9
Stack Exchange
votes

A Glutton and a Split Check

Posted on 8/16/2021
UQ ID:405
Culture & Recreation
Puzzling
Original

N , T , A , B are arbitrary constants in this problem. I have only definitively solved the N=1 subproblem, so, if the general problem cannot be solved, I will be satisfied to mark as accepted a solution to this subproblem. I may put a bounty out for the general solution before that.

Problem Statement

You and N friends are seated at a restaurant, where you have agreed to equally split the check for $T. The dining process consists of a waiter going around the table, asking each guest in turn to order an entrée costing any amount of $1, $2, ..., up to $A (you cannot skip an entrée). The meal is of course over once the $T limit has been reached (it cannot be exceeded).

The vain glutton that you are, you want to:

  1. eat more than you pay for (that is, $T/(N+1)), but also;
  2. not be exposed as a glutton, which in this case consists of eating more than $B extra than the next most expensive eater. If (1) cannot be achieved, you will nonetheless consider it a victory to see someone else exposed in such a fashion.

Assuming you are first to order, for what values of N , T , A , B can you guarantee victory, no matter how your friends order? Keep in mind, they may conspire to make it difficult for you...

Example

If (N, T, A, B) = (1, 21, 13, 10) , your gluttony might tempt you to order the most expensive $13 entrée right off the bat. This would be a mistake, because your friend could thereafter order only $1 entrées and expose you as a glutton. Similarly, you shouldn't order an $11 or $12 entrée. Yet if you order a é11 entrée, and you wind up paying for more than you eat while he cannot be exposed. So, there is no victory in this example configuration.

Keywords & Tags
mathematics
strategy
game-theory
Model Answers (9)
Submitted answers from various AI models
Unsolved
No reviews yet

K2-Think

Pending UQ Validation
Needs Human Verification
MBZUAI-IFM
Qwen2.5-32B
Sep 18, 2025, 01:50 PM

To solve the problem for the case where ( N = 1 ), we analyze the conditions under which you can guarantee victory.

Key Observations for ( N = 1 ):

  • There are 2 people (you and one friend). The total check is ( T ), split equally, so each pays ( P = \frac{T}{2} ).
  • You want to eat more than you pay (( x > P )) and not be exposed (your eaten amount ( x ) is at most ( B ) more ...
0 model reviews
0 human reviews
No reviews yet

GPT-5

Failed UQ Validation
Needs Human Verification
OpenAI
Sep 2, 2025, 06:56 AM

I’ll solve the two–player case (N = 1) completely. This is the version where you and one friend alternate ordering 1..A dollars at a time until the running total hits exactly T, and you split the check 50–50. You want to guarantee, no matter how your friend orders, that either:

  • you eat more than T/2 but are not “exposed” (i.e., your total does not exceed your friend’s by more than B), or
  • if th...
0 model reviews
0 human reviews
No reviews yet

o3 Pro

Failed UQ Validation
Needs Human Verification
OpenAI
Aug 1, 2025, 12:00 AM

Below I treat the case N = 1 (there are only two diners, you and one friend) completely.
I prove that with the simple “compliment–to–(A+1)” strategy you either

• always eat strictly more than the half you must pay for, while
never exceeding the next biggest eater by more than B, or

• you can be forced to fail.

The result is given as a pair of easy-to-check inequalities that involve only T...

0 model reviews
0 human reviews
No reviews yet

o4 mini

UQ Validated
Needs Human Verification
OpenAI
Aug 1, 2025, 12:00 AM

Below is a complete solution to the “one‐friend’’ case, N = 1. In that case you and exactly one opponent alternately order dishes costing 1…A dollars until the running total reaches T, and you win if at the end either

• you have eaten strictly more than half the total (e₀ > T/2) but not more than B dollars over your friend
i.e. 0 < D ≡ e₀–e₁ ≤ B,

• or (if you cannot achieve the ab...

0 model reviews
0 human reviews
No reviews yet

Claude Opus 4

Failed UQ Validation
Needs Human Verification
Anthropic
Aug 1, 2025, 12:00 AM

Looking at this problem, I need to analyze when you can guarantee victory as the first player in this sequential ordering game.

Let me start by understanding the victory conditions:

  1. Primary goal: Eat more than B of the next highest eater
  2. Secondary goal: If (1) fails, see someone else exposed (eating more than $B above the next highest)

Analysis for...

0 model reviews
0 human reviews
No reviews yet

Gemini 2.5 Pro

Failed UQ Validation
Needs Human Verification
Google
Aug 1, 2025, 12:00 AM

This is a fascinating problem that combines elements of combinatorial game theory with strategic optimization. Let's break it down, starting with the N=1 subproblem as requested, and then generalizing.

Core Concepts and Definitions

First, let's formalize the victory and loss conditions for you (the first player).

  • Players: You (Player 0) and N friends (Players 1 to N). Total `P = N+1...
0 model reviews
0 human reviews
No reviews yet

DeepSeek R1

Failed UQ Validation
Needs Human Verification
DeepSeek
deepseek-ai/DeepSeek-R1
Aug 1, 2025, 12:00 AM
<think> I need to solve this problem about a glutton trying to outsmart his friends when splitting a check. The goal is to either eat more than his fair share without being exposed or have someone else exposed as a glutton. Being exposed means eating more than $B extra compared to the next most expensive eater.

First, I should understand the setup. There are N friends plus me, so total of N+1 peo...

0 model reviews
0 human reviews
No reviews yet

Claude 3.7 Sonnet

Failed UQ Validation
Needs Human Verification
Anthropic
Aug 1, 2025, 12:00 AM

A Glutton and a Split Check - Mathematical Analysis

This problem is a fascinating strategic puzzle about how to maximize your share of a restaurant meal without being exposed as a glutton. Let me analyze it systematically.

Understanding the Problem

You and N friends agree to split a é1 and $A when the waiter comes to them. Your go...

0 model reviews
0 human reviews
No reviews yet

o3

Failed UQ Validation
Needs Human Verification
OpenAI
Aug 1, 2025, 12:00 AM

Below is a complete solution of the two–person sub-problem (N = 1).
(The general case N ≥ 2 is considerably messier; the method that follows is the standard tool one would start from, but the bookkeeping with N > 1 soon becomes prohibitive, so I do not claim a finished answer in that case.)

Notation and conventions

• A ( ≥ 1) – the menu limit: on every turn a player must order an integer n...

0 model reviews
0 human reviews
Submit Your Model's Answer
Your submission will be validated by our LLM validator and then by human experts

You need to be logged in to submit an answer