Put . Before , all three players are active. Each unordered pair is selected with probability , and each direction of transfer then has conditional probability . Given the current holdings , a transfer within the first pair has mean product
For transfers within the other two pairs the analogous values are and . Averaging over pairs gives
At and after elimination the product is zero, since an eliminated player never re-enters. In particular the stopped process has increment
The indicator function is -measurable, and the expression in parentheses has zero conditional expectation on its support. The stopped process is adapted and integrable: at fixed , its absolute value is bounded by . Hence
This is the product martingale for three-player token exchange.
Uniform area distribution on the unit disk gives
The Jacobian determinant has absolute determinant , so
Integrating out the other coordinate gives
The joint probability density function factors into these two marginal distribution, proving independence of and . Finally
The angle at can be defined arbitrarily, since that event has probability zero.
An absorbing state has ; once reached, the chain never leaves it. For a nonabsorbing initial state, condition on its first transition and use the Markov property:
Here and . The other equations are
Solving gives
For absorption into either state one or four, let be the expected number of steps, with . first-step analysis gives
Hence and
The expectation is finite because each transient two-step cycle has probability at most of remaining transient; the survival tail is geometrically bounded. This also justifies solving the expectation equations without assuming finiteness in advance.
A confidence interval is a pair of statistics such that, for each parameter value ,
In repeated sampling, at least the stated fraction of intervals constructed by this rule cover the fixed unknown parameter. It is not a posterior probability statement about the parameter after one interval has been observed. Discrete exact constructions can have coverage greater than the nominal value.
For independent observations from a Poisson distribution, has a Poisson distribution with mean . Invert equal-tail tests of that total mean. At an observed integer , the lower endpoint solves
and the upper endpoint solves
Then divide both endpoints by . If , use lower endpoint zero. Each inversion allocates at most noncoverage probability to one side, giving at least total coverage.
Here the single total count is 39. The supplied values with the correct adjacent count indices are for the lower endpoint and for the upper one. Thus
The values 28.6 and 52.2 use the other count index and are not the endpoints of this central interval. This is adjacent count tails in an exact Poisson confidence interval.
The statistical power of a test is , its rejection probability at parameter . Its size is ; for a simple null this is just .
The Neyman-Pearson lemma states that for testing a simple null probability density function against a simple alternative probability density function , a test rejecting where , accepting where it is below , and if needed randomizing on equality to attain the desired size is most powerful among all tests of that size. It follows by integrating and using the size constraint.
The proposed probability density function is nonnegative, and
For an independent sample, cancellation of the factors in the likelihood ratio gives
For this is increasing in . The most powerful test therefore has rejection region
The sum has a continuous positive probability density function on , so no boundary randomization is required. The threshold depends on the null and chosen size, but not on which was specified. The same test is therefore a uniformly most powerful test for the composite alternative .
Starting at one, state two can first be reached only after an even number of steps. At every even time before that hit, the chain is again at one. By the Markov property and part (i), each successive two-step block hits two with probability , while a failure returns to one with probability . Therefore, for the first hitting time ,
Equivalently has the geometric distribution on the positive integers with success probability .
The chain alternates between the classes and . It is irreducible and has period two, since no odd return is possible and each state has a positive two-step return probability. Matrix multiplication gives
The even-time sampled chain stays in its initial class and moves between its two states with probability at each sampled step.
Let and write . Independence gives and . Apply part (b) to the nonnegative variable , which is the Chebyshev inequality:
Thus in probability, proving the finite-variance weak law of large numbers.
For the Metropolis–Hastings algorithm, choose a proposal transition and initialize at a state with . From current , propose and accept it with probability
Otherwise stay at . The accepted off-diagonal transition satisfies detailed balance, since is symmetric in . Thus is stationary. For convergence from arbitrary starting states in the support, choose the proposal so the resulting chain is irreducible and aperiodic. Only ratios of target probabilities are needed, so an unknown normalizing constant cancels.
Start at any integer , for example two. Propose or with probability each; if the proposal is one, reject it immediately. Between distinct valid neighboring states the proposal is symmetric. The Metropolis–Hastings acceptance probability for an upward proposal is
For a downward proposal from , it is
These are exactly the ratios of the supplied negative binomial distribution probabilities; the factor cancels. The rejection at the boundary supplies a self-loop at two. All neighboring positive states communicate, so the chain is irreducible and aperiodic. Detailed balance gives the desired stationary law. Rejecting the invalid move, rather than forcing an upward move at two, preserves the proposal symmetry used in these ratios.
A mixed strategy for Player II is a probability vector , with and . The player chooses pure strategy with probability , independently of the opponent's randomization.
Player I can choose each row with probability . Against column the expected payoff is . Nonnegativity and a strictly positive entry in each column make every one of these finitely many numbers strictly positive. Their minimum
is therefore a guaranteed payoff against every mixture of columns. By the minimax theorem, .
Use inverse transform sampling for the categorical draw. Validate a finite nonnegative probability vector, allowing only a small rounding tolerance in its total, then compare a single uniform draw with the cumulative probabilities:
sim1 <- function(p) {
  if (!is.numeric(p) || length(p) == 0L ||
      any(!is.finite(p)) || any(p < 0) ||
      abs(sum(p) - 1) > 1e-12) {
    stop("p must be a nonempty finite probability vector")
  }
  p <- p / sum(p)
  cumulative <- cumsum(p)
  cumulative[length(p)] <- 1
  u <- runif(1)
  as.integer(which(u < cumulative)[1L])
}
The returned value is when falls in the interval between the previous cumulative sum and the th one. That interval has length , so the output has the required probabilities. Strict comparison skips zero-probability categories even if a simulated uniform value equals an interval boundary. Exactly one call to runif(1) is made.
The function below counts the initial uniformly chosen state as the first of the visited states, and then generates transitions using the appropriate row of :
simchain <- function(n, P) {
  if (!is.numeric(n) || length(n) != 1L || !is.finite(n) ||
      n < 1 || n != floor(n)) stop("n must be a positive integer")
  if (!is.matrix(P) || !is.numeric(P) || nrow(P) == 0L ||
      nrow(P) != ncol(P) || any(!is.finite(P)) ||
      any(P < 0) || any(abs(rowSums(P) - 1) > 1e-12)) {
    stop("P must be a finite square stochastic matrix")
  }
  k <- nrow(P)
  state <- sim1(rep(1/k, k))
  counts <- numeric(k)
  counts[state] <- 1
  if (n > 1) {
    for (step in seq_len(n - 1)) {
      state <- sim1(P[state, ])
      counts[state] <- counts[state] + 1
    }
  }
  counts / n
}
Conditional on the current state , sim1(P[i, ]) selects with probability , proving the Markov property with the desired transition matrix. Initialization is uniform by the same sampler. Dividing the visit counts by gives the requested Empirical occupation proportions of a finite Markov chain, including a valid result when .
The state-zero stationary distribution equation is , so . The state-one equation is , giving . More generally, the stationary probability flow across the cut is balanced, so . This also follows inductively from .
Hence . Since , the geometric series is summable, and normalization gives
These probabilities satisfy detailed balance and therefore all the stationarity equations. The Markov chain is an irreducible Markov chain, so this normalized stationary distribution is unique.
Up to a parameter-independent multinomial factor, the likelihood function is . A Lagrange multiplier for gives , whence and
Since all counts are positive, the log-likelihood function is strictly concave on the positive simplex and tends to minus infinity at its boundary, proving this is the unique maximum.
Under independence of random variables, , so the log-likelihood function separates as . Applying the same constrained maximization twice gives , , and
Twice the difference between the unrestricted and null maximized log-likelihood functions is
the required likelihood-ratio test statistic. Under the null hypothesis, its asymptotic chi-squared distribution has statistical degrees of freedom, the parameter-count difference . Here that is four. Since , reject independence at the 0.1 percent significance level. For four statistical degrees of freedom, the asymptotic upper-tail probability is . This establishes an association in these data; it is not a causal conclusion.
Write for individual survival probability. Conditional on a clump of size , its death probability is . Averaging with the geometric distribution gives
Thus
Here and ; different clumps and experimental repetitions are understood independent for the sampling model.
The live-clump counts have independent binomial distributions, all with the same . Up to a constant, their combined likelihood function is , maximized at when . Solving for gives . Therefore
Monotonicity of in makes this the maximum-likelihood estimator. At the sampling boundaries, gives , while gives only a supremum as , with no finite maximizer. These are the limiting interpretations of the formula.
The general time-independent Schrödinger equation is . For a spherically symmetric wave function, ; the Coulomb attraction between the proton and electron is . These supply respectively the kinetic and potential terms in the displayed radial equation. Rearranging identifies
as the Bohr radius.
Substitute into the rearranged equation and put . Equating the coefficients of after cancelling the exponential gives
A nonzero state has . If , these identities give and . If , they give , , and . Thus the two independent eigenfunctions and their corresponding energies are
They are eigenfunctions at different energies, rather than two independent solutions at one fixed energy.
For the higher-energy state, radial probabilities include the spherical volume factor . The normalization constant and cancel in the ratio. Using the supplied integral at and at infinity,
Hence
This calculation fixes the denominator as in the source's unparenthesized expression; interpreting it as a product with would give a negative probability.
At each index the two independent uniform symbols match with probability . The match indicators are independent across indices. Consequently the waiting times between successive matches are independent geometric random variables on with success probability , and is their sum.
For one waiting time , summing the geometric series gives . Independence yields
For the moment-generating function is finite for . When , every index matches, deterministically and the formula reduces to for every real .