Use complete rational ratios for L<1 convergence and L>1 divergence, then audit indecision at L=1.
Qualify eventual nonzero terms and absolute ratio limits, not limits inferred from finite ratios.
Givens and symbols
algebra
epsilon
real_complete
G1 — a_n=(1/2)^n/(n+1), b_n=2^n/(n+1) for every n≥0; also c_n=1/(n+1), d_n=1/(n+1)^2.
L — proved absolute-ratio limit
q — domination ratio between L and one
Rules and sources
Some rule statements have reader clarifications in this revision: they are paraphrases, not verbatim quotations. Referenced original source bytes and hashes are unchanged.
mko_definition · cited — Declare the existing limit L=lim|a_(n+1)/a_n| on an eventually nonzero tail. When L<1 choose L<q<1 and establish eventual geometric domination to obtain absolute convergence. When L>1 the term magnitudes eventually fail to tend to 0, so the series diverges. L=1 gives no decision: reciprocal-square and harmonic tails have opposite statuses with that same limit. Use absolute ratios and an actual limit argument, not a finite favorable prefix. Zero denominators and missing limits are separate boundary cases. · mko-ratio-test-series
algebra · prerequisite — Use real-field identities; division requires a nonzero denominator and multiplication by a positive number preserves inequalities.
epsilon · prerequisite — Convergence requires for every positive epsilon an integer N controlling every n≥N by a strict error bound; use the Archimedean property of the reals.
real_complete · prerequisite — Every nonempty bounded-above real set has a supremum, so bounded monotone real sequences converge, including bounded partial sums of nonnegative series.
The errors of the two rational ratios relative to 1 tend to zero, giving the same limit 1; their denominators do not tend to zero, and neither ratio is constant.
S08 · implies
Before
\sum c_n\text{ diverges by dyadic blocks};\quad\frac1{(n+1)^2}\le\frac1{n(n+1)}\ (n\ge1)
After
\sum d_n\text{ converges by telescoping upper comparison};\quad L=1\text{ is inconclusive}
Conditions: none
Rule: epsilon · Uses: G1, S07
Leaf basis: epsilon
Each harmonic dyadic block is at least 1/2; reciprocal-square tails are bounded above by a telescoping series of sum one, giving opposite statuses at the same L.
Substeps and lemmas
No substeps in this guide.
Reverse audits
S08 · premise_audit · G1, S07 — An inconclusive test does not mean the series has no convergence status.
S07 · premise_audit · G1 — A limit of one is not confused with every ratio equaling one.
S06 · premise_audit · S05 — A single ratio above one is not treated as a divergence proof.
S05 · premise_audit · G1 — The greater-than-one bound holds on the entire tail.
S04 · premise_audit · S03, real_complete — Bounded-monotone real convergence is assumed, with the upper bound independently supplied by a geometric sum.
S03 · premise_audit · S01, S02 — Domination covers every n, not merely one favorable ratio.
S02 · premise_audit · S01 — All comparisons in L<q<1 are strict.
S01 · inverse_equivalence · G1 — Absolute values are retained and no zero denominator is used.
Error controls and negative controls
Absolute ratios
Qualify a nonzero tail before proving its absolute-ratio limit.
Critical value
At L=1 use independent arguments instead of forcing a decision.
L=1 means convergence. → reject — c is divergent harmonic.
Ten ratios below one prove convergence. → reject — The later tail may differ; a limit or whole-tail domination is required.
Zero terms permit unconditional adjacent division. → reject — A zero denominator makes the ratio undefined and requires separate treatment.
Result and limitations
a converges absolutely, b diverges, and harmonic versus reciprocal-square series give opposite outcomes at L=1.
Verification and source binding
1 declared checks for this guide; only the listed expressions are covered.