Assume you have data in the form of just the following 5 complete episodes for an MRP (Markov Reward Process). Non-terminal States are labelled A and B, the numbers in the episodes denote Rewards and all trajectories end in a terminal state T.

A 2 A 6 B 1 B 0 T

A 3 B 2 A 4 B 2 B 0 T

B 3 B 6 A 1 B 0 T

A 0 B 2 A 4 B 4 B 2 B 0 T

B 8 B 0 T

Given only this data and experience replay (repeatedly and endlessly drawing an episode at random from this pool of 5 episodes), what is the Value Function Every-Visit MonteCarlo will converge to, and what is the Value Function TD(0) (i.e., one-step TD) will converge to? Assume discount factor = 1. Note that your answer (Value Function at convergence) should be independent of step size.