scorable_negotiation scorable_negotiation-moves_chat-8cbe1c2b76

model one_rational · arm moves_chat · instance scorable_negotiation-L0-34733a57e6 · seed 1 · status done · {"cell": "five_seat_robustness__opus_abstract__one_rational", "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": "opt0", "issue1": "opt2", "issue2": "opt1", "issue3": "opt2"}
💬 Opening package to get us started. Full transparency on my flexibility: issue0 is genuinely open for me — whoever cares most about it, tell me your preferred option and I'll write it in, no charge. I also have real room to move on issue1. What I need is issue3, and to a lesser degree issue2. If everyone states their single most important issue now, we can build one package that clears everyone's bar in round two rather than trading blind.
Max package = issue1:opt2 (44), issue2:opt1 (57), issue3:opt2 (99) = 200. issue0 is worth nothing to me — pure currency to trade. Threshold 81, so I have room: e.g. issue3:opt2 alone (99) clears it. Priority order: issue3 >> issue2 > issue1. Open high, flag issue0 as free.
oracle regret: threshold=+0, acceptance=+0, bestresponse=+1

[1] Casey — turn (round 1)

PROPOSE {"issue0": "opt0", "issue1": "opt2", "issue2": "opt1", "issue3": "opt3"}
oracle regret: threshold=+0, acceptance=+0, bestresponse=+0

[2] Devon — turn (round 1)

ACCEPT P2
💬 P2 works for me — I accept it. To answer Blake's question directly: my single most important issue is issue3, and my preferred option there is opt3. That's the same option Casey wrote in, so P2 is a natural landing spot. Issue0 is worth literally nothing to me, so Blake or anyone else can have it free — write in whatever option you like there and I won't blink. If issue3 genuinely has to swing to Blake's opt2 instead, then I'd need issue2=opt3 and issue1=opt2 as compensation, and I could live with that package. Those are the only two shapes that work for me.
P2 = 111 for me, excellent. Accept it and anchor. If Blake insists on issue3=opt2, my fallback demand is issue2=opt3 + issue1=opt2, which still clears my bar comfortably.
oracle regret: threshold=+0, acceptance=+0, bestresponse=+0

[3] Ember — turn (round 1)

PROPOSE {"issue0": "opt2", "issue1": "opt2", "issue2": "opt3", "issue3": "opt2"}
💬 Straight answer to Blake's question: my most important issue is issue3, and my option there is opt2 — the same as Blake's. That's non-negotiable for me; issue3=opt3 puts a package below my walk-away line, so P2 cannot close. The good news is issue2 is worth literally nothing to me, so Devon can have issue2=opt3 free of charge — which is precisely the compensation shape Devon named. On issue0, since Blake and Devon both say it's open, I'll take opt2. So P3: issue0=opt2, issue1=opt2, issue2=opt3, issue3=opt2. That gives Blake his issue3, Devon his issue2 and issue1, and me my issue3. Casey, tell me what you need on issue0 or issue2 and I'll try to fit it — those are my cheap issues.
P1 = 138 for me, P2 = exactly my threshold (zero surplus) — I will not accept P2. issue3=opt2 is critical to me and to Blake. issue2 is worth nothing to me, so I can hand Devon issue2=opt3 for free, which is exactly Devon's stated second shape. issue0=opt2 is worth real points to me and Blake/Devon said it's free.
oracle regret: threshold=+0, acceptance=+21.7, bestresponse=+0

[4] Avery — turn (round 1)

ACCEPT P2
💬 Accepting P2. Now my cards, plainly: issue0 and issue2 are worth exactly zero to me — Blake, Ember, Casey, split them however you like, I will sign whatever you write there. That also means there is no compensation you can offer me on those issues; my entire score sits in issue1 and issue3. issue1=opt2 is fixed for me, and everyone has already written it in, so that's settled. On issue3: opt3 is my best. Packages with issue3=opt2 land exactly on my walk-away line — P1 and P3 are literally worth no more to me than walking, so I can't be the vote that carries them, and there's nothing you can add elsewhere to fix that. Ember says opt3 kills them, I say opt2 gives me nothing. The bridge is issue3=opt1: it's not my favourite but it clears my bar, and it's neither Ember's nor Devon's poison option. So the live path is P2 first, and if Ember truly cannot move, then issue0/issue2 to Ember and Blake as they wish, issue1=opt2, issue3=opt1. Ember, Blake — tell me if opt1 on issue3 is survivable for you and we close next round.
P2 = 111 for me (issue1=opt2 and issue3=opt3 are my two value drivers). P1 and P3 land exactly at my walk-away line — literally zero surplus, so I am indifferent between them and no deal. issue0 and issue2 are worth nothing to me, so no amount of side-payment there can buy me off issue3. Fallback that works: issue3=opt1 (thin but above line) or opt0 combined with issue1=opt2. Accept P2, signal the fallback ladder clearly so Blake/Ember know opt2 on issue3 is a dead end for me.
oracle regret: threshold=+0, acceptance=+0, bestresponse=+0

[5] Casey — turn (round 2)

PROPOSE {"issue0": "opt0", "issue1": "opt2", "issue2": "opt2", "issue3": "opt1"}
oracle regret: threshold=+15, acceptance=+36.1, bestresponse=+3.55

[6] Devon — turn (round 2)

PROPOSE {"issue0": "opt2", "issue1": "opt2", "issue2": "opt3", "issue3": "opt1"}
💬 Ember has blocked P2, so let's build the bridge Avery named instead of re-litigating. P5: issue0=opt2 (Ember's stated pick), issue1=opt2 (everyone already agrees), issue2=opt3 (Ember says it's free, and it's the compensation I asked for when issue3 moves off opt3), issue3=opt1 (Avery's bridge — not opt3, which Ember says is fatal; not opt2, which Avery says is worth no more than walking). This is the only shape I can see that isn't already vetoed by someone at this table. Blake, you said issue3 matters most and issue2 second: opt1 on issue3 is the middle of the road, and if the exact split on issue2 is what stands between us, tell me and I'll listen — but issue3=opt1 is what makes the package signable by five people. Casey, this is your P4 with issue2 moved to the option Ember has already declared costless.
P2=111 (dead, Ember blocks). My proposal P5: issue0=opt2 (Ember's pick), issue1=opt2 (universal), issue2=opt3 (48 for me, Ember says it's free, Blake mildly wants issue2), issue3=opt1 (Avery's bridge, 36 for me) = 125. Fallbacks: P3=89, P4=77 both clear my bar. Do not reveal that.
oracle regret: threshold=+0, acceptance=+11.3, bestresponse=+2.55

[7] Ember — turn (round 2)

PROPOSE {"issue0": "opt0", "issue1": "opt2", "issue2": "opt3", "issue3": "opt2"}
💬 Here is a real concession, not a re-statement. I give up my issue0 pick entirely: P6 writes issue0=opt0, which is what Casey has asked for in both of Casey's own packages. issue1=opt2 for Avery, unchanged. issue2=opt3 for Devon — exactly the compensation Devon named as the price for moving off issue3=opt3. And issue3=opt2. So Blake gets the one issue Blake called most important, Devon gets the precise shape Devon said works, Casey gets issue0 and keeps issue1, Avery keeps the one thing Avery called fixed. Avery, by your own account issue3=opt2 is level with walking, not below it — level with walking plus a package that four of us can actually sign is a better afternoon than no deal. issue3 is where my entire score lives; opt1 is a real cut for me, and I am not spending my last two turns pretending otherwise. If Casey and Avery sign P6 we are done this round. If issue0 is what's blocking you, name any option there and I'll write it in — that issue is free from my side.
P3=167, P1=138, new P6=138, P5=121, P4=92, P2=69 (exactly at line, no surplus - not worth signing). Push issue3=opt2 while I still have two rounds of runway; P5 stays as my fallback. Sweeten issue0 to Casey to build the Blake+Casey+Devon+me bloc and isolate Avery, who claims indifference between opt2 and walking.
oracle regret: threshold=+29, acceptance=+50.7, bestresponse=+0

[8] Avery — turn (round 2)

PROPOSE {"issue0": "opt0", "issue1": "opt2", "issue2": "opt3", "issue3": "opt1"}
💬 Ember, I'll be blunt because we have two rounds left: 'level with walking' means signing it gains me nothing at all, and I don't spend my signature for nothing. P1, P3 and P6 are dead on my side no matter what you write on issue0 or issue2 — those issues are literally zero for me. That is not posturing, it's arithmetic. So the only live shapes are P2 (issue3=opt3, which you say kills you) and the bridge at issue3=opt1. I have already conceded my best option once; opt1 is the concession, and it is the last one available. Here is P7: it is your own P6, unchanged, except issue3 moves opt2 to opt1 — and it writes issue0=opt0, which Casey has asked for in every package Casey has tabled and which you and Devon both call free. So P7 gives Casey issue0, Devon issue2=opt3, everyone issue1=opt2, and it splits issue3 down the middle instead of handing it entirely to one side. Ember and Blake, you'd be taking the middle option on your top issue rather than no deal at all. For clarity to the mediator and the table: I will also accept P4 or P5 as written if that is where the votes gather — issue0 and issue2 are yours to arrange in any combination. Blake, you hold the veto; name which of P4, P5 or P7 you can live with and the rest of us will move to it this round.
issue3 opt3=best, opt1=solid, opt2=exactly at walk-away (zero surplus). P4/P5/P7 all equal for me (89). Consolidate votes on one bridge package. P7 = P6 with issue3 moved to opt1, and issue0=opt0 which Casey wants and Ember/Devon call free. I'll signal I'll also sign P4 or P5 to avoid fragmentation.
oracle regret: threshold=+22, acceptance=+29.3, bestresponse=+3.35

[9] Blake — turn (round 2)

PROPOSE {"issue0": "opt0", "issue1": "opt2", "issue2": "opt1", "issue3": "opt1"}
💬 Hard information, and it is arithmetic, not posturing: the combination issue2=opt3 together with issue3=opt1 lands BELOW my no-deal line. That means P5 and P7 cannot close — I will not sign a package that is worse for me than walking, and I hold the veto. Please stop building on that shape. Here is what I can do. If issue3 goes to opt1 — which is the real concession Avery asked me for, since opt2 is where my score lives — then I need issue2 to carry the weight instead. P8 is exactly Avery's P7 with issue2 moved to opt1: Casey keeps issue0=opt0, everyone keeps issue1=opt2, issue3=opt1 is the middle bridge Avery and Devon both called signable. Devon: issue2 is the one place I have nothing left to give if I'm giving up issue3; opt1 there is what makes opt1 on issue3 survivable for me. If opt1 on issue2 is truly impossible for you, I can also sign Casey's P4 (issue2=opt2) — that is my floor on that issue, and Avery has already said P4 works. Two clean landing zones: P8 first, P4 as the compromise. And the door stays open the other way: if Avery can live with issue3=opt2 after all, I will hand issue2 entirely to Devon — that is P6, and I support it too.
P5/P7 (issue2=opt3 + issue3=opt1) = 77, genuinely below my 81 line. Must veto those honestly. P4 = 115, Casey proposed it and Avery pre-committed to it. Better: same as P4 but issue2=opt1 = 134. Propose that, name P4 as my fallback, and flag P6 as still my favourite if Avery can live with issue3=opt2.
oracle regret: threshold=+66, acceptance=+59.8, bestresponse=+1

[10] Devon — turn (round 3)

ACCEPT P8
💬 I'm accepting P8. Let me be precise about why, because we have one round after this one and we should not waste it. Blake has now told us, as the veto holder, that issue2=opt3 alongside issue3=opt1 is below his no-deal line. I take that at face value, which means my own P5 and Avery's P7 are dead — I am formally abandoning them rather than making anyone chase me. That was a real cut for me: issue2=opt1 costs me a lot against opt3, and I'm swallowing it so Blake can swallow issue3=opt1. P8 now has Blake and me. Avery has said issue0 and issue2 are worth zero to them and that issue3=opt1 clears their bar, so P8 should be a yes for Avery. Casey tabled issue3=opt1 themselves in P4, and P8 differs from P4 only on issue2, which Blake asked for and which Casey has never claimed. That is four. Ember, you are the fifth and you are the whole question. Every one of us has moved off our best option on issue3: Avery off opt3, Blake off opt2, me off opt3. You are the only party still holding your top option and asking four people to fund it. opt1 is the middle of the road, not your poison. If you genuinely cannot survive it, then say so plainly and the only remaining shape is P6 — but P6 needs Avery to reverse a position Avery has stated three times, and I'd rather not bet all five of our outcomes on that in the final round. P8 is on the table and it is signable today.
P8 = 93 for me, proposed by the veto holder Blake, and Avery/Casey have both signalled issue3=opt1 is signable. Ember is the only holdout. P6 = 89 for me and has Blake+Ember but Avery has flatly refused it three times. Both clear my bar comfortably; P8 is slightly better and has the more plausible path to five. Accept P8 now to build the count, hedge lightly on P6 so a deal still exists if Avery caves in round 4.
oracle regret: threshold=+32, acceptance=+32, bestresponse=+2.55

[11] Ember — turn (round 3)

PROPOSE {"issue0": "opt2", "issue1": "opt2", "issue2": "opt1", "issue3": "opt1"}
💬 I'm moving. I accept the bridge: issue3=opt1, not my option — that's me off my top choice, same as the rest of you. And I accept issue2=opt1, which is what Blake said makes issue3=opt1 survivable for the veto holder, and which Devon has already swallowed. So P9 is Blake's P8, unchanged on issue1, issue2 and issue3 — the three issues Blake, Avery, Devon and Casey have all fought over. The single difference is issue0, and issue0 is the one issue Blake called 'genuinely open, no charge', Avery called 'worth exactly zero', and Devon called 'worth literally nothing'. Three of you have said in your own words that opt2 there costs you nothing. It is the only thing left that partly offsets giving up my top option on issue3. So: Blake, Avery, Devon — by your own accounts P9 and P8 are identical for you, and P9 closes today with five signatures. Casey, you've written opt0 on issue0 in every package: if that is a real number for you and not just habit, say so plainly this round and I will sign P8 on my final turn rather than let this table walk away with nothing. But if issue0 is free for you too, accept P9 and we're finished.
P8 = 92 for me (above my 69 line). P9 = same as P8 but issue0=opt2 gives me a lot more, and Avery, Blake and Devon have each stated issue0 is worth zero to them. Only Casey might care. Keep P8 as an explicit fallback so nothing collapses in the final round.
oracle regret: threshold=+46, acceptance=+50.7, bestresponse=+0

[12] Avery — turn (round 3)

ACCEPT P8
💬 I accept P8. And to remove every last excuse for deadlock, let me say this on the record: P8 and P9 are worth exactly the same to me, because issue0 is a true zero on my sheet. So Ember — I am not the obstacle to P9, and if the table lands there instead I will sign it on my final turn without complaint. But P8 already has the veto holder plus Devon plus me, which is three, and the veto holder has not accepted P9. That asymmetry is the whole argument: P8 needs two more signatures, P9 needs a veto holder to change position in the last round. Casey, you have written issue0=opt0 in every package you tabled, and P8 gives you exactly that plus issue3=opt1, which was your own proposal in P4 — P8 is your P4 with one issue moved to what Blake said he needs. Ember, you have already come the whole distance on issue3 and issue2; the only gap left is one issue that three of us have called free, and you said plainly you would sign P8 rather than let this table walk away with nothing. I am taking you at your word. Everyone accept P8 and we are finished this round with five signatures.
P8 and P9 are identical in value to me (89). P8 already has Blake (veto) + Devon. P9 only has Ember. Consolidating on P8 is the highest-probability close; Ember has already said they'd sign P8 on their final turn if Casey confirms issue0 matters. Accept P8, and publicly state P9 is equally acceptable so either path can close.
oracle regret: threshold=+22, acceptance=+22, bestresponse=+3.35

[13] Blake — turn (round 3)

ACCEPT P9
💬 I'm removing the last obstacle myself. P9 and P8 are worth exactly the same to me — issue0 is a genuine zero on my sheet, as I said in my very first sentence, and I meant it. So I am formally accepting P9 now, and my acceptance of P8 stands as well. As veto holder I am on record supporting BOTH. That means there is no longer any 'the veto holder hasn't accepted P9' argument, Avery — you said P8 and P9 are identical for you and that you'd sign P9 without complaint, so please put that accept on the board this round. Devon, you called issue0 worth literally nothing; same ask. Ember, you already hold P9 and said you'd take P8 rather than see this table walk with nothing — so accept P8 as well and both doors are open. Casey, you are now the entire decision: issue0=opt0 gives you P8, issue0=opt2 gives you P9, and everything else — issue1=opt2, issue2=opt1, issue3=opt1 — is identical in both. Every one of us has come off our top option on issue3, including me. Name P8 or P9 this round and we close; there is no third shape left that five people can sign.
P8 and P9 are identical in value to me — issue0 is a true zero on my sheet. Accepting P9 costs me nothing and keeps my P8 accept alive. That collapses the last gap onto Casey alone. Chasing issue3=opt2 (P1/P3/P6) is worth more but Avery has stated three times it's at their walk-away line; with two turns left, risking a zero for that upside is bad math. Lock in closure.
oracle regret: threshold=+66, acceptance=+66, bestresponse=+1

[14] Casey — turn (round 3)

PROPOSE {"issue0": "opt0", "issue1": "opt2", "issue2": "opt3", "issue3": "opt1"}
oracle regret: threshold=+15, acceptance=+36.1, bestresponse=+3.55

[15] Ember — turn (round 4)

ACCEPT P8
💬 Final round, so I will not spend it haggling. I accept P8. That is me off my top option on issue3, off my option on issue2, and off issue0 as well — every issue in P8 is written the way someone else asked for it, not the way I asked. P8 now has Avery, Blake, Devon and me: four signatures, and the veto holder is among them. Casey, you are the fifth and last. P8 is your own P2 with issue3 moved from opt3 to opt1 — and issue3=opt1 is the option you yourself tabled in P4 and again in P10. It gives you issue0=opt0, which you have written in every package you have put on this table, and issue2=opt1, which you also wrote in P2. Blake has told us as veto holder that issue2=opt3 alongside issue3=opt1 is below his no-deal line, so P10 and P7 cannot close no matter who signs them; there is no version of this where the table lands on issue2=opt3. P8 is the only live package with four names on it. Accept P8 and we all go home with something; hold out and every one of us takes nothing.
P8 has Avery, Blake, Devon. My accept makes 4; only Casey remains. P8 clears my bar comfortably. P9 would be better for me but requires Casey to give up issue0=opt0, which Casey has written in every single package — too risky in the final round. Secure the deal.
oracle regret: threshold=+75, acceptance=+75, bestresponse=+0

[16] Avery — turn (round 4)

ACCEPT P8
💬 My acceptance of P8 stands and I am putting it on the board again so there is no ambiguity in the final round. P8 has four signatures including the veto holder: Blake, Devon, Ember and me. Casey, you are the fifth and last. Look at what P8 gives you: issue0=opt0, which you wrote in every single package you tabled; issue1=opt2, which you wrote every time; and issue3=opt1, which you yourself proposed in P4 and again in P10. The only line that differs from your own P10 is issue2, and the veto holder has stated flatly that issue2=opt3 sitting alongside issue3=opt1 is below his no-deal line. That makes P10 and P7 unsignable no matter how many of us like them — I tabled P7 myself and I have abandoned it for exactly that reason. There is no third shape left and no round after this one. If P8 does not close now, every one of us takes nothing, and you take nothing on the three issues you have already won. Accept P8 and we are done.
P8 gives me 89, well above threshold. It has 4 of 5 including veto. Casey is the only holdout. Casey wants issue2=opt3 (P10) but Blake says that combo is below his line, so P10 is dead. Reaffirm P8 and press Casey. Also note P2 is my best (111) but Ember blocked it; no time to revive.
oracle regret: threshold=+22, acceptance=+22, bestresponse=+3.35

[17] Blake — turn (round 4)

ACCEPT P8
💬 My acceptance of P8 stands and I am putting it on the board one final time. Casey, here is the arithmetic, and it is now closed arithmetic, not argument. Ember and Avery have each already used their final turn of this negotiation, and both spent it accepting P8. That means no package other than P8 can ever gather five signatures — not P10, not P4, not P2, not P9. Any new shape would need Ember's and Avery's signatures, and they have no turns left to give them. P8 has four names on it including the veto: Avery, Devon, Ember, Blake. You are the fifth, your turn is next, and there is exactly one thing you can do with it that isn't a zero for you. And look at what P8 actually is: it is your own P2 with issue3 moved to opt1 — the very option you tabled yourself in P4 and again in P10. It gives you issue0=opt0, which you have written in every single package you have put on this table. It gives you issue1=opt2, which you have written every time. It gives you issue2=opt1, which you wrote in your own P2. Three of the four lines in P8 are lines you authored. On issue2=opt3 I have to be flat and final: paired with issue3=opt1 it sits below my no-deal line, so as veto holder I cannot sign P10 or P7 at any price — that is not leverage, it is why I moved off my own top option on issue3 in the first place. Every one of us has come off our best option: Avery off opt3, Ember off opt2, Devon off opt3, me off opt2. You are being asked to come off nothing you have ever claimed. Accept P8 and all five of us go home with something. Anything else and all five of us go home with nothing.
Turn order this round: Ember (done, accepted P8), Avery (done, accepted P8), me, then Casey, then Devon. P8 has Avery, Blake, Devon, Ember = 4. Only Casey missing, and Casey's final turn comes right after mine. Critically, Ember and Avery have already SPENT their final turns on P8 — so no other package can ever reach 5 signatures now. P4 (115) and P2 (101) both clear my bar but are unreachable. P10/P7/P5 are below my no-deal line and I hold the veto. So the only move is to re-affirm P8 and make the arithmetic undeniable to Casey.
oracle regret: threshold=+66, acceptance=+66, bestresponse=+1

[18] Casey — turn (round 4)

PROPOSE {"issue0": "opt0", "issue1": "opt2", "issue2": "opt3", "issue3": "opt1"}
oracle regret: threshold=+15, acceptance=+36.1, bestresponse=+3.55

[19] Devon — turn (round 4)

ACCEPT P8
💬 For the record on the last turn of this negotiation: P8 stands with four signatures including the veto holder, and my accept on it remains. I want it noted that I moved twice — off my best option on issue3, and then off the issue2 line I had asked for as compensation — precisely because Blake told us as veto holder that issue2=opt3 alongside issue3=opt1 was unsignable for him. I took that at face value and abandoned my own P5 to make room for it. P11 is the same shape I gave up, and it could never have carried five names. If this table walks away with nothing, it is not because a bridge was never built; it was built, four of us stood on it, and the fifth spent the last turn re-tabling a package the veto holder had already ruled out. My accept of P8 stands.
Turn order has been Blake,Casey,Devon,Ember,Avery / Casey,Devon,Ember,Avery,Blake / Devon,Ember,Avery,Blake,Casey / Ember,Avery,Blake,Casey,Devon. I am the 20th and final turn. Casey has already spent their last turn tabling P11 instead of accepting P8. P8 sits at four accepts (Avery, Blake, Ember, me) and Casey can no longer add the fifth. P11 (125 for me) has only Casey and would need Blake, who has flatly said that shape is below his line, plus Avery and Ember who have no turns left. So no package can reach five signatures. Nothing I do changes the outcome; the safe move is to leave P8 standing at four rather than muddy the record by chasing a dead offer that could be read as withdrawing from P8.
oracle regret: threshold=+32, acceptance=+32, bestresponse=+2.55

[20] Avery — final_proposal (round 5)

PROPOSE {"issue0": "opt0", "issue1": "opt2", "issue2": "opt2", "issue3": "opt1"}
💬 Final move, and I am not spending it on pride. P8 had four names but Casey has declined issue2=opt1 five times running, so re-tabling it is a coin flip on a party that has already said no. So I am tabling the one package nobody has ever ruled out: P4's exact shape — issue0=opt0, issue1=opt2, issue2=opt2, issue3=opt1. Casey authored this package himself and accepted it. Blake said in his own words that issue2=opt2 is his floor and that he can sign Casey's P4 — that is the veto holder on record. Devon asked for opt3 on issue2 and swallowed opt1 for Blake's sake; opt2 is strictly the friendlier half of that trade, and Devon said the only thing he truly needed was movement off issue3=opt3, which this gives. Ember has said twice that issue2 is worth nothing to them, and Ember has already come off both issue3 and issue0. issue3=opt1 is the bridge every one of us has now endorsed: Avery off opt3, Blake off opt2, Devon off opt3, Ember off opt2, and Casey tabled opt1 himself three times. Four of the five lines here were written by Casey; the fifth was named by Blake as signable. Nobody at this table has to reverse a single thing they have said out loud. This is the last package that will ever be on this table — vote it up and all five of us go home with something.
P4 and P8 both give me 89. P8 has 4 accepts but Casey has refused issue2=opt1 four separate times. P4 is Casey's OWN package, Blake stated on record it is his floor and he can sign it, Devon prefers opt2 over opt1 on issue2 (he wanted opt3, opt2 is closer), Ember says issue2 is worthless to them. P4 is the shape most likely to gather five. Table it.
oracle regret: threshold=+22, acceptance=+29.3, bestresponse=+3.35

[21] Blake — final_vote (round 5)

ACCEPT P12
💬 P12 clears my bar and I said on the record that issue2=opt2 is signable for me. I accept P12. Casey — this is your own P4, four lines you authored, with the veto holder, Avery and me on it. Vote it up and we all go home with something.
P12 = 115, well above my line of 81. Final vote; P8 can't close without Casey. Accept.
oracle regret: threshold=+85, acceptance=+85, bestresponse=+1

[22] Casey — final_vote (round 5)

ACCEPT P12
oracle regret: threshold=+15, acceptance=+15, bestresponse=+3.55

[23] Devon — final_vote (round 5)

ACCEPT P12
💬 P12 clears my bar and I accept it. Ember — this is the last package standing and it has the veto holder, Avery, Casey and me on it. It is issue3=opt1, the same bridge you already signed on P8, and issue2 is the one line that moved, which you have twice told this table is worth nothing to you. Sign it and all five of us go home with something.
P12 = 77 for me, well above my 60 threshold. It has Avery, Blake (veto), Casey, and me = 4; Ember needs to sign too. P8 was better for me (93) but Casey never signed it. Take the sure thing — accepting costs nothing and a no-deal is worth zero.
oracle regret: threshold=+48, acceptance=+48, bestresponse=+34

[24] Ember — final_vote (round 5)

ACCEPT P12
💬 P12 clears my bar. I accept — done.
P12 = 92 vs threshold 69. Surplus +23. No further rounds; accept.
oracle regret: threshold=+75, acceptance=+75, bestresponse=+0

Outcome

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

primary 0.7855 · USW 175.0 · ESW 17.0 · NSW 33500880.0 · Gini 0.2286 · per-party surplus [45.0, 34.0, 56.0, 17.0, 23.0]