araver

Thank you. I can see your trademark from previous ones and I like it. Afraid my grasp on English is not good enough to solve, but it is good enough to enjoy the riddles.

murder

CHAOS If 2 then _ H _ _ S because S is known, COLOR-0, SPARK-0

Declaring open-season on questions again Ask for a particular "a *|* b =?" and I'll oblige. (Of course, not the pair the OP asks for ) A recap on what was so far (including a question from amateur that I missed, sorry) : 21 *|* 7 = 11 17924 *|* 10751 = 851 1089 *|* 121 = 2069 10 *|* 23 = 6 1 *|* 0 = 2 999 *|* 909 = 1359 14 *|* 14 = 14 2 *|* 1 = 0 1 *|* 1 = 1 and a *|* b = b *|* a

Just passin' the time. SOUTHCAROLINA PWTICBSQ

Think of it this way: You've died in 2 mafias in the same week on the first night. The odds of that happening again are ... Care to make a wager?

My first strategy Step 2 - Final adaptive questions (at most 3) -------------------------------------------- All but one case in the table have at least 1 sure position and at least 1 truth-teller among these positions. Target that one known truthteller (let's call him Oracle) with the rest of the questions. First, let's eliminate most of the cases. If there are 3 or more sure positions, then at most 4 unknowns remain. Therefore 3 questions targeted at the Oracle about 3 unknown positions give a total of 6 sure positions and the last can be deduced by whatever type (R or T) is missing from the (3R, 4T) distribution. If there are 2 sure positions, then it's one of the following 7 cases: 20) NYNNYY- 4 possibilities: TTRTTRR, TTRRTRT, RTRTTRT, RRTTTRT. 2 sure positions: ----TR-. 25) NYYNNN- 4 possibilities: TTRRRTT, RTRRTTT, RRTRTTT, RRRTTTT. 2 sure positions: -----TT. 27) NYYNYN- 4 possibilities: TTRTTRR, TTRRTRT, RTRTTRT, TTRRRTT. 2 sure positions: -TR----. 41) YNYNNN- 5 possibilities: TRTRRTT, RTTRRTT, TRRRTTT, RRTRTTT, RRRTTTT. 2 sure positions: -----TT. 45) YNYYNN- 5 possibilities: TRTRRTT, RTTRRTT, TRRTRTT, TRRRTTT, RRTRTTT. 2 sure positions: -----TT. 57) YYYNNN- 5 possibilities: TRTRRTT, TRRRTTT, RTRRTTT, RRTRTTT, RRRTTTT. 2 sure positions: -----TT. 61) YYYYNN- 6 possibilities: TRTRRTT, TRRTRTT, RTRTRTT, TRRRTTT, RTRRTTT, RRTRTTT. 2 sure positions: -----TT. For each case, 3 well-targeted questions clear the rest of the arrangement. If there is only 1 sure position, then it's one of the following 10 cases: 9) NNYNNN- 4 possibilities: TTTRRRT, RTTRRTT, RRTRTTT, RRRTTTT. 1 sure position: ------T. 21) NYNYNN- 5 possibilities: TTRTRRT, TTRRRTT, RTRTRTT, RRTTRTT, RTRRTTT. 1 sure position: ------T. 23) NYNYYN- 5 possibilities: TTRTRRT, TTRRTRT, TTRRRTT, RTRTRTT, RRTTRTT. 1 sure position: ------T. 29) NYYYNN- 5 possibilities: TTRTRRT, TTRRRTT, RTRTRTT, RTRRTTT, RRTRTTT. 1 sure position: ------T. 37) YNNYNN- 5 possibilities: TRTTRRT, RTTTRRT, TRRTRTT, RRTTRTT, TRRRTTT. 1 sure position: ------T. 43) YNYNYN- 5 possibilities: TRTRTRT, RTTRTRT, TRRTTRT, TRTRRTT, RTTRRTT. 1 sure position: ------T. 47) YNYYYN- 5 possibilities: TRTRTRT, RTTRTRT, TRTRRTT, RTTRRTT, TRRTRTT. 1 sure position: ------T. 53) YYNYNN- 6 possibilities: TRTTRRT, TRRTRTT, RTRTRTT, RRTTRTT, TRRRTTT, RTRRTTT. 1 sure position: ------T. 59) YYYNYN- 4 possibilities: TRTRTRT, TRRTTRT, RTRTTRT, TRTRRTT. 1 sure position: ------T. 63) YYYYYN- 4 possibilities: TRTRTRT, TRTRRTT, TRRTRTT, RTRTRTT. 1 sure position: ------T. For each case, 3 well-targeted questions clear the rest of the arrangement. At last, the infamous branch with 0 sure positions. 19) NYNNYN- 5 possibilities: TTRTTRR, TTRRTRT, RTRTTRT, RRTTTRT, TTRRRTT. 0 sure position: -------. 3 more questions and we're done: Ask the last (7th) in the line if the first one is random. If Yes - TTRTTRR, RTRTTRT, RRTTTRT. Now the mask is ---TTR-. Ask an Oracle if the last is R (one case). If not, ask if second is R (one case for each answer). If No - TTRTTRR, TTRRTRT, TTRRRTT. Now the mask is TTR----. Ask an Oracle if the last is R (one case). If not, ask if the sixth is R (one case for each answer).

Yup ... or else his head might explode Seriously, yes, you cannot ask that kind of questions (according to most opinions out there) since they force a third type of answer: either "I do not know" or "I can't answer" or head-exploding (I like this one most of all). It's similar to a Turing machine or a program that has at least one branch that does not terminate (loops). On the right input, it will give you a Yes/No answer (sooner or later) On the wrong input, you might be waiting for an eternity without knowing for sure if it's going to respond eventually. You could ask it whether Fermat's Last Theorem is true (which would take a while, but eventually the machine would say Yes) or if P=NP (and still wait for the response as I'm doing now ). Safe bet is to try only question that can be answered in all cases.

UTAH TCJUTCJU I cannot stop pondering some things, so if you're curios check the spoiler.

I believe there are several classical definitions for random answerers in these types of puzzles. RandomFlipFlop: He doesn't look at the question. First time he flips a coin then simply says alternatively either Yes or No. RandomWeak: Before answering a question he tosses a coin. If it is heads he lies, if it is tails he tells the truth. RandomStrong: He doesn't look at the question, simply flips a coin and says either Yes or No. I believe bushindo's random's are RandomStrong (from the wording of the puzzle) since this is the most challenging situation of the three So randoms are not predictable by others and their answers don't depend on their previous answers. EDIT: Just seen that bushindo already posted a reply.

bushindo, scratch my previous strategy, I was barking up the wrong tree So I'll be going the long road too, and it seems like it might take me 10 questions, too.

Quite correct so far. In the meantime FLORIDA QTKEAGNQ

3 more: NEWMEXICO FXOGXKEQ NEVADA ADAPFXAD MISSOURI TSPWTKNJ 39 down, 12 to go. EDIT: I'm going to slow down here, since there aren't many left. Posting 1 at a time for the remaining.

Ah, just re-read it:

3 more: NEBRASKA RBSMCPFC MONTANA PNUCPCMP MINNESOTA JNOFUPTC 36 down, 15 to go. Didn't thought I'd make it this far.

A couple more states ... I've got to see what the high-scores are so far. CALIFORNIA MKHQTNIB COLORADO QTBDPEQN DELAWARE BYCSEFGL 33 down, 18 to go. @dawh - Care to give some hints to the rest of the players? Might help.

Actually +5 maurice

TROIS if 1 then _ _ _ I _ since TR cannot be first (TREAT-0), O cannot be third (WHOSE-0) and S cannot be last (STAGS-0) maurice, real-timing again? You're a faster typer than I am.

As I'm experiencing a series of electrical black-outs right now (about one every 30 min), I lost a part of what I had written, so it's going to take a while And since it's late night here, I'm probably going to finish this tomorrow. However, here's a sneak peak of what I'm trying to accomplish:

ANVIL Yeah, I sometimes get fixed on one idea.

Actually, I have a very strong hunch that But I'm very far from actually writing such a successful strategy (at least the first part of it), so for now I guess I'm just bragging You are correct, random answerers are a pain . However it's not completely unavoidable You might want to try for yourself, then come back to this challenge.

ANVIL