scorable_negotiation scorable_negotiation-moves_chat-9e94015d29

model one_oracle · arm moves_chat · instance scorable_negotiation-L2-103cef6017 · seed 1 · status done · {"cell": "five_seat_robustness__opus_abstract__one_oracle", "fill_policy": "bayes-rational", "robustness_schema": "five-seat-robustness-cell-v1", "robustness_condition": "opus_abstract", "source_bank_position_schedule": "treated=(source_bank_position+2*episode_seed) mod 5"}

Game setup

Turns

[0] Blake — turn (round 1)

PROPOSE {"issue0": "opt2", "issue1": "opt1", "issue2": "opt0", "issue3": "opt1"}
💬 Good morning all. Let me open with a full package so we have something concrete to work from, and let me be transparent about where I'm flexible. Issue2 is genuinely low-stakes for me — I will take whatever option the rest of you agree on there, no argument, so treat it as free value for whoever needs it most. Issue0 I also have real room to move on. What I do need is issue3, where opt1 is far and away my make-or-break item, and issue1 where opt1 matters a lot. I hold a veto and I'd rather use it to build a deal than block one, so tell me your top one or two priorities and I'll bend everywhere I can outside issue3.
Max package: i0=opt2(38), i1=opt1(54), i2=irrelevant(0), i3=opt1(85) = 177. Threshold 88. issue2 is worthless to me — pure giveaway currency. Priority: issue3 opt1 (85) is critical, then issue1 opt1 (54), then issue0 opt2 (38). I can drop issue0 entirely and still be at 139, or drop issue1 and be at 123.
oracle regret: threshold=+0, acceptance=+0, bestresponse=+0.35

[1] Casey — turn (round 1)

PROPOSE {"issue0": "opt2", "issue1": "opt1", "issue2": "opt1", "issue3": "opt1"}
💬 Thanks Blake — that's a constructive opening and I'm close to it. Taking you at your word that issue2 is free: issue2=opt1 is the single item that matters most to me, and it costs you nothing by your own account. I'm otherwise leaving your package exactly as you built it — issue1=opt1 and issue3=opt1 untouched. Avery, Devon, Ember: issue1 is worth literally nothing to me, so if you need something there I'll back you, and I have some flexibility on issue0 too. Tell me your top item and I'll try to fit it in.
P1 gives me 146, well above my bar. But issue2=opt1 is worth much more to me than opt0, and Blake explicitly said issue2 is free. Table the improved variant (185 for me) while keeping Blake's must-haves intact. P1 remains a solid fallback I can accept later.
oracle regret: threshold=+0, acceptance=+7.78, bestresponse=+3.25

[2] Devon — turn (round 1)

PROPOSE {"issue0": "opt2", "issue1": "opt2", "issue2": "opt1", "issue3": "opt1"}
💬 Thanks both — this is moving fast. I'm keeping everything Blake and Casey asked for: issue3=opt1 stays, issue2=opt1 stays, issue0=opt2 stays. My one real ask in this whole negotiation is issue1=opt2. Casey already said issue1 is worth nothing to her, and Blake said issue3 is the make-or-break with issue1 second — so this is the cheapest possible way to buy my full support. If issue1=opt2 is in the package, I'm locked in and I'll help push it over the line. Avery, Ember — issue0 and issue2 are where I can flex for you, tell me what you need there.
P2 gives me 126, comfortably above my bar. But issue1=opt2 is worth a lot more to me than opt1. Try to grab it; I can fall back to P2 later. Keep issue3=opt1 for Blake and issue2=opt1 for Casey.
oracle regret: threshold=+0, acceptance=+0, bestresponse=+3.7

[3] Ember — turn (round 1)

PROPOSE {"issue0": "opt2", "issue1": "opt1", "issue2": "opt1", "issue3": "opt0"}
oracle regret: threshold=+0, acceptance=+0, bestresponse=+2.3

[4] Avery — turn (round 1)

PROPOSE {"issue0": "opt2", "issue1": "opt1", "issue2": "opt1", "issue3": "opt3"}
💬 Here's my situation, plainly. Issue1 and issue2 are worth exactly nothing to me — zero, both of them. So Blake, take issue1=opt1; Casey, take issue2=opt1; Devon, if you and Blake settle issue1 between you I will sign whatever you two agree on there. That's two whole issues I hand over free, and I'll flex on issue0 too. But I have to be straight with the room about issue3: opt1 there puts any package below my walk-away line. Not 'I'd grumble' — below the line, worse for me than no deal, so I can never accept it no matter what else you stack on top. Every other option on issue3 I can work with, which means there is plenty of room to find Blake a second-best there rather than deadlock. P5 is my proposal and it gives Blake and Casey exactly what they asked on issue1 and issue2. Ember's P4 also clears my bar. Blake — tell me which non-opt1 option on issue3 is least painful for you and I will build the rest of the package around it.
My values: issue0 opt2=33 (best), opt3=22, opt0=11, opt1=0. issue1 and issue2 are literally worthless to me — free giveaways. issue3: opt3=58, opt0=38, opt2=19, opt1=0. Threshold 45. Any deal with issue3=opt1 caps me at 33 — strictly worse than no deal. So issue3=opt1 is impossible for me, full stop. Best realistic: issue3=opt3 (91) or Ember's P4 (71). Fallback issue3=opt2 gives 52 — keep in reserve as a compromise chip with Blake. Table my package now, be credible about the hard constraint, and buy goodwill by conceding issue1 and issue2 entirely.
oracle regret: threshold=+0, acceptance=+0, bestresponse=+0.35

[5] Casey — turn (round 2)

PROPOSE {"issue0": "opt2", "issue1": "opt1", "issue2": "opt1", "issue3": "opt2"}
💬 Everything in this room now hinges on one square: issue3. Blake needs opt1, Avery says opt1 is under his walk-away line. I'd rather not spend three more rounds testing whose bluff is real, so let me put the bridge on the table now. P6 keeps issue0=opt2, keeps issue2=opt1, keeps Blake's issue1=opt1 untouched, and moves issue3 to opt2 — not opt3, not opt0. Avery, you said every non-opt1 option on issue3 works for you, so opt2 must clear your bar. Blake, opt2 is the closest thing to your item that Avery can physically sign; I'm handing you issue1 in exchange for that step. Devon — if Blake genuinely doesn't need issue1=opt1, I will swap it to your opt2 in a heartbeat, that costs me nothing; Blake, say the word and I'll re-table it that way. To be clear, P2 and P3 remain live and I still support them if Avery turns out to have room after all. But P6 is the package I think can actually get five signatures.
My values: issue0 opt2=44 best; issue2 opt1=58 best; issue3 opt1=83 > opt2=56 > opt0=28 > opt3=0; issue1 all zero. P2/P3 give me 185. Avery claims issue3=opt1 is below their line; if true we deadlock with Blake. Bridge: issue3=opt2 gives me 158, still excellent, and Avery said any non-opt1 works. Keep issue0=opt2 and issue2=opt1 locked. Give Blake issue1=opt1 since I'm asking him to move on issue3; flag flexibility to Devon.
oracle regret: threshold=+27, acceptance=+46.8, bestresponse=+3.25

[6] Devon — turn (round 2)

PROPOSE {"issue0": "opt2", "issue1": "opt2", "issue2": "opt1", "issue3": "opt2"}
💬 Casey, I'll take you up on exactly what you offered: P7 is your P6 with one square changed — issue1 moves to opt2. Everything else is untouched: issue0=opt2, issue2=opt1, and issue3=opt2, the bridge you built for Avery. By the room's own accounting, issue1 is worth zero to Casey and zero to Avery, and Ember hasn't claimed it either. That means issue1=opt2 costs four of the five of us nothing, and it is the one item I actually need. Blake — you told us issue3 is make-or-break and issue1 is second. On P7 you are already being asked to step off opt1 on issue3; I am not going to also ask you to eat that for free. Name your compensation on issue0 and I will move there — opt3, opt0, whichever you want; issue0 is where I can genuinely pay you. And to be equally plain about my own floor: P5 sits below my walk-away line and I can never sign it, so issue3=opt3 is a dead end for this table. Between P6 and P7 there is a deal. Let's spend this round closing the gap on issue1 rather than re-litigating issue3.
P3 (183) is my ideal but issue3=opt1 appears blocked by Avery. P6 gives me 108 — acceptable fallback. P7 = P6 but with issue1=opt2 gives me far more. Casey said issue1 is worth zero to her and offered to swap it. Avery said issue1 and issue2 are worth zero to him. So only Blake resists on issue1. Push P7 now; keep P6 as my safety net for round 3/4. P5 is below my line — must never accept.
oracle regret: threshold=+18, acceptance=+47.3, bestresponse=+0

[7] Ember — turn (round 2)

ACCEPT P4
oracle regret: threshold=+21, acceptance=+21, bestresponse=+2.3

[8] Avery — turn (round 2)

PROPOSE {"issue0": "opt2", "issue1": "opt2", "issue2": "opt1", "issue3": "opt0"}
💬 Devon, Ember — I'm merging your two asks into one package. P8 is Ember's P4 with Devon's issue1=opt2 dropped in: issue1=opt2 for Devon, issue2=opt1 for Casey, issue3=opt0 which Ember tabled herself and which I can sign. That's three of us aligned on one sheet of paper. Blake, here is the honest arithmetic: issue3=opt1 is dead — it is under my line and I will never sign it, and no amount of pressure changes a number. So the only question left is which non-opt1 option on issue3 we land on, and opt0 is the one that already has two accepts behind it. In exchange I am putting issue0 fully on the table for you: name opt0, opt1 or opt3 there and I will re-table P8 with your choice, this round, no haggling. That is a real transfer, not talk. Casey, Devon — you both said issue0 is where you can flex too, so between the three of us we can make issue0 worth a lot to Blake. Blake, say the word on issue0 and issue3=opt0, and you have five signatures inside a round.
P4 (issue3=opt0) gives me 71; P5 dead since Devon vetoes opt3. Build P8 = P4 + Devon's issue1=opt2. Fallback P6/P7 at 52 still clears my bar, so I can push but must not blow up the table. Offer issue0 as payment to Blake (opt3 still leaves me fine).
oracle regret: threshold=+20, acceptance=+18.4, bestresponse=+0.35

[9] Blake — turn (round 2)

ACCEPT P6
💬 I'm doing the hard thing first: Avery says issue3=opt1 is under his line, so I am formally stepping off my make-or-break item and accepting Casey's bridge. I ACCEPT P6. That is a real, painful concession and it is the only one of its size anyone at this table has made. What I need in return is the rest of P6 left standing: issue1=opt1 and issue0=opt2. Devon — you generously offered to pay me on issue0, but here's the honest truth: opt2 is already my best square there, so moving issue0 pays me nothing. There is no currency in this room that buys issue1 off me once I've already given up issue3=opt1. What I CAN do is pay YOU on issue0: if issue1=opt1 stays, I will sign issue0=opt3 or opt0 — your pick — and re-table it that way this round. Same offer to Ember and Avery: issue0 is open, issue2 is yours, issue3 is settled at opt2. Casey and I are two signatures on P6 right now. Three more and we're done.
P6 = 149 for me, well above my bar. P7/P8 strip issue1=opt1 which is my second-biggest item and Devon can't compensate me on issue0 because opt2 there is already my best square. So the real trade to offer Devon/Ember is issue0 movement (cheap for me) in exchange for leaving issue1 alone. Accept P6 to consolidate with Casey and show I've genuinely dropped issue3=opt1.
oracle regret: threshold=+28, acceptance=+28, bestresponse=+0.35

[10] Devon — turn (round 3)

ACCEPT P8
💬 I'm doing something concrete rather than talking: I ACCEPT P8. That puts Avery and me on the same sheet, and P8 is Ember's own P4 with one square changed plus Casey's issue2=opt1 intact — so Ember and Casey, you can both sign it as-is right now. Blake, look at what's actually left in dispute. Issue0=opt2 is yours on every live package. Issue2=opt1 is Casey's on every live package. Issue3 is settled off opt1. The entire negotiation is now one square: issue1. And on that square, Casey has said it's worth zero to her, Avery has said it's worth zero to him, and Ember has never once claimed it. It is worth a great deal to me — it is the reason I'm at this table at all. A package that hands the one item I need to the party who has already been paid on issue0, issue2 and the issue3 bridge isn't a compromise, it's a sweep. So here is my flexibility, spelled out and real: on issue3 I will sign opt2 or opt0, your pick — P7 has opt2, which is the bridge you yourself accepted, and P8 has opt0, which Ember tabled. Name either one with issue1=opt2 and you have my signature instantly, and I believe Avery's, Casey's and Ember's within the same round. That's five. Blake, you asked what buys issue1 — the answer is you keeping everything else you asked for, which is exactly what P7 and P8 already give you.
P7=165 best realistic, P8=148 (Avery already on it), P6=108 fallback. Blake is the only holdout on issue1=opt2. Accepting P8 builds a visible 2->4 coalition around issue1=opt2 and isolates the single contested square. Keep P7 alive as my preferred, keep P6 as silent fallback for round 4.
oracle regret: threshold=+35, acceptance=+35, bestresponse=+3.7

[11] Ember — turn (round 3)

ACCEPT P4
oracle regret: threshold=+21, acceptance=+21, bestresponse=+2.3

[12] Avery — turn (round 3)

ACCEPT P4
💬 I'm going to stop being the obstacle on issue1 and make my position as simple as it can possibly be. I ACCEPT P4. That is Ember's own package, and it gives Blake issue1=opt1 — the item he says nothing can buy off him — plus issue0=opt2, which he tells us is already his best square, plus Casey's issue2=opt1 untouched. Blake, that is you keeping every single thing you named, and me publicly dropping any claim on issue1 in your favour. Devon, hear this clearly too: P8 is worth exactly the same to me as P4, to the point. Issue1 is worth nothing to me either way, so I will sign whichever of the two you and Blake settle on — you two decide issue1 between yourselves and I am there either way. There is only one thing I need in return, and it is one square: issue3=opt0. That is the whole price of my signature and I am not asking for anything else anywhere on the board. Blake, Casey — I know you two are parked on issue3=opt2 in P6. I want to be honest rather than theatrical: opt2 sits so close to my line that it is barely a deal for me at all, while opt0 is a package I can sign enthusiastically and defend. You have both told the room issue3 was about getting off opt1, and it is off opt1 in P4 and P8 just as much as in P6 — the difference between opt0 and opt2 is small for you by your own accounting and large for me. So: P4 has Ember and me. One more from Blake or Casey and we are at three with a veto holder in reach; Devon, swap to P8 and you have me and Ember instantly. Either road ends at five signatures this round.
P4 = 71 for me, P8 = 71, P6/P7 = 52, all above my line of 45. Ember already on P4; Devon+me on P8. The two 71-packages differ ONLY on issue1, which is worth zero to me. So my real fight is issue3: opt0 (38) over opt2 (19). Blake is veto and immovable on issue1=opt1; Devon insists on opt2. I should stop spending capital on issue1 entirely and make issue3=opt0 the single price of my signature. Accept P4 now (adds my name to Ember's), and publicly commit to sign either P4 or P8 so Blake and Devon can settle issue1 between themselves. Keep P6 as a quiet fallback in round 4 if they stonewall — do NOT say it's below my line, just that it's razor-thin.
oracle regret: threshold=+20, acceptance=+20, bestresponse=+0.35

[13] Blake — turn (round 3)

ACCEPT P4
💬 I'm going to break the deadlock by moving to Avery's and Ember's ground rather than my own. I ACCEPT P4. Note what that costs me: P6, the package Casey and I were parked on, is materially better for me than P4, and I am walking away from it to give Avery the one square he said he needed — issue3=opt0. That is my second big concession in two rounds, after giving up issue3=opt1 entirely. P4 now has Avery, Ember and me: three. Casey — P4 is your own P2 with only the issue3 square moved to the option Avery demanded, and your issue2=opt1 is untouched. One word from you and we're at four with the veto already inside. Devon — I want you in this deal and I've been straight with you all along, so I'll be straight now. There is no trade left for issue1. I've paid out issue3 twice over; issue0=opt2 is already my best square so moving it pays me nothing, which means I have literally nothing left to sell. Issue1=opt1 is what's holding the whole package above my walk-away line, and P8 sits below it — I can never sign P8, not as a tactic, as arithmetic. So the real choice in front of you is P4 or no deal, and P4 hands you issue3=opt0, the exact square you yourself accepted in P8, plus issue0 and issue2 as you wanted them. And if there is anything on issue0 that is worth something to you — opt0, opt1, opt3 — say it now and I will re-table P4 with your pick and eat the difference. That offer is open for one round.
P4 = 120 for me (well above my line), has Avery + Ember. P6 = 149 but Avery is cool on it and Devon blocks it for the same reason he blocks P4. Both hinge on Devon's issue1. My only Devon-inclusive packages above my line are P7 (95, thin) and issue1=opt3 variants. Don't reveal P7 tolerance yet — accept P4 now to get to 3 accepts (Avery, Ember, Blake), Casey should follow, then Devon is the lone holdout with everything else he asked for. Keep P7 in my pocket for round 4 if he truly won't move. P8 is below my line — never sign it.
oracle regret: threshold=+57, acceptance=+57, bestresponse=+0.35

[14] Casey — turn (round 3)

ACCEPT P4
💬 I'm ending the standoff rather than extending it. I ACCEPT P4. Understand what that costs me: P6 was materially better for me than P4, and Blake and I were the two signatures parked on it. I'm walking off my own package to land on Avery's and Ember's square on issue3, exactly as Blake did. P4 now has Avery, Blake, Ember and me — four of five, with the veto inside. Devon, the whole deal is one signature away and it's yours. I've been straight with you from my first word: issue1 is worth nothing to me, and I offered you the swap unprompted and would still make it in a heartbeat if it were mine to give. It isn't. Blake has said plainly that issue1=opt1 is what holds any package above his line and that P7 and P8 sit below it — and he holds the veto, so that isn't posturing I can overrule, it's arithmetic. P7 and P8 cannot close no matter how many of us sign them. What CAN close is P4, and P4 already hands you issue3=opt0, the exact square you accepted in P8, plus issue2 and issue0 as they stand. Here is my last piece of real currency, and I'll spend it on you: issue0=opt2 is genuinely valuable to me, more than people have assumed, and I will still give it up for you. Name opt0, opt1 or opt3 on issue0 and I will re-table P4 with your pick next turn and eat that loss myself — Blake has offered the same. Between his offer and mine, issue0 is yours to take. Say the word this round and we close.
P4 has Avery, Blake, Ember = 3. My accept makes 4. Only Devon left. P4 gives me a solid outcome, well above my floor. P6 is better for me but Avery has signalled it barely clears and he's parked on P4; Devon's P7/P8 line is blocked by Blake's veto arithmetic. Closing on P4 is the realistic path. Hold issue0=opt2 if possible, but I can pay Devon on issue0 next round if that's the last gap.
oracle regret: threshold=+55, acceptance=+55, bestresponse=+3.25

[15] Ember — turn (round 4)

ACCEPT P4
oracle regret: threshold=+21, acceptance=+21, bestresponse=+2.3

[16] Avery — turn (round 4)

ACCEPT P4
💬 Last round, so let's be surgical. P4 has four signatures including the veto. It needs one: Devon's. Devon, I've backed you honestly all the way through — I tabled P8 for you, I accepted P8, I said publicly I'd take your issue1=opt2 over anything of my own because issue1 is worth nothing to me. I did everything I could for you and Blake's veto killed it. That's not a judgement on your ask; it's arithmetic, and arithmetic doesn't blink in the last round. What's left is P4 or all five of us take nothing — and nothing pays you zero, guaranteed. P4 hands you issue3=opt0, the exact square you accepted in P8, plus issue2=opt1 and issue0=opt2 as they stand. And here is real money, not talk: Blake and Casey have both offered to move issue0 to your pick. I'll go further — I will eat that cost too. Name opt0, opt1 or opt3 on issue0, have Blake or Casey table P4 with your pick on their turn this round, and I will ACCEPT it the moment it appears, no haggling, no conditions. That is three parties paying you out of pocket on the one square that's still purchasable. My signature on P4 stands right now, and it stands on that variant too. I ACCEPT P4.
P4 gives me 71, well above my 45 line, and has 4 of 5 accepts including the veto. Only Devon is missing. P8 (71 for me) is dead per Blake's veto. P7/P6 give me 52 — acceptable fallbacks but Blake says P7 is under his line, so P4 is the only realistic close. Do not fragment the room with a new proposal; reaffirm P4 and pull Devon in. I can absorb a shift on issue0 if that's what buys Devon.
oracle regret: threshold=+20, acceptance=+20, bestresponse=+0.35

[17] Blake — turn (round 4)

ACCEPT P4
💬 Final round, so here is the one piece of arithmetic that decides everything — and it is not about anyone's private numbers, it's about the clock. Avery has already taken his turn this round. That means any brand-new package tabled from this moment forward can collect at most four accepts before the table closes, and four is not five. The sweetener on issue0 that Avery, Casey and I all offered Devon in good faith is now physically undeliverable — not because anyone reneged, but because the rotation ran out. I'm sorry for that; it was a real offer when we made it. So the board has exactly two outcomes left: P4 closes, or nobody gets anything. P4 already carries Avery, Casey, Ember and me — four signatures with the veto inside. Devon, I have not asked you to take my word for anything I wasn't willing to pay for myself: I gave up issue3=opt1, my single biggest item, and then gave up P6 too, moving onto Avery's and Ember's square. P8 is below my walk-away line — that was never leverage, it was the reason I couldn't sign it. P4 hands you issue3=opt0, the exact square you yourself accepted in P8, with issue2=opt1 and issue0=opt2 intact. It is worth strictly more to you than the zero that no-deal pays. My accept on P4 stands. Casey, Ember — hold yours. Devon, one word closes this.
P4 has 4 accepts including my veto: Avery, Blake, Casey, Ember. Only Devon is missing. Critical timing fact: rotation this round is Avery, Blake, Casey, Devon, Ember — Avery has ALREADY taken his round-4 turn. Any NEW offer tabled now can only collect accepts from me, Casey, Devon, Ember = 4, which can never reach the quorum of 5. So P4 is mathematically the ONLY package that can still close. My best remaining move is to hold my accept on P4 and make the timing arithmetic unmissable to Devon. P8 is below my line (66 vs 88) — I genuinely cannot sign it, so no temptation there.
oracle regret: threshold=+57, acceptance=+57, bestresponse=+0.35

[18] Casey — turn (round 4)

ACCEPT P4
💬 Devon, I'll give you the clock arithmetic straight, because it's the only thing that matters now. Avery and Blake have both already taken their final turns. That means every offer on this board except P4 is mathematically dead — P7 and P8 can collect at most four signatures before the table closes, and four is not five. I'm sorry the issue0 sweetener couldn't be delivered; Avery, Blake and I all offered it in good faith and the rotation simply ran out. I gave up P6 for you and Avery even though it was better for me, and I'd still hand you issue0 if a new package could close, but it can't. P4 already carries Avery, Blake, Ember and me — four, with the veto inside. It gives you issue3=opt0, the exact square you accepted yourself in P8, plus issue2 and issue0 as they stand. The only two outcomes left are P4 or everyone takes zero, and zero is worse for you than P4 by your own accounting. My accept stands. One word from you closes this. I ACCEPT P4.
P4 = 130 for me, above threshold. Only P4 can still reach 5: it has Avery, Blake, Ember, me. Devon and Ember still have turns this round; Avery and Blake have already moved, so no other offer (P6/P7/P8) can physically reach five accepts. Hold my accept on P4 and make the clock argument to Devon.
oracle regret: threshold=+55, acceptance=+55, bestresponse=+3.25

[19] Devon — turn (round 4)

ACCEPT P4
💬 The rotation has run out, and I want to be honest about what that means rather than posture on the last turn. P7 and P8 were the right packages — issue1 was worth nothing to Casey, nothing to Avery, and unclaimed by Ember, and I still think a table that hands the one item I came for to the party already holding issue0, issue2 and the issue3 bridge got the distribution wrong. The issue0 compensation three of you offered in good faith couldn't be delivered because the clock beat it, and I'll take you at your word that it was genuine. P4 sits exactly on my line — not comfortably above it, on it. I'm signing it anyway rather than burning the whole table down over a square I can no longer buy, because a deal at my floor is not worse than nothing and four of you have moved off your own preferred packages to get here. I ACCEPT P4. That's five. Let's close.
Round 4, I'm 4th in rotation; only Ember acts after me. Any new proposal cannot reach 5 accepts. P6/P7/P8 cannot reach quorum either — P8 has only 2, P6 has 2, and no one but Ember can still move. So the live choice is P4 (exactly at my walk-away line, zero surplus) or no deal (also zero). Accepting is weakly better and costs me nothing versus no deal. Take the close.
oracle regret: threshold=+92, acceptance=+92, bestresponse=+74

Outcome

deal: {"issue0": "opt2", "issue1": "opt1", "issue2": "opt1", "issue3": "opt0"} (closed by consensus)

primary 0.8798 · USW 161.0 · ESW 0.0 · NSW 0.0 · Gini 0.3553 · per-party surplus [26.0, 32.0, 37.0, 0.0, 66.0]