Skip to content

"Should I rank my favorite second?" — the plain RCV-IRV betrayal incentive, runnable

The question RCV voters keep asking — and the honest answer is sometimes, yes. Under RCV-IRV, ranking your true favorite first can elect your worst candidate; ranking the compromise first can rescue them. This is favorite betrayal, and here it is in the simplest possible 3-bloc form you can run yourself. Then: STAR and Ranked Robin elect the compromise from the honest ballots — no betrayal required.**

▶ Live on BetterVoting: - Honest ballots (STAR + RCV-IRV + Ranked Robin, same voters): vote · results ↗ (election 3xgkck, BV2227) - The betrayal (RCV-IRV, 2 voters rank the compromise first): vote · results ↗ (election bgcmxx, BV2228)

→ The deep-dive concept page: Favorite Betrayal — the full explainer. Related: center squeeze · strategic voting · what makes a good winner?.

The setup

Three candidates on a spectrum — Left, Center, Right — and 34 voters. Center is everyone's second choice and the Condorcet winner (beats Left 22–12, beats Right 21–13 head-to-head), but has the fewest first-choices.

Each row is one voter's ranking, most-preferred first (N: prefix = N identical ballots).

12:Left>Center>Right
5:Center>Right>Left
4:Center>Left>Right
13:Right>Center>Left

1. Honest ballots → RCV-IRV elects your worst (bv2227_3xgkck_honest_irv)

The 12 Left voters vote sincerely, favorite first:

ROUND 1
Right   13   Hopeful
Left    12   Hopeful
Center   9   Rejected      ← fewest first-choices, squeezed out
Right   18   Elected
Left    16   Rejected

Center — the candidate a majority prefers to either wing — is eliminated first, and Right wins. The Left voters' honest ballots elected their least-favorite candidate. (Right was their bottom choice.)

2. Rank your favorite second → you win (bv2228_bgcmxx_betray_irv)

Change just 2 of the 12 Left voters: they betray their favorite and rank Center first, Left second (Center > Left > Right). Nothing else moves.

ROUND 1
Right   13   Hopeful
Center  11   Hopeful
Left    10   Rejected      ← now Left is fewest, eliminated instead of Center
Center  21   Elected
Right   13   Rejected

Now Left is squeezed out, their ballots flow to Center, and Center wins 21–13. Two voters, by hiding who they actually liked best, flipped the result from their worst choice to their compromise. That is favorite betrayal — and it's exactly why the advice under IRV is "putting your favorite first is only safe when they're either very strong or have no chance at all."

3. The fix: STAR & Ranked Robin elect Center from the honest ballots (bv2227_3xgkck_honest_star)

Score the same honest preferences 0–5 (favorite 5, compromise 3, worst 0) — no strategy:

[Divergence from STAR]
  STAR                   = Center
  Choose-One (Plurality) = Right   (differs from STAR)
  RCV-IRV                = Right   (differs from STAR)
  Note: no ballots had tied scores, so RCV-IRV vs STAR here is a genuine
        method difference, not a tie-breaking artifact.
  Note: Ranked Robin (RCV-RR) agrees with STAR, so RCV-IRV is the lone
        outlier — the classic center-squeeze signature.
  Full round-by-round reports (generated for review):
  RCV-IRV rounds: cases_tabulated/bv2227_3xgkck_honest_star_RCV-IRV_tabulated.txt

--- STAR Voting Method (single winner) ---

[STAR Voting]
 Tabulating 34 ballots.
Count × Left,Center,Right
   13 ×    0,     3,    5
   12 ×    5,     3,    0
    5 ×    0,     5,    3
    4 ×    3,     5,    0

[STAR Voting: Scoring Round]
 The two highest-scoring candidates advance to the next round.
   Center        -- 120 -- First place
   Right         --  80 -- Second place
   Left          --  72
 Center and Right advance.

[STAR Voting: Automatic Runoff Round]
 The candidate preferred in the most head-to-head matchups wins.
   Center        -- 21 -- First place
   Right         -- 13
   Equal Support --  0
 Center wins.
   Runoff math:
     34  ballots cast
   −  0  Equal Support (no preference between the two finalists)
     ──
     34  voters with a preference  (majority = 18)
           Center 21 (62%)  ·  Right 13 (38%)

[STAR Voting: Winner — STAR Voting Method (single winner)]
 Center

STAR → Center. Ranked Robin → Center. RCV-IRV → Right. On the identical honest preferences, the two methods that read the whole ballot elect the compromise the electorate actually prefers; only instant-runoff needed the betrayal. Under STAR you score your honest favorite a 5 and your compromise a 3, and the compromise still wins — no games, no regret.

The takeaway

  • The incentive is real, not a debating trick: under RCV-IRV, sincerely ranking a viable favorite first can elect your worst outcome, and betraying them can prevent it.
  • A common wrong rebuttal: "if your favorite gets eliminated round 1, ranking them second changes nothing." True but beside the point — the betrayal bites precisely when your favorite is strong enough to squeeze the center, not when they're hopeless.
  • STAR and Ranked Robin don't have this hole. They elect the Condorcet-winning compromise from honest ballots, so honesty is safe. That's the pitch — and the reason Later-No-Harm ≠ Favorite Betrayal is worth keeping straight.
  • Fair note: STAR has its own rare strategic seams (min/maxing), conceded in strategic voting and STAR's honest limits. The claim isn't "STAR is strategy-proof" — it's "STAR doesn't punish plain honesty the way IRV can here."

Reproduce it

python3 STARVote_LH_tabulation_engine/starvote_larry_hastings.py method_comparisons/favorite_betrayal_irv/cases/bv2227_3xgkck_honest_irv.yaml
python3 STARVote_LH_tabulation_engine/starvote_larry_hastings.py method_comparisons/favorite_betrayal_irv/cases/bv2228_bgcmxx_betray_irv.yaml
python3 STARVote_LH_tabulation_engine/starvote_larry_hastings.py method_comparisons/favorite_betrayal_irv/cases/bv2227_3xgkck_honest_star.yaml