The Third Dynasty of pokes
23 November 2018 - 20 January 2019
- Some relevant selections from the obituary:
- Morgan Chamberlain IV, Connecticut financier, passed away on the 22nd of November, after a brief illness. They were 83.
- Shortly after a string of high-profile business successes in the late 1980s, Morgan retired from public view and began a reclusive life. Public records show that Morgan remained in Connecticut for the rest of their life, but specifics of their whereabouts remained unknown outside of the Chamberlain family.
- Dr. Chamberlain is survived by a spouse, Casey, of four years; eight children: Morgan, Benjamin, Elizabeth, Isabel, James, Joshua, Anne, and Robert, all from previous marriages, and twelve grandchildren.
- Meanwhile, within the newly-reconstructed City Hall: Case number 2018-1022, regarding the estate of Morgan Chamberlain IV, begins to be heard. The courtroom is already at capacity with lawyers representing parties who would love to materialize any claim they have on the sizeable fortune of the deceased.
- Replace “Activist” with “Attorney” and “Veteran” with “Judge”.
The following players were active at the start of the Dynasty:
Brendan*, derrick*, Kevan*, pokes*, Trigon
The following players were active at the end of the Dynasty:
edelopo, Kevan*, pokes*, Trigon, Zaphod
Summary of Final Ruleset
Cases were built to imitate real-life lawsuits. First, the Judge would open the case, selecting Clients to be the Plaintiff and the Defendant, and Private Messaging all Attorneys that represented each. From there, any such Attorney was encouraged to post comments describing what they did in the case using voting icons. Finally, after the arguing had occurred, the Judge would [[Ruleset 163#Closing Cases|close the case]. The least guilty party would win, importantly causing the Client to transfer a portion of their winnings to the Attorneys representing them.
This cycle of opening, arguing upon, and closing cases ended up being the primary feature of the dynasty, as the victory condition involved gaining $100,000.
Posts of Interest
Proposals and CFJs
- The Clientele started the dynasty off with Clients, which would grow into the main draw of the dynasty. Each Client had a Worth (a dollar value) and a Reputation. Clients could be represented by Attorneys. This would be done by making requests to the judge.
- Habeas Corpus reworded the clients rule completely and fleshed out the requests mechanic, allowing Attorneys to choose a set of clients they would like to represent and having the Judge choose secretly from that set.
- Court Is In Session allowed the Judge to create, but not yet resolve Cases, the main mechanic by which money would shift between Clients and Attorneys.
- Business as Usual added four businesses to the pool of clients.
- The Guilty Party put more detail into the Cases mechanic. It gave Clients villainy scores related to their reputations. For each case, it gave Clients guilt, a variable between 0 and their villainy.
- Lawyer Up gave Attorneys two stats: money and integrity. Attorneys could Set up Shop by setting their money and integrity such that the sum of their integrity and their money divided by 1,000 did not exceed 10.
- Slush Puppy gave Attorneys the last of their stats: their Slush Funds, tracked privately by the judge.
- Gavelogue finally completed how Cases would work. Each Attorney could vote on cases. FOR icons decreased their client's guilt by two. The winning party would take a sum of money from the loser and give a portion to the attorney.
- I'll Allow It added underhanded tactics. Players voting with AGAINST icons on a case would now decrease their client's guilt by one at the cost of one integrity.
- Better Call Saul added consequences for having low integrity. Now, Clients could mistrust their Attorneys.
From this point on, most proposals are just quick afterthoughts, as gameplay shifted to the cases.
- Case Adjourned raised the stakes by making each case only able to have Clients represented by Attorneys in them. The first case to adhere to this rule was Case 9.
- Laundry Day allowed players to launder money in their Slush Fund, giving 90% of it back to them.
- Who Was That Masked Barrister? did away with some of the secrecy and inserted a list of representations into the ruleset.
- Noisy Withdrawal allowed dropping of Clients to be done by the Attorneys themselves.
- Open Registration finished the work of the previous two and allowed the acquisition of clients to be done by the Attorneys as well.
- The Winner Takes It All instituted a win condition -- earning $100,000.
- Out to Dry gave Attorneys one last reaction image for cases: VETO would absolutely throw the case.
- The burden of proof made the fines when defendants won the case $0. The first case to adhere to this rule was Case 24.
- Digging Dirt gave a change to underhanded tactics: instead of decreasing a Client's Guilt by 1 (which was useless if both Clients were already at zero), it increased the other Client's guilt by one.
- Proven Guiltier allowed guilt to be between 0 and 2 times the Client's villainy to raise stakes.
- Time Is Money added strategy by causing FOR icons to cost $2,000 to the Attorney.
Similar to the missile crisis dynasty, this dynasty had a sizable component of me doing relatively complicated things on secret information. This made me nervous then and now about accidentally throwing the game by handling these steps incorrectly. Near the beginning of the dynasty I began implementing the opening and closing of cases as a computer program, to find out if this would be a net positive or negative for my stress of handling these steps. Hopefully, less stress. (At least compared to the missile crisis dynasty, this had almost no opportunities for wording-based scams that would then require a decision about whether to CfJ the scam, unfairly blowing it up if it would be considered legitimate but CfJed, and unfairly allowing it to go on if it's not legitimate and not CfJed.)
I think that, overall, it was a good thing to have the programmatic implementation. There were some pros:
- The main pro is that it made opening/closing cases much faster, and made it so that I didn't have to re-understand the list of steps 38 times, and could avoid a significant amount of double- or triple-checking myself.
- If the steps the code took passed a human double-check, I had much more confidence that they were done correctly than if I had done two human checks.
- It was fun.
and some cons:
- Manual interfacing of numbers with the GNDT and ruleset: I had a local database of gamestate that needed synchronization with the gamestate here. I think there were only a few mismatches, none significant, but that there might have been was an ongoing concern. In some cases, it helped: most recently, I had entered Kevan's money incorrectly, but the local database was still correct. So, on the next update, when I went to type it in, my old typo got fixed without me even being aware of it - Kevan was the first human to figure out what happened.
- Manual translation of rules into code: each new enacted proposal also required me to update the code accordingly. Thankfully nothing was that hard, but if something that would have been hard to code did come through the proposal pipeline, I'd be on the hook for implementing it.
- Double-checking was required anyway. I never really fully trusted the code, but it gave me enough output to check that the steps were being done correctly.
If I were to do it again: I would use some construct that can log each individual change to the database, like the GNDT does, so I can review it in more detail. I made use of version control to keep some snapshots of my local gamestate. But VC still needs me to choose when to snapshot, which is less granular. Thankfully I think I was still able to catch all the errors, but it would have been easier to do the forensics if I had a GNDT-like view of the changes.
Overall, even with the support, I was often too burnt out or low on free time after closing and generating cases to author many significant proposals. I had a weak excuse in that proposals coming from me while having secret information could be biased, or seen as being biased, consciously or unconsciously. --Pokes (talk) 14:46, 19 January 2019 (UTC)
For the curious, I'll document the code here: The Third Dynasty of pokes: helper code
As mentioned on Slack: I'll defend burden of proof as (a) increasing speed: the average fine increased (b) decreasing RNG effect: otherwise the first successful defendant against Connecticut gets an instant win. Without burden of proof the winner would have been Zaphod, earning over a million dollars for case 36. --Pokes (talk) 14:17, 20 January 2019 (UTC)
- I will admit that The Burden of Proof did some good for the game, especially with cases involving States. However, I guess I was just a little peeved that I lost money on Cases 29 and 33 when I was the winner. Case 34, on the other hand, was totally my fault. Without BoP, though, I would've gotten $11,739 from Bananasoft and then $14,562 from Goldberg Tech, which would have made me an actual competitor at about $60,000.
- But I guess hindsight is 20/20. I've learned to actually pay attention to the ruleset if I want to win. Trigon (talk) 21:35, 20 January 2019 (UTC)
- Burden of Proof definitely sped the game up, but it feels like it added more randomness than it removed - it felt very harsh as a player that if you lost the 50/50 coin toss for Prosecution/Defence, you made zero money that round while others got ahead. And tripling the payouts just reduced the number of times someone had to be lucky to hit $100,000.
- I think if we had ended up in a situation where the Defendant looked like they had a 50/50 chance of winning a million-dollar State case tied at 1 Guilt, proposals would have come into play to stop that from happening. (I had vaguely assumed that the State would always win because it had multiple Attorneys and they could make multiple AGAINST motions, but looking at the rule that's not how it actually worked.) --Kevan (talk) 12:28, 21 January 2019 (UTC)
Well, that was an interesting dynasty. Like most, there was lots of ironing out to be done before we got to some actually serious gameplay. I have a history of not paying attention to dynasties until it is far too late for me to really be successful, so I didn't really get to experience the way the game was played before I started representing Benjamin.
Drawing inspiration from pokes' commentary, here's what went right with the dynasty:
- Fun central mechanics (cases, clients, etc.).
- Repeatable gameplay, relatively easy for attorneys to pick up. Additionally, it didn't get boring, even over it's long runtime (this dynasty lasted over a month!).
- Many, many opportunities for fun flavor text. If you didn't play this dynasty, read through the cases. Some of the situations are just wild.
and here's what went wrong:
- There wasn't much to do for anyone but the Judge. I stated that it was easy to pick up? Well that's because it was dead simple. Some of the proposals at the end of the dynasty tried to add a bit more strategy to the cases mechanic for Attorneys, but even then it wasn't all that complicated. Here's hoping the next dynasty has a little more strategy.
- This isn't anyone's fault, but the RNG was just vicious. For example, cases 16 to 23 were nearly all Elizabeth.
- The burden of proof slowed gameplay to a crawl. Like I said, it was still fun, but I think by the upper 30s of case numbers, we were all feeling a little ready to move onto the next dynasty.
I also left with a sour taste in my mouth because my scam didn't work.
- Ah, I just didn't want it to sound like I'd won by doing anything clever. --Kevan (talk) 18:09, 21 January 2019 (UTC)
It did feel like tactical gameplay never really got going here - once Out to Dry enacted I was expecting more bribery to throw cases (Defendants with stronger Clients offering to throw the case in exchange for half the payout), but that never seemed to happen. To some extent Burden of Proof should have helped that (as Defendant you can only ever win $0, so there's a big incentive to negotiate) but the two offers I made privately to other players were rejected. --Kevan (talk) 12:28, 21 January 2019 (UTC)