chris's dashboard

chris set one of chris' replies as an official response to "Type of gaming and how to stop it." in Hubdub

chris replied on May 08, 2008 11:44 to the problem "Type of gaming and how to stop it." in Hubdub:

Blocking by IP address is a myth. Most people are on dynamic or shared connections. Almost all users at the university of Edinburgh for example connect to us through just 2 IP addresses, whilst our office uses just 1. In addition both me and my flatmate share a connection (which is also dynamic). Blocking further registrations from the same IP would therefore not only fail to solve the problem, it would also certainly affect other users.

We do monitor the creation of new accounts to try to spot trends such as this, and I don't think it's currently a problem. The most likely reason for accounts with 0 balance are accounts which have been created but never had the email address confirmed.

chris replied on May 01, 2008 11:46 to the problem "Is "Will Jesus return by the end of 2008" a good question????" in Hubdub:

It can demonstrated if the return of Jesus is reported by mainstream media sources. I suspect the jaded and cynical journos at CNN (and possibly even Fox) would require some pretty smart miracles to convince them, which is why the odds on yes are so low, and it's not a *great* question certainly, but it's a valid question and very much in the area of "if you don't like it, don't put money on it"

chris replied on April 16, 2008 17:19 to the problem "Congo plane crash settlement" in Hubdub:

Right, having done some further investigation, this was settled on initial reports which claimed to be accurate but have now obviously been revised.

To be fair is pretty unusual for air crashes since airlines generally have good figures about the numbers of people involved. In the case of other types of accident we would have regarded early eyewitness reports as suspect. We should therefore have suspended the question pending more authoritative information, but it being an air crash we assumed the given figures to be more definite than they turned out to be.

The issue is not the same as the merger question, since the press reports there were definitive reports of an actual event (ie. a decision by the company), and the disagreement was caused by a subsequent U-turn. Here the reports were broadly speculative, and therefore were simply not good enough sources to settle the question.

I appreciate your concern crudsy that these types of question are more open to interpretation than sporting or money-market questions, and I can see the appeal for many users in sticking to these well-defined 'logical' markets, and we would like to feel that the site supports them in playing the game their way.

Other users are interested in areas where less well-defined 'logical' questions exist, and we intend to continue serving them as well. As far as possible this will be through providing rules and procedures which make these questions better-defined, but there is always going to be some room for disagreement on certain questions. I guess if this sort of resolution process doesn't appeal to you, that you are best off avoiding these questions

chris replied on April 16, 2008 12:52 to the problem "Congo plane crash settlement" in Hubdub:

Hmm, both situations are looking rather confused at the moment - it may well come to which happened first. Suspending may have been a better solution, but I think it's probably best if we leave it settled for now and resettle if the bus crash in a day or so if necessary.

chris replied on April 13, 2008 21:20 to the question "How do we deal with markets with "highly" subjective language?" in Hubdub:

I think objective might be rather difficult to define. I think avoiding contentious language would be better. I think offhand use of the word 'traitor' to describe Jimmy Carter is extremely contentious, and probably dangerously close to libel, so I would feel quite comfortable removing it.

In any case it adds unecessarily to the length of the question, and incorrectly capitalizes the leading 't' in 'Traitor', so is just begging for an editing.

A comment on the question "Any objections to me making predictions?" in Hubdub:

But I work *all* hours. My idea of an hour off is to create questions on Brazilian football :( – chris, on April 13, 2008 13:00

chris replied on April 02, 2008 22:00 to the discussion "Brand New Leaderboards Live!" in Hubdub:

It is, but the code which showed the update time on the website was broken. it's fixed now. thanks

chris replied on April 01, 2008 18:01 to the discussion "Brand New Leaderboards Live!" in Hubdub:

Right. (page recently updated):
- For both quarterly gains and weekly gains we gather together all predictions which were open at any point in the quarter/week
- For each of these we calculate an opening value, which is either the sale value of the prediction at the start of the week/quarter, or the amount wagered for new predictions
- We also calculate a closing value, which is either the current value (for open predictions in the current week/quarter), the sale value of the prediction at the end of the applicable quarter (for previous periods), or the close value (win value for wins or 0 for losses) where the prediction was settled in that quarter
- For the quarterly values we just total up the differences between these opening and closing values, and give you the result
- For the weekly values we total up the differences, and then divide this by the total of the opening values
- For the categories we just count predictions in that category.
- For the weekly game, if you do not have enough open predictions during the week, you do not qualify for inclusion in the leaderboards.

We would like to reflect the amount of the net worth invested in the weekly totals, in order to further reward people playing the game, and are thinking of either making the weekly boards gain/net worth, or something like if you don't invest x% of your net worth, you do not appear in the weekly boards (with x being 20% for the category boards and 40% for the others). We don't really want to do either of these until we have a scheme in place for people to move their excess wealth out of the game though...

chris replied on March 22, 2008 17:17 to the problem "Large user images on profiles" in Hubdub:

Good spot. I've got some code to fix this. will drop it in shortly.
Cheers

chris replied on March 16, 2008 22:44 to the question "Grand Theft Auto - How Much Violence is it Responsible For?" in Hubdub:

I'm exactly with Ryan on this. I'd have said starting odds should have been nearer 20%, but meh, it's running now

chris replied on March 12, 2008 14:07 to the question "Escalation: What color is this tie?" in Hubdub:

that is a pink tie. an ugly ugly pink tie.

chris set one of chris' replies as an official response to "multiple settlements on same question" in Hubdub

chris replied on March 11, 2008 15:22 to the problem "multiple settlements on same question" in Hubdub:

Correct wager only ever pays out once, and it was actually settled the same way 4 times, so nobody gained or lost anything at all.

The reason it was done was because we were trying to resolve an issue with the cutoff time.

chris marked one of Nigel Eccles' replies in Hubdub as useful. Nigel Eccles replied to the question "Contentious questions (part 3)".

chris replied on March 10, 2008 19:31 to the problem "Is there a time zone issue?" in Hubdub:

done

chris replied on March 10, 2008 19:30 to the problem "Is there a time zone issue?" in Hubdub:

done

chris replied on March 10, 2008 19:28 to the question "Ridiculous suspend dates!" in Hubdub:

I would have preferred the Chinese Democracy question to have a slightly shorter end date since very few people will keep their money in for the whole life of the question and it would have had some validity as a question closing at the end of next year.

It has always been our intention to have long-running questions such as these though, where people are able to buy and sell predictions based on the story as it develops, much like the financial markets in long-lifed bonds, where the original purchaser is very unlikely to be the owner at payout time. In this case for example there isn't really much chance of the release happening in the next year but I might like to take a bet at a low percentage on the basis that many bands end up reforming and putting out disappointing new material. An announcement of a tour might then encourage other people to make a prediction, moving the market so that I can sell my original prediction at a profit without having to wait for the close.

The fact that the odds cannot easily be guessed is what makes prediction markets valuable - effectively our percentage should factor in the considered views of a lot of people on the basis of all the information available at the time.

I don't understand what your objection to them is. As with all markets which don't interest me, I simply don't play them

chris replied on March 10, 2008 18:32 to the problem "Is there a time zone issue?" in Hubdub:

My bad - I didn't realise the flag stayed in place. It's sorted now

chris set one of chris' replies as an official response to "Is there a time zone issue?" in Hubdub