talonschild
Drive-By NESer
That's why my personal opinion is that Ministers will use polls - complete or no - to inform their judgment. Sometimes decisions need to be made. If you don't like it, vote 'em out. But give Ministers some room to act.
Team CFC Constitution
- Thread starter talonschild
- Start date
Ah, the whole "will of the people" argument. 
Polls are needed when there is a choice to be made with more than one reasonable option, and it is not clear from discussion which way the team leans. As I previously posted (in this thread?) there are times when an expert needs to just handle it, and times when we need a strong consensus.
We should have a running debate on grand strategy, starting off with aspirations which we may or may not be able to realize depending on the map and other factors. This needs to be clear from the beginning because there are things we should do in the first 10-20 moves depending on the overall strategy.
Whether to go to war or not, who to be allies with, whether to trade resources -- those are significant actions where we should be on the same page to the extent possible. City placement, civic and religion switches, and similar are also important. Tech path should be decided by strategy. With tech trading off, it is more important than ever to beeline as much as possible. Early imbalances can be killer in the right circumstances.
If we end up polling mundane stuff that ties the turnplayer's hands, then we've made a mistake.
If we have citizens who want to accentuate the democracy aspect of the game, there are a plethora of non-game things that can absorb a lot of effort and result in a lot of fun. Things like city and unit names can give comic relief, or send subliminal messages.
One interesting democratic system I see used a lot is the organization of speed vote chess on chess.com. A move must be made every 10 minutes (20 minutes total think time). We have team captains whose main responsibility is to organize the input of all the players and call the consensus move. A lot of the time the captain isn't the highest rated player, it's someone who can listen to the arguments for moves by the high rated players and deduce whose move is more convincing. The teams I'm on that use this system win a lot more games than the ones who don't, even when the opponents who don't use the system have higher rated players.
Here, a similar concept might be carried by a President or Prime Minister who has control of the decision making process, deciding what issues can be handled via looking at the preponderance of comments on one side or the other and which need votes. Unlike RL examples, this leadership position might not have any independent decision making power, or at least might be restricted from choosing any option which doesn't have other supporters.
This rambled a bit and I'm sleepy. Hopefully it's coherent enough to be worthwhile.
Polls are needed when there is a choice to be made with more than one reasonable option, and it is not clear from discussion which way the team leans. As I previously posted (in this thread?) there are times when an expert needs to just handle it, and times when we need a strong consensus.
We should have a running debate on grand strategy, starting off with aspirations which we may or may not be able to realize depending on the map and other factors. This needs to be clear from the beginning because there are things we should do in the first 10-20 moves depending on the overall strategy.
Whether to go to war or not, who to be allies with, whether to trade resources -- those are significant actions where we should be on the same page to the extent possible. City placement, civic and religion switches, and similar are also important. Tech path should be decided by strategy. With tech trading off, it is more important than ever to beeline as much as possible. Early imbalances can be killer in the right circumstances.
If we end up polling mundane stuff that ties the turnplayer's hands, then we've made a mistake.
If we have citizens who want to accentuate the democracy aspect of the game, there are a plethora of non-game things that can absorb a lot of effort and result in a lot of fun. Things like city and unit names can give comic relief, or send subliminal messages.
One interesting democratic system I see used a lot is the organization of speed vote chess on chess.com. A move must be made every 10 minutes (20 minutes total think time). We have team captains whose main responsibility is to organize the input of all the players and call the consensus move. A lot of the time the captain isn't the highest rated player, it's someone who can listen to the arguments for moves by the high rated players and deduce whose move is more convincing. The teams I'm on that use this system win a lot more games than the ones who don't, even when the opponents who don't use the system have higher rated players.
Here, a similar concept might be carried by a President or Prime Minister who has control of the decision making process, deciding what issues can be handled via looking at the preponderance of comments on one side or the other and which need votes. Unlike RL examples, this leadership position might not have any independent decision making power, or at least might be restricted from choosing any option which doesn't have other supporters.
This rambled a bit and I'm sleepy. Hopefully it's coherent enough to be worthwhile.
Majic
Warlord
Polling is a way to make less vocal members feel involved and it worked out fine in AMAZON afterall. However I agree with not having everything decided by a poll, it takes time and makes us less flexible.
Trying to adress few points about voting here rather than in Starting Location thread. First a disclaimer: I do not support heavy polling. However, in important non-urgent decisions I think it is prudent to clarify the decision with a vote unless there is an unanimous decision or absolute majority of active team members, not just the ones voicing their opinion in the thread, can be clearly identified from the discussion.
I am fully in favour of trying to reach consensus through discussion but if there are several differing opinions in the thread, there are two major issues that interpreting votes from the discussion has. I'll use Starting Location discussion as an example, though at the moment we seem to be very close to absolute majority so IMO, it might not be necessary to have an official vote on that after all.
First issue is that it is very difficult to assess opinions correctly from a lenghty discussion. E.g. in Starting Location thread the first assesment did not actually reflect the count at the time very well. Also from the current count at least Slaze's opinion seems to be missing. Another problem in the post-hoc analysis of the discussion is that some of the opinions are rather old and made before majority of the simulation results were posted. A clear cut voting period after sufficient discussion usually reflects the end result of the discussion better.
Second problem that drawing conclusions about discussion is that it may well not reach a consensus but is in the end just another way of finding out simple majority just without an official poll. Think of the count kept in the Starting Location thread. If we don't end up with absolute majority (which it is starting to seem we will), what we have in our hands is a count for simple majority. (Not saying that it we shouldn't keep tabs. It helps to see where we stand currently. Thanks, Yossarian.)
Now to the alternatives. We don't have to rely on simple majority polls (IMO we shouldn't). There are several much better alternatives. All of them obviously have their weaknesses, but I think we can manage. For those interested in maths of different voting, check this comparison table in Wikipedia. IMO, most suitable method for us would be modified Borda count. Reposting the short explanation to the thread the discussion really belongs:
)
I am fully in favour of trying to reach consensus through discussion but if there are several differing opinions in the thread, there are two major issues that interpreting votes from the discussion has. I'll use Starting Location discussion as an example, though at the moment we seem to be very close to absolute majority so IMO, it might not be necessary to have an official vote on that after all.
First issue is that it is very difficult to assess opinions correctly from a lenghty discussion. E.g. in Starting Location thread the first assesment did not actually reflect the count at the time very well. Also from the current count at least Slaze's opinion seems to be missing. Another problem in the post-hoc analysis of the discussion is that some of the opinions are rather old and made before majority of the simulation results were posted. A clear cut voting period after sufficient discussion usually reflects the end result of the discussion better.
Second problem that drawing conclusions about discussion is that it may well not reach a consensus but is in the end just another way of finding out simple majority just without an official poll. Think of the count kept in the Starting Location thread. If we don't end up with absolute majority (which it is starting to seem we will), what we have in our hands is a count for simple majority. (Not saying that it we shouldn't keep tabs. It helps to see where we stand currently. Thanks, Yossarian.)
Now to the alternatives. We don't have to rely on simple majority polls (IMO we shouldn't). There are several much better alternatives. All of them obviously have their weaknesses, but I think we can manage. For those interested in maths of different voting, check this comparison table in Wikipedia. IMO, most suitable method for us would be modified Borda count. Reposting the short explanation to the thread the discussion really belongs:
- Everyone lists as many candidates as they like in order of preference.
- When voting has ended every voters first preference gets 1 point, second preference gets 0.5 points, third one third of a point and so forth.
- All the points are summed and one with most points is the winner.
)In discussion-based decision, there is no way to gauge non-posters position. Especially if the conclusion is drawn from opinions stated in a surprising place (like a discussion in an espionage thread). People could just have missed that there is a decision process going on. In voting based decision as long as the poll has fair (more than 24 h if possible) and clear timeframe, I think it is fair to count those as empty votes.
Coming up with a plan and following it or adjusting it as needed is a better approach than getting too focused on counting the votes of the majority.
I guess this is as good a place/time as any to mention that winning the game is sort of a microcosm of winning a war in the game. There are 3 things you want to try to out do your opponent(s) in:
1. Micro - Squeezing the most possible out of your Workers and tiles to yield maximum benefits. In War this is mostly the battle of Production, ie trying to get more/better units in the field as quickly as possible.
2. Diplo - This is more complex, but it is basically always being first to set up treaties, agreements & alliances that enhance your opportunities and limit your opponents opportunities. In War this is getting a good dogpile going against your enemy to make defeating them easier and faster.
3. Cerebro - This is the most complex IMO, but also the most important. We have to realize that even the best micro (or tactics in War) is not what gets you ahead of the opponent. Our opponents will be able to maximize their micro as well. The way you get the advantage is by suprising your opponent, tricking him, outsmarting him. We must fool our opponents into wasting time, resources etc. In War this means feints and surprise manuvers.
I mention all this because we should start thinking overall strategy to inform decision making. Saying "Lets have a Vote" without a grand plan risks making us go around in circles... "Vote go this way" then "Vote to go a different way" then "No back to this way!" and so on. Lets talk grand strategy, then decide the best path to follow it.
I guess this is as good a place/time as any to mention that winning the game is sort of a microcosm of winning a war in the game. There are 3 things you want to try to out do your opponent(s) in:
1. Micro - Squeezing the most possible out of your Workers and tiles to yield maximum benefits. In War this is mostly the battle of Production, ie trying to get more/better units in the field as quickly as possible.
2. Diplo - This is more complex, but it is basically always being first to set up treaties, agreements & alliances that enhance your opportunities and limit your opponents opportunities. In War this is getting a good dogpile going against your enemy to make defeating them easier and faster.
3. Cerebro - This is the most complex IMO, but also the most important. We have to realize that even the best micro (or tactics in War) is not what gets you ahead of the opponent. Our opponents will be able to maximize their micro as well. The way you get the advantage is by suprising your opponent, tricking him, outsmarting him. We must fool our opponents into wasting time, resources etc. In War this means feints and surprise manuvers.
I mention all this because we should start thinking overall strategy to inform decision making. Saying "Lets have a Vote" without a grand plan risks making us go around in circles... "Vote go this way" then "Vote to go a different way" then "No back to this way!" and so on. Lets talk grand strategy, then decide the best path to follow it.
On past teams we had a simple policy on non-posters. "Silence means Consent" as in if you don't speak up when the decision is being made, that is the same as giving your approval on the decision once made... No coming in after the decision is made and trying to overturn the decision, critisizing, complaining etc. That does not mean we have to adopt such a policy here, but that was what has been the custom in the past.
As I've said in several accounts, I fully agree that most of the time voting is not the way to go. However, we will eventually have to vote. Why I've brought this up now is that if we don't agree now what method we'll use when that time comes we run a risk of a) wasting time on deciding the method at that time, b) people feeling they did not have a fair say in the matter due to the way the the vote was organized. This can be avoided by deciding rules now when we still have ample time. Aside from the voting method we should also agree when/why votes are called for, who sets up the election and what is sufficient time limit.
Now for the most actions, IMO our aim should be to minimize time our decision cycle takes (OODA loop for those familiar with military planning, PDCA in industry). Length of the decision cycle has determined battles and even wars in real world. Since Civ does not happen in real time, absolute time may not be quite as essential. I've been trying to think ways how to achieve that, but everything I can come up with boils down to that, for any possible decision, we need to practically always have someone available who has mandate to make immediate decision on the matter if need be. E.g. only way I can think of enables us to react sufficiently fast when our random scout spots a SoD obviously heading for us is to actually give someone right to DoW on the spot if needed. Downsides are obvious. This would put an awful lot of responsibility on few people and gives a lot of power in their hands. The power part can be mitigated in several ways if required. Responsibility, not so much.
OT: Cerebro sounds like what college students could call each other in a geek fraternity
Now for the most actions, IMO our aim should be to minimize time our decision cycle takes (OODA loop for those familiar with military planning, PDCA in industry). Length of the decision cycle has determined battles and even wars in real world. Since Civ does not happen in real time, absolute time may not be quite as essential. I've been trying to think ways how to achieve that, but everything I can come up with boils down to that, for any possible decision, we need to practically always have someone available who has mandate to make immediate decision on the matter if need be. E.g. only way I can think of enables us to react sufficiently fast when our random scout spots a SoD obviously heading for us is to actually give someone right to DoW on the spot if needed. Downsides are obvious. This would put an awful lot of responsibility on few people and gives a lot of power in their hands. The power part can be mitigated in several ways if required. Responsibility, not so much.
OT: Cerebro sounds like what college students could call each other in a geek fraternity
cav scout
The Continuum
- Joined
- Apr 4, 2008
- Messages
- 2,630
As long as our decision cycle gives our turn player enough time to play we are fine. We can't really get inside our rivals' decision cycles though as this is a turn based game...
talonschild
Drive-By NESer
The pedant in me says you meant "macrocosm", Sommerswerd.
Our decision-making time may still matter in simultaneous turns sections.
And I think that we should be able to conclude this discussion within two days.
Our decision-making time may still matter in simultaneous turns sections.
And I think that we should be able to conclude this discussion within two days.
Yes macrocosm fits better thanks
As Sommers said, this game is not just about city/unit management. And even the empire management benefits greatly from faster decision cycle. E.g. we spot a SoD, our ally gets DoW'd or we find out that we're in a settler race and we either settle this turn a tile away from our original plan or lose the position completely. All of those situations and others I can't even imagine now would require decision cycle maximum of few hours. Plus diplomacy and outsmarting our enemies happen very much in real time and our response time there can be of critical importance.
This is an excellent point and related to what I'm trying to say here. Currently our decision cycle in organization issues is frankly put sluggish. This team is a bit over a month old (counting from first post in private forums) and we still don't have our working agreement/constitution. It is not clear, who will be our officials. We haven't even decided on what official roles we do want to have with the team. We aren't discussing our general strategy in organized fashion. If the game is about to start within a week, we can't afford to get caught pants down.
I'm not saying we haven't made any progress. We have. We have a nice prototype of turn logging application, we have nice discussions going on starting location and the ruleset and we have many volunteers for several roles. We have achieved this through people stepping up and taking up the tasks and doing something about them. But in the process we have postponed tasks that need to be done before the game starts. And the amount of tasks is not going to let up when it starts, on the contrary. Now, we can continue to do the tasks in self organizing fashion if we agree to and if we organize ourselves in that fashion. (Agile multiplaying anyone? Sorry, SW Engineering reference
) The other option is to agree on organization structure with named responsibles for various tasks. How we want to go about this is up to us to decide. And I think it is high time to reach some decisions on the organization and decision making issues.Now, please excuse me. I'm going to do something concrete about pending tasks.
I have one open question about current vote on capital location. What do we do with previously voiced opinions if everyone does not have time to vote due to tight deadline before turn flip? Do we count them as best we can or do we only count given votes? Though I had the gall to set up the vote and even force my favourite voting system, this is as far as I'm willing to go. I just don't have the guts to either ignore or count the previous opinions.
YossarianLives
Deity
I also have no idea how we should count those who had previously voiced an opinion, but do not vote in the new thread before the deadline. That's a tough call. In case you want it for reference, this was my latest tally, taken before a lot of the persuasive arguments that made a number of people reconsider their votes:
Yossarian, I've updated the tally of people's expressed preferences since that last count was made. It's clear that there's been a big jump in the polls for the banana spot, but there's no longer a clear majority. My count is based on opinions in the starting location and turns 1-50 thread since your last tally:
In place: 6 or 7? Sommerswerd, Bistrita, Tobaism, Talonschild, Cavscout, Caledron, Arkipeller?? (maybe changed his mind in his final post in the turn 1-50 thread but I'm not certain)
PH: 4 or 5? YossarianLives, Bowsling, Socralynn, Jovan Kukic?? Whb
Banana: 7 Vransam, Bcool, RegentMan, Grant2004, Slaze, DaveShack, Aivoturso
Aivoturso, I think we have too much ambiguity to use this tally to change the settlement decision. First there are a few players we're unsure of the first choice on. Then after that we aren't sure about 2nd or 3rd choices, and if the players would even choose to vote them given the strategy of the Borda count.
I'd suggest that because the tally is tight and ambiguous that the most fair way to make the decision is to trust the results of the poll. We may get a reprieve here and be able to extend the voting period to allow greater participation if the re-load request is approved, but we shouldn't be counting on that.
For the future I think we can have an interesting debate about the preferred polling method. I like the Borda count in situations where the number of voting options is fixed, and you have to vote for each one. This eliminates the tactics of whether you should vote for a 2nd choice or not. In cases where the voting options are not fixed (who would have imagined a vote for the north pole?) I prefer runoff polling, even if we need to declare our 2nd choices immediately to avoid multiple rounds of voting. In this way a person with a 1st choice out of left field isn't penalized in the final tally. That said I agree 100% with the use of the Borda count for this decision. We need to make a call, and I'm happy to see someone step up to make it happen.
In place: 6 or 7? Sommerswerd, Bistrita, Tobaism, Talonschild, Cavscout, Caledron, Arkipeller?? (maybe changed his mind in his final post in the turn 1-50 thread but I'm not certain)
PH: 4 or 5? YossarianLives, Bowsling, Socralynn, Jovan Kukic?? Whb
Banana: 7 Vransam, Bcool, RegentMan, Grant2004, Slaze, DaveShack, Aivoturso
Aivoturso, I think we have too much ambiguity to use this tally to change the settlement decision. First there are a few players we're unsure of the first choice on. Then after that we aren't sure about 2nd or 3rd choices, and if the players would even choose to vote them given the strategy of the Borda count.
I'd suggest that because the tally is tight and ambiguous that the most fair way to make the decision is to trust the results of the poll. We may get a reprieve here and be able to extend the voting period to allow greater participation if the re-load request is approved, but we shouldn't be counting on that.
For the future I think we can have an interesting debate about the preferred polling method. I like the Borda count in situations where the number of voting options is fixed, and you have to vote for each one. This eliminates the tactics of whether you should vote for a 2nd choice or not. In cases where the voting options are not fixed (who would have imagined a vote for the north pole?) I prefer runoff polling, even if we need to declare our 2nd choices immediately to avoid multiple rounds of voting. In this way a person with a 1st choice out of left field isn't penalized in the final tally. That said I agree 100% with the use of the Borda count for this decision. We need to make a call, and I'm happy to see someone step up to make it happen.
vranasm
Deity
just wanted to point out that the new voting process works reasonable only if there are X candidates and everyone has to vote for all and just order them.
The problem here is that in free voting system it's easy manipulate, let's take the capital decision.
If I vote bananas, SIP. I gave 1 to bananas, 0.5 SIP, 0 to PH
someone else votes SIP, gives 1 SIP, 0 bananas, 0 PH
best strategy if you want win of your option is to not vote for other options since then you don't give another vote advantage.
It feels like competing while trying to figure out the majority of votes.
I certainly could just say Bananas and be finished, but would be a bit unfair to people saying SIP only (since they obviously don't want win for any other place, they choose best strategy how to achieve this).
Eventually people will grasp this concept and the consensus will be on plain majority since everyone will start using the "1-vote" strategy.
The problem here is that in free voting system it's easy manipulate, let's take the capital decision.
If I vote bananas, SIP. I gave 1 to bananas, 0.5 SIP, 0 to PH
someone else votes SIP, gives 1 SIP, 0 bananas, 0 PH
best strategy if you want win of your option is to not vote for other options since then you don't give another vote advantage.
It feels like competing while trying to figure out the majority of votes.
I certainly could just say Bananas and be finished, but would be a bit unfair to people saying SIP only (since they obviously don't want win for any other place, they choose best strategy how to achieve this).
Eventually people will grasp this concept and the consensus will be on plain majority since everyone will start using the "1-vote" strategy.
Caledorn
Emperor
I agree with this! I "gamed" the voting process in the exact way you describe here by removing my second vote, as I grasped exactly what you say here.
I also think we should include something about only voting for the presented options so we don't get "North Pole" votes. There is absolutely no reason why someone's votes should count less than another's. We are all equals as far as I am concerned, and everyone's vote should count equally.
We have been given a day extra more or less because of civfr's request for a reload, so we may want to use that extra day to ensure we all agree on the voting system.
Note: If I reload the game, the timer will be back at 96 hours. I strongly recommend we do not abuse this to postpone our decision, and that we complete our turn within 24 hours after the reload to avoid any ire from any of the other teams. It will reflect very badly imho on any team that does not complete their turn within 24 hours of the reload (which so far has only yes-votes).
Sent from my HTC One X using Tapatalk 2
I have used a different multi-vote system in the past. You get N votes (N <= number of options), and place them all on one or split them according to how strongly you value each option. For a large number of voters this differentiates very well.
But I'd be equally OK with keeping it simple. Use the built-in poll system with either single or multi choice, and take the poll number as the answer. Set the poll to auto expire, and public.
But I'd be equally OK with keeping it simple. Use the built-in poll system with either single or multi choice, and take the poll number as the answer. Set the poll to auto expire, and public.
I see your concerns. The vote "manipulation" is a known property of unrestricted borda count and I did my best to explain this in the voting thread. It is true that if everyone strongly supports a single solution, borda count with unrestricted number of votes effectively reduces to simple majority. However, I don't think that is the situation most of the time. Voters that support multiple options get their fair say too. This voting method requires the voters to decide which is more important, pushing for their favourite or ensuring that one of their viable candidates wins. IMO, this is not necessarily a bad thing. However, if the team feels this is not fair or optimal way of reaching the consensus I'm up for any good alternatives. Other voting methods that I believe are suitable for us are Approval Voting and Instant-runoff Voting (IRV). Personally I still think Borda is good enough. IRV is good too and it fills Later-no-harm criterion (the manipulation problem with Borda), but tallying the votes takes a bit longer (unless you have a software to do that of course) and it fails a few criteria Borda count doesn't IMO, the most important for us being participation criterion.
BTW, I must say that I'm also surprised and sufficiently impressed by the ingenious way Socralynnek disqualified his first selection. Never seen that happen before.
BTW, I must say that I'm also surprised and sufficiently impressed by the ingenious way Socralynnek disqualified his first selection. Never seen that happen before.
vranasm
Deity
the difference here is the thinking...
for me the SIP is just reserve when the bananas don't win and will try to live with it (even if I think it will be ... suboptimal decision), but still means I have something to play...
but what to think about someone saying "SIP only"...does it mean that if by some miracle bananas win they will stop playing?
Now I feel guilty that I didn't use "bananas only" since I harmed actually my best option. SIP is surely something I don't want to end with, but will try to live with.
for me the SIP is just reserve when the bananas don't win and will try to live with it (even if I think it will be ... suboptimal decision), but still means I have something to play...
but what to think about someone saying "SIP only"...does it mean that if by some miracle bananas win they will stop playing?
Now I feel guilty that I didn't use "bananas only" since I harmed actually my best option. SIP is surely something I don't want to end with, but will try to live with.
Caledorn
Emperor
I don't think anyone will sulk that much, vranasm. Speaking for myself, my enthusiasm is not lessened by a banana capital over a SIP capital. I feel pretty certain both are strong choices. And even if one of them were outright bad, I am dedicated to this game and want to play it out to the end no matter what
Sent from my HTC One X using Tapatalk 2
Check the part 4 of the voting instruction
Just edit your message before the deadline (a bit over two hours to go) if you want to. I believe that as long as the votes are open, it is fair that anyone can change their vote at any point. Otherwise the latecomers have tactical voting options the early voters don't have.
BTW, I voted for two options and I'm sticking with it. Effectively my vote (Bananas-SIP) is a vote against PH and slightly in favour of bananas over SIP. Personally I'd like the voting method to support this kind of thinking as well.
