About Long Bets

suggest how to make Long Bets better

The discussions on Long Bets last as long as the Bets. What reads well over time?

Long Bets are about ideas, expressed by persons. Respect for both is appropriate in these discussions. Debating ideas is fine; attacking persons is not. Relevance counts. Take account of posterity's opinion, which will be showing up soon enough.

I'd like to be able to read each bettor's supporting arguments for their position.

Good point. This first version of Discussion is not as densely linked with each Bet as a later version will be. For now you have to take yourself over to the original Bet Page to see the arguments.

Perhaps a useful fix for now would be to add to each Bet Discussion a direct link to the relevant Bet Page.

Just like that, Catherine Bacon at Long Bets did an instant workaround, so Post #2 on each Bet Discussion is a link to the original Bet Page.

How about adding some kind of online polling system to let the rest of us vote in support of propositions? That way, we could see what popular opinion says at any moment in time (and maybe track changes in it?)

We agree, and voting on bets is a feature you should see in the near future.

You're in perfect synch with Tiffany Shlain, who made the same voting suggestion at a talk that Kevin Kelly and I gave in San Francisco about Long Bets. Tiffany is the impresario behind (and exuberantly in front of) the Webbies---the tongue in cheek awards for best Web Sites.

At present the idea for voting is this (improvements welcome):

On a Bet Page, if you find you strongly support one side of the bet and are willing to go on the record with your vote, you can vote (once only) for one side or the other. Your vote and your true name then become part of the permanent record of that Bet. At the eventual conclusion of the Bet, your name is there in glory or shame, having supported the right or wrong opinion.

Of course a running tally of votes on each side is kept on the Bet Page for all to contemplate.

Sound interesting?...

How about a forum on improving forecasting itself?

Let's face it, this industry has a DISMAL track record.

In response to the question "Why is the minimum betting amount so high?" the LBF has answered in part "Maintaining institutional and technical continuity to keep track of Long Bets and manage the whole service over decades and centuries is an accumulatively
expensive process." This may be true, but the answer isn't in keeping with the spirit of the rest of the enterprise. A better answer would be "The minimum amount is high because we want Long Bets to initiate well-reasoned discussions about weighty issues. A low betting threshold would encourage frivolous wagers that are more suitable for another forum."

Well said; also true.

And some who don't have $1,000 but want to bet are finding either a sponsor or a pool of friends (whose multiple voices might show up in the argument) to put up the money. Either workaround adds to the nonfrivolousness of their bet.

Yes, the idea was to filter out the casual throw-away wagers so that only serious willing-to-stand-behind bets were left. And to raise the bar of attention. And to elevate the likelihood of some revenue to pay for the long-term service this wants to be.

I would like to have a feature where I could get an email each time a new bet is posted. That would keep me coming back to the site--much easier to get my attention that way than for me to have to remember to log on.

(By the way, the design - both flow and graphic design - of this site is very nice.)

I have implemented a Delphi Method voting mechanism at
http://www.outlander.com/delphi that includes some of the issues on longbets(this system is currently in a very rough prototype but does work).

The Delphi Method is a process of anonymous iterative voting. I'm asking people to estimate the odds they place on each event(which I felt was more accurate than a simple less no estimate). People can change their estimates at any time.

There are still quite a few features that need to be added to Executive Delphi Support. I have worked out a scoring mechanism similar to that used by the weather bureau to assess accuracy of weather forecasters that I will be implementing along with better authentication. Also, there is a need to allow folks to express the inter-relationships between predictions. Many predictions are based on a set of assumptions about the world--this is especially true of long term predictions. A major nuclear war or a comet impacting the earth could change the odds of many predictions--as could smaller events like a major
economic collapse.

I would also encourage folks to look at www.ideosphere.com
a prediction board that has been operational for quite a while. This page is more mature, but has IMHO some problems with user interface and scoring mechanisms(i.e. it doesn't distinguish well between folks that are accurate and people that are lucky).

One big concern I have about Longbets is how Judging of bets will be handled and how it will be made certain that all bets are clearly worded enough that judging will be clear at a future date. I think this discussion forum will help clarify wording but It think this process may need to be formalized.

What we've found at www.ideosphere.com over time is that wording and sources need to be handled quite carefully. For example, during the Million Man March, we had a situation in which there was a bet on how many people would be there, the government sources were under that limit and the third party sources were over that limit. I had written that claim and and made clear I wanted third party sources to be used but had I not done so, this claim could have been difficult to judge.

You're right....much technological forecasting is written in astrologically vague terms and Long Bets should aspire to something more concrete. Far too many futurists make a living out of woolly forecasts that are too vague to evaluate properly when the time comes.

Several bets (like no. 6) need to be clarified.

The U of Iowa political science dept. has a futures market that pays off. It is a political market that allows people to vote with their wallets.

Each candidate is valued at a share of $1 with the final election providing a $1 payout in theory. This would allow for group participation sub the $1000 dollar level. I know you don't want a gambling arena, but the small stakes would make the public forum more interesting and lively. It would also allow for swings in opinion over time for or against by the $populi.

These could then be tracked. Historical opinions about the future are interesting to track. write now all bets are viewed as 50-50% by their participants. Let's hear the chattering classes pitch their pennies and see public opinion sway.

are self referencing bet accepted ?

for examples, "I bet that i will not win this bet." is a bet that nobody will win.
or "I bet that somebody will accept this bet." is a bet that I can't lose.

In my humble opinion, modify the rule set to avoid this kind of bet is an improvement.

(sorry for my poor english.)

Adrien.

Nick Gogerty is right. The Iowa futures market is definitely worth a look---
www.biz.uiowa.edu/iem
Its gambling-like activity is allowed because it is strictly educational, though real money is involved.

Another site using real money is Britain's rather wonderful
www.flutter.com
It is a flat-out gambling site which deals in odds on all sorts of issues, many far beyond sports. It is illegal for Americans to use.

(Hmm. I see the Flutter link no longer leads to the British gaming site but to a site with advertising for various gaming services. I wonder if access to flutter.com proper from the US is now blocked.)

Probably some self-referencing bets could be intriguing and permissable. It's a category worth exploring. For instance, this one should be acceptable:

"The Long Bets Foundation will last until 2050 and pay off this bet."

Will it be possible to join a bet, which is to say, actually contribute the minimum stake to join one side or the other? If so, there should be some mechanism to close betting, or at least show when the bet was joined, so that early joiners receive more credit for prescience.

I agree that being able to 'support' a bet sounds on the face of it like a good idea. However there would be some complications, the most basic of which being that at present the bets are 'matched' with equal amounts being bet by the initial bettor and by his oponent.

If I (for example) want to support a bet to the tune of $1000 then my designated non-profit organisation gains no benefit if I win unless someone has supported the other side to an equal amount. Presumably there would therefore have to be 'open' and 'matched' supporting bets?

Also I think that 'supporting' bets should only be allowed if they include some argument / details not already covered by the original bet. That text possibly being placed on the main 'bet' page under the appropriate side.

It is dubious as to how well the number of 'open' supporting bets would indicate the strength of support for one side, because it serves little purpose to 'support' a bet that already has a large number of unmatched bets as it is unlikely that your 'supporting bet' would ever be matched.

It would all seem to be rather tricky

"The Long Bets Foundation will last until 2050 and pay off this bet."

But the only way someone betting against that proposition would get money for his designated non-profit organisation would be if the Long Bets Foundation doesn't last till 2050 and some _other_ organisation pays off the bet.

It is indeed possible that betting against the longevity of the Long Bets Foundation is a seriously self-defeating bet---no foundation, no payoff. That prospect could be part of what makes it an interesting bet.

But there are potential backups, and such a bet might help force their existence and clarity about them. One backup is the organization holding the stakes, Capital Research. Capital has already been around longer than 50 years, is doing very well, and is likely to last another 50, if civilization does.

Also Long Bets is closely connected to The Long Now Foundation, which has other long-term duties such as the 10,000-year Clock. The assumption at Long Now has been that if Long Bets poops out, Long Now will see its obligations through.

It's encouraging to see discussion about joining a bet, because a feature long planned is "Bets in Parallel"---the ability to join an existing bet, perhaps at a reduced rate, with each additional bettor adding a further argument.

It's a complex feature to add, so it will take time, and first should come the Voting feature. Response to it may say more about how Bets In Parallel should be designed.

I'm not up on the finance stuff, so forgive me if I'm asking an obvious question.

Can the designated charity be one that does not operate in the USA?

Yes, you are welcome to name a charity in a foreign country.

but the 'Bet 4' discussion forum doesn't seem to be showing the latest post [From Steward Brand on 27th].

[edit]

Aha, I see it shows the latest post details even _after_ that post has been deleted. That's not particularly vital I suppose, but you might want to fix it.

So far I don't mind that a deleted post leaves a brief cloud of smoke where it was, as you alertly noticed. I wrote something dumb, then unwrote it, and you were enabled to catch me at it. No harm done.

It is either a mild bug or a mild feature or both...

"I wrote something dumb, then unwrote it, and you were enabled to catch me at it. No harm done."

I spy another phantom post. ^_^
Now I'm itching with curiosity as to what it's contents were.

I must have deleted my memory as well. I don't recall. Meanwhile, William Pietri and Jim Home are shaping software and feature improvements to the the site, and that phantom remnant from deleted posts may disappear. I'm counting on you to alert us if it doesn't---thank you for your help catching glitches.

While reading some bet discussions, I found myself muttering "You think it's not going to happen? Bet against it!"

Here's a mechanism that (at first glance) minimizes bookkeeping.

1) Skeptic opens a "Hogwash" account with a minimum deposit of US$1000. The charity is designated and the funds are transferred to Long Bets.

2) Skeptic can subsequently commit funds from the account to join the No side of any open bet. Commitments are irrevocable and not subject to negotiation.

3) Any money remaining in a Hogwash account is transferred to the charity or to the Long Now Foundation two thousand years after the last activity.

4) Since Hogwash bets are not subject to negotiation, a bet proposer must be confidant of her/his bet's wording, and Long Bets must be confidant that the terms are ajudicable. Thus the bet proposer must explicitly state that Hogwash bets will accepted (up to the limit of the original bet) and Long Bets must agree that the bet conditions are concrete.

Enhancement:

Mark posts in the Bet Discussion thread with icons reflecting the degree of financial obligation, such as a little bologna for US$100, a rolling-eyes face for US$10, a tilted-head for US$1, and hot air for US$0.

I have just signed up for the site and I can't see any place to amend my profile. Since I will be on the site for a long time, what do I do if I need to change my details?

1. Bet 70 is below bet 69 in the list.
2. There shouldn't be 'bet details' link on the non bet forums (e.g. this one).

Soon you will be able to edit your profile on the site. In the meantime if you have changes you want to make you can email me at catherine@longbets.org and I can make them for you.

The bets are listed in the order they were approved for the site. The original numbers are assigned when a bet is proposed. In this case bet 70 was approved before bet 69. It can easily be moved if this sequence offends.

It (ordering of bets) doesn't offend me, but it does look untidy and may make it (slightly) harder for people to use the forum.

If "date of approval' is to be the ordering criteria I suggest that should be prominently displayed in the forum list.

I wondered what procedures are in place to ensure that participants in a bet will be able to come to agreement on whether the terms of a bet have been met or not.

No offense to the Long Bet staff, but managing this system and the related financial stuff isn't the same as being an expert in the topic of any particular bet.

Suppose that the person betting FOR something ends up in a 'yes it is', 'no it isn't' argument with the person betting AGAINST. How would that be dealt with?

One possible idea is that both bettors would have to agree on a judge that both view as a) impartial, b) knowledgeable in the field. They would then be able to call on that judge rule on the bet outcome as well as make technical rulings on ambiguous points - if the bettors cannot come to agreement on those points themselves.

Obviously some bets are harder to decide than others, the 'Turin Shroud' bet for example I view as potentially problematic.

So far, the Bet conclusion procedure is...

1) If both original bettors are alive and if they agree who won the bet, that settles the matter.

2) Otherwise Long Bets will take responsibility for deciding the outcome, That may be by a blue-ribbon panel of judges, by judges selected for expertise in the particular bet, or by other means as seems fairest at the time. (If the Turing Test is passed by then, maybe computers will do it.)

Public discussion will continue with each Long Bet after its conclusion, so the issues can continue to be hashed out there, and the judgement itself debated.

As people[1] may well be following the forum regularly for new posts, I suggest that changes to the bet's status and changes to the bet details be noted by a LongBets' staff member in the relevant thread.

[1] Me, for one.

As well as being able to vote 'for / against' bets it would be interesting if you could also vote 'important / trivial'.

Perhaps a place could be found on the home page for the bet voted 'most significant'.

It would be nice to have threaded discussions in some area, especially in the 'Prototype Bets' and 'Improve Long Bets' areas. Its a bid hard to follow a thread right now.
-Eric

"Important/trivial" is an interesting distinction for voters to draw out.

It might be worth speculating how that would play out.

Single-thread discussions with reply-to header continuity helps keep the discourse convergent and in one long timeline---easier to peruse in retrospect, which is most of the idea here.

I like the basic idea of a timeline approach - the current problem (unless I'm missing something) is that when you get thousands of posts, finding the one you want (or the newest,etc) is a bit tricky.

For example, is there a way to see only the latest posts across all boards? For that matter, in a single board?

Quote "For example, is there a way to see only the latest posts across all boards? For that matter, in a single board?"

A sensible compromise might be to show the posts in reverse order in a thread. Most recent page opened first, most recent post at the top.

This would be a great start!

Let's see if we can try reverse order with this discussion, see how it feels.

I'll check with James Home and William Pietri.

Another improvement would be to make the latest post as listed in the discussion index a link to that post. This way, for those of us tracking a few threads we could see the latest post and quickly respond...

While I see most of the posts to be constructive criticism I'd like to pause and complement you and the team on a very cool idea. I've read the Clock of the Long Now and found it to be thought provoking. In fact, thinking in time chunks of 10,000 years provides great perspective, particularly when you look back and then try to look forward...

With one letter change in a line of code, the order of posts in discussions is now reversed.

What do you think? Improvement? Confusing?

My first impressions about having posts with most recent on top:

* It emphasizes the present instead of the past.

* Fewer clicks to get around (good!)

* Picking up where one left off is trickier, helped by noticing how many posts are "new".

* There is an intellectual but not necessarily inuitive parallel with how the past is buried in archaeological and geological stratigraphy. Present is on top, past is down deep, future is up in the air.

* Reading convention, though, is the opposite. We read from top to bottom.

* So, what weighs strongest for you in actual use?

I think that most people are used to dealing with time from the present backwards, rather than an arbitrary starting point (april 24th or what ever).

On the other hand, as new people join the discussion, it might be difficult for them to dig deep and understand the progression. However, for those who are interested, this is a one-time event (getting up to speed) vs. the daily/weekly events of seeing what is new.

So, my vote, keep the new up front for a while and see how it goes!

I think having the most recent _page_ shown first is fine. But having the most recent _post_ shown first is a little odd at first.

What many forums use is a traditional 'old posts at top' arrangement, but they have links to allow users to jump to the most recent posts.

Could an email be sent to a bet submitter each time someone adds a discussion item? Bruce Damer (Bet 42)

Interesting idea.

I imagine only some would want such a service. And some might want it valved---let me know when there are 5 new posts in the discussion of my bet.

There's other features to be added first---such as voting on the bets (coming)---but I could see some version of notification to bettors of dicussion posts being added as Long Bets gradually becomes larger and more feature-rich.

I was excited when I first read the New York Times article on your website and its content. However, I immediately surmise that outside of the value you are bringing in this altruistic endeavor, the day to day appeal may well wane for many. Let me explain.

While the nature of the posited bets is interesting, the results of the bets have two aspects which warrant comment. Firstly (and obviously) they are a long way off, and secondly the bets hold in the interim only conjecture and discussion regarding the topics in question. This in itself is worthwhile as it fosters discusssion. However, this discussion can be found in other places: journals for example.

A summary of the role you have for this site then can be simply this: a site for scientific and other discourse with bets attached to them that benefit charities. We bet, we discuss, and then we eventually determine a winner. The news of a definitive winner would likely in most cases be accounted for by newspapers and other media (most of the bets are newsworthy in themselves).

There is to my mind a possibility of the site going stale (because of the lengthy wait for resolutions). I have a suggestion which might add more immediate results (something most seek) and at the same time this idea would offer us potentially enormously valuable knowledge. My suggestion does not replace the admirable activities of the site as it operates now, but would be in addition to it.

I propose a category of $100 postulations (not bets) which must be limited to no more than a four year (or so) time frame. Someone may make a prediction about anything (within the guidelines of the current bets) and would pay $100 for the privilege. The money goes to his/her charity of choice. There is no limit to the number of postulations/predictions a person could enter. The benefits would be myriad, and fascinating.

A few people would emerge as accurate prognosticators, consistently getting their speculative statements correct (it would of course be up to you to reject any which are obvious - eg. the sun will not nova within five years). The worthiness of determining those scientists and others whose ability to think rationally and demonstrate accurate predictions consistently is clear: these people, whether academics or not, might then be sought out for their opinion on other matters. In a sense then the site would expose minds which the world can benefit from (and might not be aware of currently).

There is a more immediate reward in short bets for both those who participate and those who observe. Because there will be results sooner and more frequently, the site will be able to generate publicity and draw new participants.

A batting average of sorts (like the ebay rating system) could be applied to individuals, and you could foster an element of friendly competition (something I believe which occurs in the communities most of the current participants work in anyway). One could call it the Rational Clairvoyancy Average (RCA). I can well imagine university teams forming, and I would think that both individuals and teams would have an enormous amount of fun. I would hope too that a few weighty RCA's emerge for our collective benefit.

Charities would greatly reap the rewards from such competitive brainwork.

In closing, I do not mean to disparage or reduce the excellent thing you have created here. On the contrary my comments reflect my entrepreneurial interest (I want this site to last and flourish) and as such the site might need to appeal to general human nature ( we like immediacy, results, and competition). Most interesting to me is finding these RCA's - thereby performing a valuable social and scientific service which can be utilized and expanded upon.

I myself am no great thinker and have as evidence been predicting the Toronto Maple Leafs as Stanley Cup Champions for years - that and the sorry state of my financial investments means that I hold the only negative RCA known to mankind which languishes in the triple digits. I am a layman, and regard you fellows (it is interesting that few women participate here) as the few who evolve the many. I take my hat off to the work you do.

If you do implement the idea, I will happily divert some funds for the launching of a few ideas. Just don't expect me in the distant vicinity of the leaderboard.

Cheers,

Adam Bronson

Alternatives:

* Votes for important / trivial are registered only if paired. Expected benefit: 'In the two centuries since the commencement of the bet, fifty million members have disagreed on whether the matter is important or trivial.'

Adam Bronson, you make the best argument I've seen for traffic in some shorter bets or predictions on this site, and a good argument for a low price of entry---statistical-level volume. Grinding slow but fine, I expect Long Bets to offer a fair portion of your idea set in the comng months.

Could you drop me a line at my registered email and / or saotome_genma@hotmail.com

Incidently I've tried to send an email to info@longbets.org on another subject and it just goes 'bouncy-bouncy' back. Possibly because of my unusual set up regarding email accounts / reply addresses.

Once the voting system is in, one way to encourage participation would be an honorary title of "Head Oracle" for however has the track record for voting earliest, most frequently, and most accurrately.

Good idea.

It's very convenient when revisiting these discussions to have new postings indicated, and the software tries to do that. In the Discussion Index, the number of Posts column shows the number of "new" posts since the user last visited. And each new post in each discussion is indicated as "new" in its header.

But I've learned the user has to do two things to make the "new" feature work right:

* Always log in before visiting Discussions.

* Always leave a visited Discussion by way of the "Discussion Index" link, not by backpaging.

This is my current opinion on all the bets.

(100% = Completely certain, 0% = Wild guess)

Bet 1 : YES 70%
Bet 2 : Awaiting clarification.
Bet 3 : NO 30%
Bet 4 : NO 50%
Bet 5 : NO 40%
Bet 6 : NO 80%
Bet 7 : NO 20%
Bet 8 : Don't know, Don't care.
Bet 9 : NO 25%
Bet 10 : NO 30%
Bet 11 : YES 80%
Bet 12 : YES 10%
Bet 13 : YES 20%
Bet 14 : Don't know, don't care to guess
Bet 15 : Evens Stevens
Bet 16 : NO 10%
Bet 20 : Awaiting clarification
Bet 22 : NO 45%
Bet 26 : YES 80%
Bet 27 : NO 100%
Bet 30 : NO 30%
Bet 39 : YES 20%
Bet 42 : YES 50%
Bet 43 : NO 100%
Bet 44 : Awaiting clarification
Bet 63 : Depends too much on external factors.
Bet 69 : NO 85%
Bet 70 : NO 35%
Bet 78 : NO 10%
Bet 77 : YES 20%
Bet 76 : YES 70% (some clarification would be helpful)
Bet 80 : NO 60%
Bet 86 : NO 100% (some clarification would be helpful)

I am also impressed by the 'long bets' idea and implementation, but the further possibility it suggests to me, as either a corollary of long bets or a separate aspect of 'the long now' projects is to track actual cases of 'bets' (ie predictions) in the public/scholarly domain and hold them to some sort of account.
Clearly you'd have to restrict the field - prophets and prophecies being just too thick on the ground - using some criterion of 'seriousness'. Part of the point would be to provide a portrait of the 'actual' horizon we work with in the backgound as distinct from when we have to speculate on those possibilities in themselves.
- Melissa Mcmahon

It would be nice if there was "Last post - Bet XX, Oct 7" or similar as a link from the front page.

I noticed when logging in today that my posts have an "edit" button on them. If Long Bets is about accountable predictions, then you shouldn't be able to edit your posts, except possibly very recent ones to provide a cooling-off period. Anything committed to the public record should stay there permanently. [Incidentally, this is a major problem with using a website as a journal of record since sites can be taken down for many different reasons]

Good suggestion. This is the kind of thing I would keep an eye on, see if there are offenses, and then limit the editing time as you propose.

As for Long Bets being a journal of record, that is in a sense its primary function, so much will go into keeping it intact and accessible. Long Now is working on a "Long Server" project to handle digital storage, including working websites, over decades and centuries.

Ok, maybe I just like to hear myself talk, or hear other people talk about me, but maybe there should be a system that lets follow particular string. I haven't even gotten through all the threads, so it is hard to keep checking the ones I've posted to to see if anyone has said anything about what I've said. It might sound a little self centered to be most interested in what other people say about what I say, but, after all, I post on the threads that most interest me in the first place. If there was some sort of way to receive e-mails if someone posts on 'bet 32' for instance, it might be very useful, or perhaps a search engine that lets you search out comments from specific posters, which would also be useful for tracking threads from other posters.

I like the idea of voting on the bets and of keeping records on short term bets. Keeping track records would be an interesting way to add some more short term gratification to the proccess, especially for those of us who can't make bets, even for charity. ($1000 would be about 15% of my annual income.)

Another idea would be to organize the bets in a couple different ways. It would be fairly easy to do it by timeline of the completion of the bet, and by the placing of the bet, and it wouldn't be too hard to organize them into broad catagories, like computer developement, cultural changes, etc. (You could cross list bets that fit into one or more catagories, and highlight the ones that the viewer has clicked on in another catagory.)

Since any money on short term bets would be going to charity- (will robotization of all important aspects of our lives make charities obsolete by the time some of these bets come to pass?)- is there really isn't a need for all bets to be equal value on both ends of the bet. If someone bets $10 and someone who has the money bets $1000 does really matter, although I suppose a certain threshold might make people think more about their bets before placing them.

Well, that is it for me. I enjoy coming hear to read. Speculation on the future is something I love doing and this site is great.

I have twice tried to enter a prediction via credit card. After I click the 'submit' button, I get an error message ...

There seems to be a problem with the information you entered.
The transaction was declined for the following reason(s):
Amount must be greater than 0.00

I cannot see any place to enter the $50.00 amount.

Thanks for your details on this, Thomas. We'll get on it.

Apologies for the frustration!

Thanks to Thomas Holaday's bug report, the credit card glitch has been sorted out, and Holaday's Prediction is On the Record.

The current vote should not be shown until
a) You vote
or optionally
b) You click a link to reveal it.

Otherwise your vote may well be influenced by seeing what the majority view is.

"your vote may well be influenced by seeing what the majority view is. "

I'm sure you're right. Why is that bad?

Bear in mind that over time the majority may well slide back and forth. (That will be fun to graph.)

"I'm sure you're right. Why is that bad?"

Because what I'm interested knowing is the opinions of those voting, not to what extent they are lemming like. ;-)

What if for a future Long Bets were to offer an open ended bet (ie What will be the news story of the year in the year 2075?) and then the money goes to the charity of the person who was deemed most accurate (via a panel of judges who will not know any of the bettors). These bets would be done at a fixed ammount (ie $100 or $250) and have a closing time (in the case of the example: 2005), bettors would be allowed to change their bets till the closing time, any bet that is deemed too close to a pervious prediction will not be accepted (at the descresion of the mediator). What do you guys think, a chance to see who can see the long term or just a silly prediction time capsle?
-Joe Kavanagh

I agree with Paul

As of June 4th, 2003, the HTML for the discussions presents the content of a postings as a pair of "td" elements with the class "txt," the first one bold (for the subject), the second one normal. If the forum author could add a unique "id" attribute to one or the other of of these elements, it would be possible to make them hyperlink targets.

Alternatively, and perhaps more suitable for a multimillenial effort, the postings could be numbered.

I was surprised and delighted to find this site. It seems
an extraordinarily important effort. My suggestion is along
the following.

No prediction or bet stands alone. Each depends precisely on the
sequence of world events, technological advances, geopolitical
changes, happenstance and so forth that precede it.
For instance, prediction #26, that 50% of all root servers will
be outside the US in the year 2050, depends on the existence of
the internet itself.

Because of this, I think it vitally important that predictions be
allowed to *reference* other predictions. Not off-hand in the
text of the description, but linked as with a database. In fact,
very quickly (before too many more people enter predictions)
this needs to be converted to a database driven system.

Entries need to be taxonomically categorized, n-levels deep.
Prediction #63 should be something like

religion:historical artifacts:turin shroud

Because very quickly in its current form, organized by ID#, this
will become unwieldy. Once the site is database driven
and cross-linked, you will be able to start at the top-level and
browse major categories of interest, and see how many predictions
lie beneath them. Also, and perhaps more importantly, each prediction
will take on a value that directly corresponds to the number of
predictions that depend on it coming true. Major predictions that
predict entirely new sciences or technologies or completely new
organizations such as #22, the existance of a unified world
government, will spawn and organize whole subtrees of thinking.
If those predictions become true, it will lend additional weight
to those subtrees.

Since people are for better or worse somewhat driven by personal
recognition, individual users might gain recognition or points--
or perhaps just a rosy glow to their name--by how many predictions
and subsequent dependencies they had spawned.

The ability to cross-reference other predictions, such as a
world policy prediction referencing a prediction of medical
innovation, will create strong cross-links between the trees and
strong sets of dependencies. You should be able for instance to
predict that something will come true in ((x <> y) + 7) where
x and y are the two predictions that you reference as dependencies
and <> is a symbol i misused to mean "the latter of". As the
time schedule for x or y slips, your prediction will slip as well,
as will all the predictions that depend on your own.

All of this should be able to be done easily and with a pleasing
user interface. In fact, the software currently being developed
by Hawken's company Groxis (i have no financial interest there)
would be a terrific way (perhaps as an option) to be able to fly
visually through the networks of predictions and connecting
dependencies.

Imagine a time, after many years, as predictions come true (or
perhaps more importantly *don't* come true), and the network of
predictions here self-adjusts and reforms and begins to prove
its value, where this site becomes vitally important-- perhaps
in 100 years *mandated*-- as a source of information for world
leaders in crafting decisions that affect our future. I know, we
fault our leaders for not looking back to history enough, what's
to say they would look at the future either?

I would be delighted to lend my efforts, large or small, to this
effort.

I applaud the efforts of the team in putting together this site
and creating the Longnow Foundation. I can't begin to describe
how much your thinking and writing have meant to me in my life.

Thank you,

Dan Whaley
San Jose, Costa Rica

I think it would be a good idea to have a page in which ...
1. All bets are listed with link to bet details
2. Each bet listing includes a 'you voted YES/NO' text or 'you have yet to vote on this bet'.

There are a lot of bets & predictions. At present it would be very easy to miss voting for a few and not realise that you have done so. Checking each individually would be rather time consuming and will get more so.

"Entries need to be taxonomically categorized, n-levels deep.
Prediction #63 should be something like

religion:historical artifacts:turin shroud "

Dan, your whole database approach to the body of material accumulating on this site is very much on our minds, and your suggestions are particularly helpful. For now we're collecting experience and desires like yours for providing a use-based set of criteria for the eventual architecture of searching and linking within Long Bets.

This feature should be available for anyone, not only bettors. Why not add a check box like "I'm interested in this bet/prediction and want to be notified of new comments" on each bet/prediction?

> I imagine only some would want such a service. And some
> might want it valved---let me know when there are 5 new
> posts in the discussion of my bet.

Someone might find a daily (or perhaps weekly) digest useful, but other than that I don't think any kind of "valve" is needed. (Actually such a "new messages" threshold would be quite bad in low volume discussions where it could take a very long time after a message was posted until a notification about it is sent.)
Besides, you can always filter your emails on your end any way you like.

An additional revenue source for Long Now would be (perpetual?) web hosting. That way, discussants in Long Bets could provide URIs with confidence that the content would remain available. The server could be kept in the Gift Shop next to the models of the Clock.

Interesting idea. With software like Sitesucker we could automatically (within reason) haul in anything linked to in a discussion (down just a level or two in the sucked site) and thus keep the links live.

Once Long Now's "Long Server" project is further along, this becomes plausible.

Please tell me that the Long Server will be powered by human winding.

I've been thinking about something similar to long bets, but not exactly the same. I'd like to be able to make a "long investment", with a pay-out to a charity, but not necessarily one that I can name when starting the thing.

Roughly what I have in mind is a long term charitable trust fund, say 200-250 years. Anybody would be free to put money into the fund at any time. Once money goes into the fund, it stays there, accumulating, until pay-out time. When it pays out, the money goes to solve one major world's problem.

The goal here is to generate enough money that, taken all at once, it can have a major impact. Let's assume a 12% average annual growth rate and 5% annual inflation, for a net growth of 7% faster than inflation. A compound interest calculator tells me that, under those conditions, each $1,000 in the fund today will be worth more than $700,000,000 in 200 years *in today's dollars*. If a bunch of people had gotten together to put $20,000-$30,000 in such a fund 200 years ago, we could be using that money today to make a serious dent in AIDS in Africa, for example.

How is this different from a prediction or bet?

1. At 200+ years' distance, I can't name a charity with any confidence it'll still be around. I need a set of impartial trustees to decide what constitutes a significant world's problem. For such an investment to succeed, it needs a fairly long-lived institution to manage it.

2. I don't want to make a prediction about something that might or might not happen, except to predict that the long term investment will gain value faster than inflation. Instead, I want to solve the problems that will exist in the future. I don't know what those problems will be, but if past history is any predictor of the future, it's probably safe to assume they'll exist.

There's one other important point about such an investment: we need more than one of them. It's possible for laws to change, governments to confiscate funds, or people to try to hijack them, so having more of them increases the odds that at least some will survive. Having a diversity of investments probably means there needs to be a cap on the seed money for any individual one. It also probably means that it needs to be possible, when creating the investment, to give guidelines to the trustees that they should follow when it pays out, so one investment may go to solve an environmental problem, while another goes to solve a social problem.

Finally, there's the question of "why not give the money to charities today to head off those future problems before they even happen?" It's a valid question, and there are a few answers:

1. Charities are not necessarily zero sum. Most people have some flexibility in how much money they decide to give to a charity. If it's a cool enough idea, they might decide to both put money into one of these funds *and* give to the charity of their choice.

2. A large pile of money can take advantage of economies of scale.

3. It's a form of insurance against unexpected catastrophies.

Very interesting site, the idea of forcing people to backup their pontification is most welcome.

It would be interesting to take things a step further. Problem with a bet is it is taken at a point in time and the situation will evolve.

What happens if you take a proposition and sell shares in it (selling short if you don't approve). Propositions could be proposed by the community, then "IPOed" by the administrators. The market value of the proposition would serve as a current level of belief in the concept by the community.

Advantages:
1) Serves as a constantly updated measure of the proposition validity
2) Real money provides real consequences (owners of bankrupt ideas go bankrupt)
3) Don't have to have $1,000's to get into the game
4) More scalable than bets (bets between two people, market can involve infinite people)
5) Allows the evaluation of very long term propositions with immediate concequences (who will collect on the universe stopping its expansion a Hubble time later?)

Credit where credit is due. This is very similar to what DARPA was considering with the "Bet on terror" concept. Actually not a bad idea, but REALLY badly packaged.

-- Chris

The ability to search various topics or the site in general would greatly enhance the user - individuals looking to discuss/debate predictions - experience.

It's an intriguing idea, Jeff.

I think you could use Long Bets to accomplish approximately what you want. You find a willing cohort and make an interesting 200 year Bet. If you each put up $500, that would put $1,000 bucks into the Farsight Fund, growing at roughly 4% annually over the centuries---significant money at the end. You could both designate the same charity type, and spell out in your argument roughly how to fulfill your general intention in 200 years. Then when the Bet consumates, the Long Bets judges would have the responsibility to find and award the most appropriate charity.

As far as Long Bets is concerned, this would be a welcome way to game the system, and the organization would be honored to help conduct your long-term charity.

For now Google works surprisingly well:

Type "long bets" and a keyword and you'll go pretty directly to the relevant Prediction or Bet, and even Discussion.

In time Long Bets expects to add serious search and browse functions to this site. That will take some funding not currently in hand---something on the order of $20,000.

the currency for bets should be in a currency that reflects the duration of the bet - since paper fiat money issued by any govt is inherently short term in the context of your organization (i am sure that no nation/tribe exists from 8000BCE, and likely no countries from the present will exist 10,000yrs from now), i would expect that any bet of duration longer than 100yrs would have to be denominated
in something that you (and/or your community of members) think will have value over the tenor of the wager - not sure what that would be, but if history is any guide (which it may well NOT be), the currency of the longnow bet could be gold/gems or spice/pepper - however, using our current behaviour (as a species) as a guide, i think that units of pure water or units of clean air might be candidates - ofcourse when man moves into an almost timeless existence by leaving behind the current carbon-based physical body and inhabits a more durable, robust and more fully-featured mode of transport/expression, then money (and therefore the basis of risk implicit in a wager) will likely be meaningless

Ban all bets on religion. The shroud of Turin has been proven to be a fake! Do you want the religionest to start making bets on the santa croce titulus in Rome,the sudarium of oviedo in Spain or the thousands of other "holy" relics, virgins milk?? The lance that stabbed christ?? Once something is fake it is fake forever!! Leigh Alan Demrow

Coming to Long Bets from other discussion forums I find it very clunky and hard to follow discussion on particular topics, as posts are only ranked chronologically rather than grouped into threads by subject as per the thousands of vBulletin boards out there.
I believe if discussions were grouped in threads it would make it a lot easier; especially as the forums grow.

Another useability feater these discussion boards could do with are hyperlinks. It is the nature of these discussions that people would like to refer supporting URL's, a feature to enable direct hyperlinks from smart software (that recognises "http://..." or "www..." and converts to a hyperlink), or by manual use of tags, would be much more convenient than copying and pasting addresses into the browser.

Testing
http://www.vbulletin.org/
[url=http://www.vbulletin.org/]vbulletin.org[/url]
vbulletin.org

Ahah! Properly formated html code is allowed... perhaps just a short note in the header or footer is required so people are aware and can include hyperlinks in their posts? I've seen a lot of posts with links but none are hyperlinks...

Here's another advantage of having predictions database-driven. The Long Bets database could be linked to other databases.
Think of the Alexandria project, if it expands from its current mapping focus to producing the modern equivalent of the Library of Alexandria.
Or, more currently, consider the Library of Congress and the work monks are doing to digitize the entire collection. (The monks work in a modern version of the scriptorium, but I will discuss that separately.)
A third example would be a database I envision matching up 220v-capable electronics gear in American garages and attics with enterprises in Africa such as a conference center and leadership training effort now under development in the Copperbelt Region of Zambia.
With a zillion linkages available, some with money attached, the possibilities are endless--and there's the rub. OK, I'm not a database expert. How could a database help social entrepeneurs sort all of this out, help them do their due diligence for a particular project, and help marshall the text and graphics for presentation to decision makers with their fingers on the money valve.
I'm new here and would appreciate some guidance on how best to participate in discussions. I don't have any extra $1000 bills lying around but am a "temporal millionaire" with a lot of control over how I spend my time--so it looks like discussions are it for me here.
Is the above discussion appropriate? Too detailed for a follow-on? Needing its own topic?

I revisited the site recently and prefer the discussions in reversed age order, as they are now. Good job.

Although Mr Bronson's idea, below, of SHORT long bets is intriguing, and I especially like his idea of finding accurate prognosticators, something about it does not seem right to me. On reflection, I suppose it is that this site, like the Long Now site, is geared to the long term, which seems incompatible with "the site going stale (because of the lengthy wait for resolutions)". The immediate results that he rightly says most seek (and I know I do) can be found at other venues- but I don't come here for them, because that is not to me the purpose of the site. A site devoted to finding people with high prediction ability is not a long bets site, it is a person-centered site, and it seems to me simpler to keep the site definition narrow on such a long-term project.

In short, this is not Slashdot or Salon or a wikiweb or a chat room, and it will seem stale by comparison, and that is OK.

I would like to stress that I am not attacking Mr. Bronson; his idea seems enthusiastic and fresh to me, and he states it eloquently and kindly, and I may even come around to it- but I just felt nervous that this simple site would get unnecessarily complex.

I bow to both Mr. Bronson and the makers and maintainers of this site.

Stefan Sittler

You folks need to have a place designated or a way to highlight the bets and predictions that have played out- two or three now, right? It should be immediatly obvious which bets/predictions have been resolved and which have not. Bold/Emphasize/Highlight them on the list, put them on another list-- do something.

Oh- Also for goodness sake close voting on completed bets. It makes an accurate tally harder to see.

Two major issues:

One, a more careful vetting of new bets would be good. Some bets have been added to the discussion area in almost unintelligible form. It would be a definite improvement if moderators would ask the submitters to clarify, edit, or simplify their proposals before the bets were unveiled.

Also, some bets are clearly simply self-promotions and should at least be rephrased so they don't seem quite so shallow. And, last but not least, bets 175 and 176 (if I remember correctly) are virtual duplicates, which probably shouldn't be allowed. (If very similar bets are allowed, they should probably be accompanied by a brief explanation of how they differ.)

Two, traffic at the site seems to have dwindled precipitously. Certainly, there's less commenting in the discussion area. The site needs to attract more traffic.

We would love to make the community of Long Bets larger. Are there any of you that would like to volunteer to as moderators?

Hi,

I'm so glad to rediscover this site. I've become a fan of the tv show Galactica. So I'm thinking again about the singularity and many long bets.

I noticed forum posts are shown newest first. What an amazing idea. We view blogs and emails, perhaps more this way. Why not forums.

I'm sorry for jumping in; I'm not sure if this topic has been discussed, or if this is the right place; how about column headers that act like a real application? :) Within an email program, we can change the sort order. Viewing thread titles or posts within a thread is very useful.

There could be some other interface improvements too. Ajax finally took off. It took long enough for people to say, hey, web interfaces, how about they act like real apps? About this topic, it might be treading forum software that won't exist for a decade. Maybe? Is anyone thinking about real web forum UI design?

Still, the design here is amazing already. The typography, layout, and readability is just beautiful.

Hi zander,

I'll volunteer. I would love to.

Hi frisky,

very interesting. Something like this could be very useful. But I think we need a new way. Stuff like [url=http://]my site[/url] is much to ask of people.

How about something like?

link=my site=www.mysite.org

Or people paste urls as they wish. When previewing a post, can we get some ajax goodness and an easy way to add descriptions?

For link color, sometimes I really would like to keep have no color at all. Imagine every word in this sentence linkable to it's definition. Useful? There's lots of uses for invisible linking.

For a pasted url link; http://.., I like that it's now black. The forum could interpret this and make it a live link but keep the color black. I think that's a good idea.

Is this the only forum for site discussion? There are many meta topics about this site. Is one forum enough?

I was just thinking. Keywords for forum posts? I like it.

Thanks for offer. We now have someone who will be working in te office on this project for the summer and will wait until he is in to discuss this next month. Please send along you email address to me to zander at longnow dot org so we can hopefully work together.

thanks again.

I'm wondering if it's possible to view new posts. That'd sure be useful. :)

It would be really good if you could type things in and find the results easier, i was looking for predictions on WW3 and i didn't want to have to spend hours looking through all the hundreds of predictions, it would be best if you could just type it in and slim the numbers down and get a better result of what you're looking for. Thanks

Even though it's not a betting site per se, it's going to be helpful to know who had the most accuracy.

There should be an option to see which predictions/bets/votes a given user has. Why can't I click on a username and see this?

- Ralph

Difficult Task – Discovering Equally Divided Interesting Debates?

I recently had an intense discussion with one of my friends, who works in World Bank. When we both got frustrated with equally balanced arguments, he challenged me to accept a bet for one hundred dollars. I told him I will not be alive to see the outcomes of our bet, why not we place a bet at Long Bets. To my ultimate frustration, he had never heard about Long Bets Foundation.

When I compare the frequency with which Equally Divided Interesting Debates generates in my own mind every month and multiply that number by supposedly population of intellectually gifted people in the world, I am surprised to see much smaller number of debates finally reaching to Long Bets’ balance sheets.

Simple reason that I may figure out is need for more exposure or “Marketing”.

Potential betters who may benefit from a place in historic exposure such as Famous personalities, CEO’s, Politicians, Industrialists can perhaps easily sponsor up to ten thousand dollars. But they may not have time, academic inclination or drive to be able to make any purposeful contribution to Long Bets. On the other hand, there may be many like me who are overflowing with interesting ideas and want to disseminate “Equally Divided Interesting Debates” to a larger audience, but may not have access to sponsors.

Once again Long Bets Management may help build a bridge.

To prove my point I will soon forward an intense debate which should spin off the Long Bets’ readers interest to a new direction. Of course I will still need support e.g. to “Market” or to reach out to potential audience. I mean, audience beyond existing readers who know about Long Bets so far.

Hope I can trust Long Bets Management.

There has been a suggestion on file in this forum for _more than TWO YEARS_ that you somehow mark bets which have already been resolved. I have seen at least two such bets so far browsing the site, and neither of them is marked in any way as to who won. The fact that this hasn't been dealt with yet makes me feel that this site is little more than a joke. I strongly support the mission of the Long Now Foundation, and I think Long Bets could use a little more attention. (In the interest of putting my money where my mouth is, I will point out that I just challenged a prediction, and fully intend to put up the money if the predictor accepts my challenge. I suspect they will not, however, given that the predictor, 1future, doesn't even appear to have provided their real name, as is required by the rules. Aren't the predictions supposed to be moderated?)

Hello everyone, I just registered for the first time this week.

Sorry for being lazy, but I just thought I'd ask if you could point me to any especially good posts that I should check out first, so I can get up to speed here?

Thanks, Michael

Hi all,

Several people have pointed out that the Long Now website seems to be out of maintainance, especially because long settled bets (e.g. 129) are not closed on the site. I mailed them about this (info@longbets.org), because no staff seems to be replying to these topics here. I also volunteered to help with upgrading the site (e.g. closing bets functionality) since I'm a software engineeer myself. I'm waiting for the reply now, but in the meantime maybe other people can suggest website improvements or join the website team. Let me know at ewjmulder@[domain] where domain is yahoo.com (trying to outsmart bot-harvesters :)

- Erik Mulder, The Netherlands

Why can't I add a post comment to bet#208?

Long Bets is actively in redesign. Below are some of the things we want to improve. We are now looking for a Java tomcat programmer who can help us with this as the original developer is now busy with other projects (but is available to help answer questions etc.)

Any other suggestions and or people who may be interested in this project please let us know. Thanks.


Next Release Features
* Migrate onto upgraded server and Java instance
* A way of listing a bet as complete and who won
* Pagination and ordering options of predictions/bets
o Ordering Options
+ by pageviews
+ by number of votes
+ by certain vote criteria
+ by date
+ by duration
+ the default could be a combination of number of votes and page views
+ what is best default?
+ by category (e.g. political, economic, technological)?
- this wouldn't necessarily/only be a pagination/ordering option, but would help organize the predictions bets into meaningful categories; at least, allow users to search predictions/bets by keyword
* Predictions are free
* Integration of Timeline Tool
* Bet minimum reduction?
* Users can sign up for mailing list
* Voting on more criteria
o desirability of outcome
o how much this bet outcome "matters" (eg. is it frivilous or really useful?)
* Re-look at discussions to make them less of a hack?
* rss feed?

good idea. +1

I would suggest that longbets include international sms support for each bets. This micro finance mechanism can improve the odd againt any bets, such Bets: Dave Winer vs Martin Nisenholtz. Wager can help increase the amount their wager and put it to good use.

hi it's me william morley aka william zoroaster and still don't see my prediction posted so it is safe to say it falls under the prohibited section whereby enhancing ourselves is not enlightnening or maybe you are still busy? can't find your email offering a refund but can you do one or the other please thanks

We here at Long Bets have been impressed with a new offering in the on line prediction area... Predictify.

After speaking with them, they were generous enough to set up a partner area for us here:
http://longnow.predictify.com

Its a great place to make shorter term predictions and or put a bounty on answers about your prediction. They have an excellent and simple interface worth checking out.

This site requires you to pay $50 when you first post a bet, this is counter-intuitive. I should be able to post for a free period to see what reaction it gets then I can decide if it is worth financing.

Why does it take so long for Spam to get removed?

Also for 50 bucks I can post my Web site in the guise of a "Prediction" getting people to visit and prop up my hit counter. Seems to me that should be against the rules