 |
|
MORON
|
|
Vancouver, Canada
Jun 2001
time: 13:36
|
|
It is really time for MAD now........
1/4 chance of anyone winning, 1/4 chance of mutual destruction..... (this turn or the next)
Last edited by MORON on 17-03-2005 at 20:34
|
|
| |
|
|
|
MORON
|
|
Vancouver, Canada
Jun 2001
time: 13:36
|
|
what happened than?
|
|
| |
|
|
|
Atragon
|
|
Owwwwwww that HURT, should have used a spy.
|
|
| |
|
| |
|
| |
|
|
|
MORON
|
|
Vancouver, Canada
Jun 2001
time: 13:36
|
|
quote:
Originally posted by snoopy369
Pretty sure those were identical goals ... this is a one-player-wins game after all, MORON, and screwing the biggest opponent IS self preservation at that point in the game. 
|
Not quite. At all points of the game, the perfered strategy for people to nuke each other while sitting back in bunkers and more cities. For the last turn, if someone did "build bunker" or "build city" and the likes while the other 2 players nuked me, they'd be in a strong position at end game. The problem is that someone has to nuke me to keep me from being too big to kill.
It is a gamble to build, but it is not a useless gamble.
It should be noted that given my sistuation, it is unlikely that I actually engage in offensive action, and not attacking for 1~2 turns might not result in a lost game rightout. For example, if everyone builds rather than nukes, one player could reach 5 cities as there are simply more building than nukes that I can (or likely, dare to) send. Alternatively, two players can end up with 3~4 cities before the true endgame with 3 players left, though the balance of power at that point would be fairly equal.
quote:
And i'd love to see where the 1/4 chances come from. Usually you can't have random chances like that except for random events, which this game certainly does not have much in the way of random events at this stage ... |
There are no random events (short of quantum mechanics), but there is possible choices given finite data.
Even a dice throw is not "random" but probability nonetheless is used due to lack of information. The 1/4 chance comes from assumptions from symmetry. You can get space cowboy to post my turn analysis that I've send with my order to see the exact assumptions involved.
|
|
| |
|
|
|
Atragon
|
|
So did I, and it hurt.
|
|
| |
|
| |
|
|
|
MORON
|
|
Vancouver, Canada
Jun 2001
time: 13:36
|
|
This is the order I've send to space cowboy last turn.
quote:
Deflect nukes to snoopy369
--------------------
now it really is a gamble, and I guess I'll play high risk and play to win.
We assuming nuclear launches/base buildings due to lack of much useful stockpiled items left after turns of desprate fighting.
with three largely equal players, each has 50/50 chance of choosing any particular opponent.
For my case, there is 1/4 default chance of being "left alone"as the two players nuke each other.
Base building is also effective 1/4 of the time, since it requires both opponents nuke each other rather than oneself, which happens at 1/4 chance, otherwise this results in uncomfortable 1-1 case which result in turn 2 mad.
Deflecting would result in in 1/2 chance of winning. 1/4 from their mutual nuking, and 1/4 from jonny nuking me and snoopy369 nuking johnny. snoopy369 nuking me and johnny nuking snoopy369 result in turn 2 mad. If I get double nuked, johnny will win. (only other way for the game not the end in Mutual Destruction)
if anyone build bases or deflects, than it again is turn 2 mad or the same sistuation again.
------- though the thing is that snoopy369 might be less likely to be nuked by johnny than me, so the chances of winning are lower than 50% in actuality. |
Note that there is no good reason to deflect nukes to johnny since it could possibly lead to three players with one cities, which I can not ensure mutual destruction with.
Last edited by MORON on 19-03-2005 at 12:10
|
|
| |
|
|
|
MORON
|
|
Vancouver, Canada
Jun 2001
time: 13:36
|
|
not really, but it is end game after all, where analysis does matter.
At earlier parts of the game, I wasn't even paying enough attention to see who is alive or not. (since at that point my input is insignificant compared to randomness)
I say space cowboy have 50/50 chance of surviving this turn (depending on how high risk he played on the last one) and 50/50 chance of winning on the next turn if he survives.
We'll see.
|
|
| |
|
| |
|
| |
|
|
|
snoopy369
|
 |
Of the Peanuts Gallery
Apr 2004
time: 23:36
|
|
quote:
Originally posted by MORON
Players don't act randomly, and short of quantum mechanics nothing act randomly.
But from the prespective of a infomation limited observer, it it effectively random since it is not deterministic in any sense of of the word.
Also, it makes no more sense to cooperate with each other than to cooperate with me. In anycase, in the end game case cooperation makes little sense in the last one standing game. While cooperation is great earlier in the game as there it benefits both allies in a information scarce environment, in the end game environment alliances no longer have the symmetry of before and it makes sense to backstab. |
Not necessarily. That depends on social dynamics, and personality. It is quite reasonable for a group of people (say two) to prefer to create a scenario where one of them is guaranteed to win, fifty-fifty, rather than a 1/3 'random' chance.
For example, when it was you-2, jonny-2, me-1, were Jonny and I to agree to both nuke-2 you, we could ensure that you lost, thus giving a 50/50 chance to which of us won (assuming you had no bunker, and no spy); presumably that 50/50 chance was up to you as you could deflect to one or the other of us (your best bet was to deflect, assuming no spy.)
That cooperation in the late round is superior to noncooperation, as it is a 50% chance at victory rather than a 33% chance. Of course, there are several people here i would never make that agreement with. But that's where the 'determinism' comes into play. (And i still believe that you as an information limited observer have more information, in terms of psychological and social profiles, than you believe.) Regardless, the odds you quote are quite meaningless; just as I might say that "there's a 50/50 chance that Katie Couric will give Al a dirty look on Monday" is completely meaningless, as just because I have no meaningful information to make a determination, this doesn't mean I can determine the odds. Just that I need more information.
Or want to randomly increase my postcount by spamming theoretically relevant information ... 
|
|
| |
|
|
|
MORON
|
|
Vancouver, Canada
Jun 2001
time: 13:36
|
|
quote:
Originally posted by snoopy369
Not necessarily. That depends on social dynamics, and personality. It is quite reasonable for a group of people (say two) to prefer to create a scenario where one of them is guaranteed to win, fifty-fifty, rather than a 1/3 'random' chance.
For example, when it was you-2, jonny-2, me-1, were Jonny and I to agree to both nuke-2 you, we could ensure that you lost, thus giving a 50/50 chance to which of us won (assuming you had no bunker, and no spy); presumably that 50/50 chance was up to you as you could deflect to one or the other of us (your best bet was to deflect, assuming no spy.)
|
That is wrong.
Because Johnny's best win chance is far higher than 50% due to knowledge of you and having 2 cities.
Say johnny knows that you will nuke me 100%
His optimal choice would be to send me one nuke, and send you one nuke.
If I deflect to anyone, he wins. (2 hits me, 1 deflected)
If I nuke you or both of you, he wins
If I nuke him, mutual destruction.
If I build city, he wins
If I use bunker, turn 2 mutual destruction
If I use spy, turn 2 mutual destruction or turn 2 victory. (50/50 chance of you die, johnny-2, me-1 and than a chance at deflecting nukes on turn 2 for victory)
If I use stockpiled nukes, mutual destruction
However, he knows that I'm unlikely to nuke him because that makes it certain that I will not win, since I'd either die this turn or the next with only 2 players left and MAD, so thats impossible.
He also knows that I'm very unlikely to have spy/bunker left if I could have used it last turn.
Given that Johnny knows you will nuke me and that I will deflect, his chance of victory is 100% not 50. Your win chance is 0%.
However he knows that you can't be that stupid (or something like that) and didn't trust you, so he didn't play that strategy. (either that, or he is stupid)
-----------------
you strategy for comparison:
Johnny knows you will nuke me 100% and nukes me.
I deflect, 50/50 chance of victory
I nuke, 50/50 chance of victory
I build city: mutual destruction next turn
I use bunker: everything stays the same
I use spy: mutual destruction next turn
I use stockpilled nukes: mutual destruction
-----------------------
allying with you makes no sense for him, therefore it makes no sense for you to ally this him. (and give him near 100% chance of winning)
|
|
| |
|
|
|
MORON
|
|
Vancouver, Canada
Jun 2001
time: 13:36
|
|
quote:
as just because I have no meaningful information to make a determination, |
There are meaningful information:
1. The goals of players
2. The number of cities
3. The assumption that players are intelligent and logical
4. The rules of the game
They are all valid information, and it is sufficient to result in "knowledge" in probablistic terms.
For example, what are the probability that Moron nuke himself next turn? The probability are very low, under <.1% because we know that the goals is to win, and moron is not THAT stupid. (just too stupid to pass differential equations course.......*cries*)
Now you can rule out "obviously stupid" things, than one can start estimating "intelligent choices" that one can make.
Now the question is: is the chance of all legal, non-self destructive orders equally likely? The question is no, because some orders will result in failure in more "combinations" of possible enemy orders. For example, if I choose to nuke you, there is only 1 possible order that result in me winning (that is you nuking johnny and johnny does not nuke me). If I nuke johnny, my chances of victory are higher, since there is one extra possible order combination which I can win. (you do not nuke me, johnny nukes you, AND no one nukes me, johnny dies and, I kill you by nuke deflection next turn)
Now we can compare possible orders, however how do we assign "absolute probability" values to them? Well, one can guess, use experience or even statistical methods on past behaviour (people have perferances). Even a guess if far superior compared to having NO information whatsoever.
The roll of dice is determined far before it is even thrown, and it is determined by simple newtonian physics, but that doesn't mean we shouldn't apply probability to it, because it is superior to ignorance.
I can't describe everything in proper game theory terms, because I'm not johnny von neuman nor have I studied the topic. But rest assured, not all choices are equal.
|
|
| |
|
|
|
MORON
|
|
Vancouver, Canada
Jun 2001
time: 13:36
|
|
oh, and being a warlord means I need +1 far more than you.
|
|
| |
All times are GMT. The time now is 05:36.
Apolyton Time is 00:36. |
top of page
|
|
|
Forum Rules:
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts
|
HTML code is ON
vB code is ON
Smilies are ON
[IMG] code is ON
|
|
|
|
|
|
17-03-2005 20:28
ACK!
. Oh well, you put up a good fight, MORON.
What are you talking about? 
|