked]
Ralf Gering
Germany
Germany
flag msg tools
designer
55Stones is a Mancala game in which moves are chosen non-simultaneously while the actual move is performed synchronously. For that reason, I don't think that the simultaneous "action" in this game has any randomness. See the comments on a recent 55Stones game at http://www.boardgamegeek.com/thread/109467 .

 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Loading... | Locked Hide Show Unlock Lock Comment     View Previous {{limitCount(numprevitems_calculated,commentParams.showcount)}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
{{error.message}}
{{comment.error.message}}
    View More Comments {{limitCount(numnextitems_calculated,commentParams.showcount)}} / {{numnextitems_calculated}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
Loading... | Locked Hide Show Unlock Lock Comment     View Previous {{limitCount(numprevitems_calculated,commentParams.showcount)}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
{{error.message}}
{{comment.error.message}}
    View More Comments {{limitCount(numnextitems_calculated,commentParams.showcount)}} / {{numnextitems_calculated}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
Front Page | Welcome | Contact | Privacy Policy | Terms of Service | Advertise | Support BGG | Feeds RSS
Geekdo, BoardGameGeek, the Geekdo logo, and the BoardGameGeek logo are trademarks of BoardGameGeek, LLC.