Recommend
 
 Thumb up
 Hide
39 Posts
1 , 2  Next »   | 

Alchemists» Forums » Strategy

Subject: When can you know for sure which color an opponent's published theory is hedged on? rss

Your Tags: Add tags
Popular Tags: [View All]
Scott Seifert
United States
Little Canada
Minnesota
flag msg tools
badge
Avatar
mbmbmbmbmb
Here's an example from a recent game on boiteajeux:

My opponent has tested four times and has gotten Green+, Blue-, and Soup twice. They then published about feathers saying it was Green+, Blue-, Red+. They were the first to publish this game, so this is all the information they could possibly know.

Assuming they weren't being deceitful (or mistaken, which is difficult on boiteajeux), is it 100% certain that they hedged on red? My instincts tell me so and I've failed to come up with theoretical test results that would give them enough information to narrow down red (hedging on a different color).
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
John Henry
United Kingdom
Wellington
West Midlands
flag msg tools
Avatar
Why would you assume they're not being deceitful?
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Cameron McKenzie
United States
Atlanta
Georgia
flag msg tools
badge
Avatar
mbmbmbmbmb
Really, the question is whether or not it is possible to know the red symbol for sure after those tests.

I believe I tried every relevant combination of ingredients that could have been tested. The best I could do was narrow down four ingredients to two possibilities each, but those possibilities were always different in red sign.

He doesn't know the red.

But it is possible he's gambling with an unhedged seal based on something he inferred from another player.

It's also possible he deliberately published green or blue incorrectly and hedged on that. Red may also be wrong in that situation but he may not be penalized if he's debunked on the color he hedged.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Gareth Reynolds
New Zealand
Christchurch
Canterbury
flag msg tools
Avatar
mbmbmbmbmb
I've not managed to work out how to find the red sign with that information. I don't claim expertise however.
There's a chance that your opponent has paid attention to the ingredients taken by others and the results received from tests conducted by others. Of course, hedging for an early publication is well within the realm of possibility.
There's also no guarantee that they've hedged on Red if only one of the tests was for Feather, I recently hedged Green when all I knew was Red.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nicola Bocchetta
Italy
Milano
MI
flag msg tools
Avatar
mbmbmbmbmb
MasterDinadan wrote:

It's also possible he deliberately published green or blue incorrectly and hedged on that. Red may also be wrong in that situation but he may not be penalized if he's debunked on the color he hedged.


Actually, if he's wrong on both blue and red and he has hedged blue, he would be penalized anyway.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Andy Leber
Canada
Orillia
ON
flag msg tools
Yin
badge
Yang
Avatar
mbmbmbmbmb
Faso74it wrote:
MasterDinadan wrote:

It's also possible he deliberately published green or blue incorrectly and hedged on that. Red may also be wrong in that situation but he may not be penalized if he's debunked on the color he hedged.


Actually, if he's wrong on both blue and red and he has hedged blue, he would be penalized anyway.


At the end of the game, but not necessarily on that debunk attempt.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nicola Bocchetta
Italy
Milano
MI
flag msg tools
Avatar
mbmbmbmbmb
Holmes108 wrote:
Faso74it wrote:
MasterDinadan wrote:

It's also possible he deliberately published green or blue incorrectly and hedged on that. Red may also be wrong in that situation but he may not be penalized if he's debunked on the color he hedged.


Actually, if he's wrong on both blue and red and he has hedged blue, he would be penalized anyway.


At the end of the game, but not necessarily on that debunk attempt.


Sorry, you're right
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Robert Stewart
United Kingdom
Newcastle-upon-Tyne
flag msg tools
mbmbmbmbmb
You can never know for sure how an opponent has hedged until you see the back of the seal, or see enough of his other seals. The best you can do is deduce which colours it would have been sensible to hedge on.

In this case, he definitely doesn't know the red aspect (assuming no-one else has given it away somehow) so it's plausible that he's published truth about the other colours and hedged red. It's also plausible that he's deliberately lying about one of the other aspects intending to debunk himself later, in which case that aspect will be hedged, or that he's bluffing, only knows one of the aspects, and has hedged red as the most likely to be debunked...
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Ben Williams
msg tools
The big-small pattern of alchemicals are set up so that if you pick a colour (say red) and change all r+ to r- and all r- to r+ in alchemicals and in test results, leaving everything else the same, you still get a valid game.

Example: R+, G+, B+ combines with r+ G- b- to make r+. Now change the sign of red: r- g+ B+ combines with R- G- B- to make R-. Now change the sign of G. R- g- b+ combines with r- g+ B- to make R-.

In particular, this implies that without doing an experiment that 'is red' and with no other information, you cannot decide between r+ and r-. I put 'is red' in quotes, because a partial result (say from selling a potion) is enough sometimes.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Philip Morton
msg tools
Avatar
RiemannZetaJones wrote:
The big-small pattern of alchemicals are set up so that if you pick a colour (say red) and change all r+ to r- and all r- to r+ in alchemicals and in test results, leaving everything else the same, you still get a valid game.

This happened in the first game I played using Master Mode debunking. There were 4-5 things published, and a player knew at least one of the things on the board was wrong, so he kept trying to demonstrate conflicts....except EVERYTHING on the board was wrong on Blue, so the conflicts kept failing.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Wei-Hwa Huang
United States
San Jose
California
flag msg tools
designer
Avatar
mbmbmbmbmb
golden_cow2 wrote:
Here's an example from a recent game on boiteajeux:

My opponent has tested four times and has gotten Green+, Blue-, and Soup twice. They then published about feathers saying it was Green+, Blue-, Red+. They were the first to publish this game, so this is all the information they could possibly know.

Assuming they weren't being deceitful (or mistaken, which is difficult on boiteajeux), is it 100% certain that they hedged on red? My instincts tell me so and I've failed to come up with theoretical test results that would give them enough information to narrow down red (hedging on a different color).


It's 100% certain that they don't know the red of feathers. However, that doesn't come close to saying that they hedged on red.

An easy example is, perhaps their Green+ and Blue- results were on four different ingredients. Then, they don't know two colors on any of the eight ingredients. At that point all publication guesses are 1 in 4, so there's no particular reason they should choose the Red hedge over the hedge in the other color they don't know.

I can construct other, more complicated examples.

I do have a counter-question, though -- why is it important to you to know what color they hedged on?
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
David desJardins
United States
Burlingame
California
flag msg tools
badge
Avatar
mbmbmbmbmb
onigame wrote:
I do have a counter-question, though -- why is it important to you to know what color they hedged on?


You might want to know if you're ahead or behind in order to decide how risky to play.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Jeff Carter
United States
St. Louis
Missouri
flag msg tools
badge
Avatar
mbmbmbmbmb
DaviddesJ wrote:
onigame wrote:
I do have a counter-question, though -- why is it important to you to know what color they hedged on?


You might want to know if you're ahead or behind in order to decide how risky to play.

How would knowing which color they hedged on tell you if you're ahead or behind?
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
David desJardins
United States
Burlingame
California
flag msg tools
badge
Avatar
mbmbmbmbmb
IdleHacker wrote:
[q="DaviddesJ"]How would knowing which color they hedged on tell you if you're ahead or behind?


If their theory is wrong, it's likely to be debunked before the end of the game, and if they hedged the wrong color they will score less points than if they hedged the right color.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Philip Morton
msg tools
Avatar
JBH1 wrote:
Why would you assume they're not being deceitful?

rmsgrey wrote:
It's also plausible that he's deliberately lying about one of the other aspects intending to debunk himself later, in which case that aspect will be hedged, or that he's bluffing, only knows one of the aspects, and has hedged red as the most likely to be debunked...

onigame wrote:
I do have a counter-question, though -- why is it important to you to know what color they hedged on?

Is anyone actually having success with this kind of fake-out publication strategy?

I tried to make it work several times, and it was successful in its secondary goal of having more debunking in my games, but it always seemed to lose to anyone who just played things straight. Having to replace the seal before the end of the game costs valuable actions, you're vulnerable to getting conflicted below conference thresholds, and I never actually felt like I managed to catch anybody out with the bad information (e.g. nobody ever endorsed a deceitful theory, or published a wrongly-hedged thing of their own based on the deceitful information). It seems like a crap shoot whether anyone else in the game will even be in a position to care about using your carefully-planted misinformation (=> tested with the ingredient you published but didn't get the result that shows you're lying).

I have had my own deductions tripped up by unintentional bad information (trusting another player who'd turned out to have put a token in the wrong spot on their grid), so I don't disagree that you should keep a certain level of skepticism when trusting any results you didn't produce yourself, but in general I would say you're fairly safe trusting the blue and green information from the publication in the OP, unless there was some obvious tell (e.g. giggling as they put out the seal). If they try to get tricky with it it's more likely in the long run to bite them than you.

Are other people finding better luck with deceitful publishes, to the point where it's a real concern that someone else might be trying it? Does it work better with a specific number of players (most of my games are 3-player)? Is my group unusually distrustful?
4 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
David desJardins
United States
Burlingame
California
flag msg tools
badge
Avatar
mbmbmbmbmb
Chrondeath wrote:
I would say you're fairly safe trusting the blue and green information from the publication in the OP, unless there was some obvious tell (e.g. giggling as they put out the seal).


Remind me to giggle if I play with you.
2 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Cameron McKenzie
United States
Atlanta
Georgia
flag msg tools
badge
Avatar
mbmbmbmbmb
It's a sort of rock-paper-scissors thing.

If the other players are only publishing based on what they know, at worst publishing something they aren't sure of and correctly hedged, you should be able to beat them simply by endorsing their publications, making sales/exhibitions based on the information they published, and buying artifacts.

If your opponent is counting on your to publish correct theories, you'll beat them if you publish misleading theories. If one assumes that the theories are correct, it will lead to multiple failed endorsements and sales.

But, if neither player can trust the information published by the other, the one who publishes more accurately will win.

You have to react to your opponent's play style. If you really think that publishing misleading results is a bad strategy (and you believe the opponent thinks the same) then there is little reason to go through the efforts of performing your own tests. Let your opponent do the tests while you reap the rewards.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
David desJardins
United States
Burlingame
California
flag msg tools
badge
Avatar
mbmbmbmbmb
MasterDinadan wrote:
If the other players are only publishing based on what they know, at worst publishing something they aren't sure of and correctly hedged, you should be able to beat them simply by endorsing their publications, making sales/exhibitions based on the information they published, and buying artifacts.


I doubt it. Has anyone actually tried that? I think the cost of endorsing and the reputation hit would be too much. You also may not know what they are hedging. And you have to get the timing right for conferences. It's also going to be slow to get enough usable information to sell potions.

I'm interested in the idea of playing a game without doing any tests at all. I'm just skeptical about how well it would work.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
thomas metro
Poland
flag msg tools
There are a diffrenet aproach depending on number of players.
In 1 vs 1 game there are less chances that your opponent make same test as you, especially on master variant it takes his time to figure out what test debunk you. So you do not debunk you asap on next round, but start to confilcit your incorrect theories with his good one and denying him from getting grands and conference points. If you do just test and publish hedged one it comes to who got better luck in testing.

In 4p is less reliable strategy mostly due to less cubes and more chances that someone is testing on same, additionally just denying one person is not beneficial as other two are untouched.

Don't forget that on conferences only number of publication matters, intentionaly using wrong token force other to denouce as they would not want to publish on tokens that left, in some situations resulting in negative points (and some penalty if they are high in reputation).
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Pawel Bulacz
Poland
Kraków
flag msg tools
Avatar
mbmbmbmbmb
When I see somebody writing a theory in the second round on something he doesn't have a clue - example he just made one test on Red+ and his theory is Red-, then it's possible that he has hedged on red in order to get some points on debunking his theory.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Cameron McKenzie
United States
Atlanta
Georgia
flag msg tools
badge
Avatar
mbmbmbmbmb
Well, it was a bit of an exaggeration. I'm not saying you can win while never doing any of your own tests or publications, but that if you are sure that opponents are never deceptive (i.e. publishing something they KNOW is wrong with an appropriate hedge) you can gain enough accurate information from them that you will gain an advantage.

I've played with people who only ever trust their own information and completely disregard the publications of others when making decisions. Of course the player who takes a chance and counts on some inferences from opponent info will do better, but not if opponent's start publishing misleading info.

Certain artifacts play into this. For example, Seal of Authority and Printing Press encourage players to publish rubbish theories because it's more rewarding or less expensive to publish.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
David desJardins
United States
Burlingame
California
flag msg tools
badge
Avatar
mbmbmbmbmb
I don't think you can win by being intentionally misleading and wrong in the theories you publish. You might be able to affect which of the other players wins, though.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Wei-Hwa Huang
United States
San Jose
California
flag msg tools
designer
Avatar
mbmbmbmbmb
Chrondeath wrote:
Are other people finding better luck with deceitful publishes, to the point where it's a real concern that someone else might be trying it? Does it work better with a specific number of players (most of my games are 3-player)? Is my group unusually distrustful?


Here's an example, based on actual behavior. I've made two potions this game so far, Green plus and Red plus. I publish a paper that says Toadstool is Green plus and Red plus. Are you bold enough to assume that Toadstool was used in both of my potions, and therefore the green and red aspects are probably correct?

Well, as it turns out, I actually used Claw in both of my potions, and Toadstool was only used in my Red potion. But I chose to take a risk on publishing Toadstool because I happen to have more Claws in my hand, so I'm hoping that I'll actually learn enough about Claw to be able to put a star ribbon on it later instead of having a guaranteed safe hedge.

Does my publication fit your definition of deceitful?
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Philip Morton
msg tools
Avatar
onigame wrote:
Here's an example, based on actual behavior. I've made two potions this game so far, Green plus and Red plus. I publish a paper that says Toadstool is Green plus and Red plus. Are you bold enough to assume that Toadstool was used in both of my potions, and therefore the green and red aspects are probably correct?

Well, as it turns out, I actually used Claw in both of my potions, and Toadstool was only used in my Red potion. But I chose to take a risk on publishing Toadstool because I happen to have more Claws in my hand, so I'm hoping that I'll actually learn enough about Claw to be able to put a star ribbon on it later instead of having a guaranteed safe hedge.

Does my publication fit your definition of deceitful?

Well, it would trip me up if I trusted it, so yes, I would count that. Does this course of action come out with a net positive expected value for you, in your games?

I'm curious what the eventual plan for this tricky publish is. Do you debunk it as soon as you can, or let it kite as long as possible? Do you end up taking earlier wake-up times for the rest of the game so that you can jump in front if it looks like someone else is going to debunk it? Does your choice to do this get influenced by the other players' visible test results (maybe seeing a Green- increases the chance they know your Toadstool's green is wrong right off, or any Blue result means they could possibly debunk you on your non-hedged color immediately)?

Comparing it to the non-tricky course (publishing Toadstool with Red+ Green- hedged on Blue, so you can still do your starred Claw publish later if that's what's motivating you):

* They both have the same immediate effect, +1 publish

* The tricky publish guarantees someone can debunk it later, possibly you, so you can eventually publish a starred seal on Toad and maybe avoid having to research an extra ingredient to get all your star seals out

* but you don't have the information to do it immediately (assuming you're in master mode), and the implication of your hand full of Claws is that you don't have a hand full of Toads to keep testing on

* Having a wrong seal on the board lets you conflict out other players' good seals...

* ...but it also lets other players conflict out your good seals

* and if there's any conflicting going on, you're effectively down a publish because your tricky publish is always involved. (This was the biggest issue when I tried it; it was too easy for other players to make me miss conference thresholds when they realized what was going on by conflicting two of my publications the round of the conference.)

* There's a chance you trick someone into screwing up their deduction when they trust you...but only if they need the green information on Toadstool (roughly, I know there are other cases) during the time period where the tricky publish is kiting and hasn't been revealed yet

* It screws up the eventual publication of the ingredient that really needs that seal--but there's a 50% chance that that's the Claw that you're looking to star, so this seems like a downside

* There's a risk that someone does manage to debunk it on blue and you lose a boatload of points. It's not as large of a risk as it looks, and it never actually happened in my attempts, but it's still a chance

Overall, in the games where I tried it, it just felt like Dick Dastardly stopping to cheat--the chance that it messes up the other players isn't that good, it opens you up to being messed with via conflicts worse than it does the other players, it makes it harder to stay ahead in the seal game because you spent resources on a publish that's definitely going away at some point, and there's a chance you end up stepping on your own face when your other publications need the teardrop you used.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Robert Stewart
United Kingdom
Newcastle-upon-Tyne
flag msg tools
mbmbmbmbmb
I believe the idea was not to deliberately publish a false Toadstool, but rather to publish a possibly correct Toadstool (1/6 chance of being right - 1/3 right on green; 1/2 on blue) that opponents will expect to be correct on both red and green.

In this example, there's a 50% chance of being stuck with a hedged correct theory on Claw if you publish a possible Claw theory, but only a 1/6 chance if you publish a green + possible Toadstool (though you do have a chance of being caught out on your unhedged colour).

The basic idea is that, since your first publication is not going to score anything at the end of the game anyway, you might as well make it something you've got a good chance of being able to replace later, and which isn't going to block you from putting a points-seal out on the first ingredient you solve completely.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
1 , 2  Next »   | 
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.