chris's dashboard

Fixed back to PST now. There'll be no stupid Daylight Savings round here

Dammit, forgot to check that in the server configuration, and what with this crazy moving daylight savings forward in the US thing it wasn't even on my horizon.

I think the best plan is probably to force the server to stay on PST all year round, and have people take account of that when they set the settlement times, but I'm open to suggestions (I doubt a significant number of settlements will be affected before tomorrow if I don't fix it until then)

Opinions people please?

It wasn't ever down - I've been using it constantly all day. Oh well...

I've fixed this now

In this case clarification wouldn't have helped - i'm a cricket fan and know the difference between a test and a series quite well. It's really more a case of I-shouldn't-be-allowed-near-the-settle-button. But yes, credible sources which are covering what's intended, and a little background would be nice.

Interesting link btw - I'd never have thought to look at a espn site for cricket, but then I like my cricket coverage biased towards England so I always go straight for the BBC

I'd be absolutely fine with in-game football running until that late. Indeed I'd be happy with it running until the final whistle. If you bet before the start of the game and are not in a position to participate during the game, you've lost nothing compared to if we had closed the market earlier. As for participants during the game, they are just benefitting to the extent that there's a market already open and ready to play.

Think about the logic of the game. If there is a strong team against a weak team, then pre-game betting will mean that even before the game the odds will be at 80-90%, so the movement during the game will be small. If it is more evenly matched then the odds would be nearer to 50%, so the overall movement up until the final whistle would be larger, but there is no single point at which things are decided. sure if a team scores their odds will likely increase, but even bad teams sometimes overcome a goal disadvantage to draw, so they won't go to 100% immediately. As time passes of course the chances of them making back that goal diminish, and the odds move accordingly, and if they concede again, they will once again move accordingly, but as long as there are enough in-game participants to keep the market liquid, there is little reason to believe that anybody can take an unfair advantage by placing a large bet at an ideal moment, purely and simply because there is no such ideal moment - whenever they do it until the final whistle they are still taking a gamble, and it's a gamble based on the odds available to them.

For an example of that look at your own football games. All the late post-season games went right down to the nail apart from the San Diego one, so reasonable odds were available until seconds before the whistle, whereas the ones which were decided earlier had odds of 95% or more for the whole last quarter (roughly) - I was there trying to make a sneaky buck, and there wasn't a sneaky buck to be made...

Agreed. I didn't read the question properly :(
I'll correct it in the morning

I imagine this was deliberate - I'd find it interesting to see in-play gaming on a political question, and would actually have set cutoff to closing time on the dot. If you've flagged it I'm sure ryan will get round to addressing whether it is deliberate or not tomorrow.

Okay, amounts are now updated on wager and cash in

It's *not* a bug - half of the work of producing a number of our pages was generating those numbers, so we introduced caching. This means that the numbers are now only updated every 2 minutes. This has had a positive effect on the servers and is therefore a *good* thing.

In order to improve things a little I fully intend to implement some code soon so that the amounts are updated when you wager or cash in, but sadly I'm just too busy updating you on getsatisfaction to get any work done :P

%#*#. I did do the credit correctly, but I didn't change the win amount. how annoying. I've corrected it now...

Admittedly a closer look at the questions revealed 5 on this week's episode alone. Personally I think that's overkill, and I would be inclined to only accept one per week, but I'm not really sufficiently up with settlement issues right now to make that call. I guess the same applies to in-game market trading - if a market is not well-traded we could call a start or middle of day closing time.

Personally I find it all the in-game trading rather perplexing. Although it's clearly not cheating it requires being available to place predictions for much more time than I (and I guess most users) can really afford to be on the site. I guess from a competition point of view it's a bit like expecting enthusiastic amateurs to compete with professionals at the top of their game. I know everybody likes to dream of taking on the top guys, but imagine if Rocky 6 was in the real world - there would have been mashed up old man all over the ring, and that wouldn't have been fun for anybody...

I changed it, and manually credited your gain just before I posted the last response

My feeling on this is that we do not want to discourage international users at this stage.

Certain questions such as these (or in fact the questions I have posted on scottish football) haven't attracted a massive amount of interest, and could therefore be gamed. Part of the reason for this lack of interest is lack of users interested in that area, and part of it is that the North American questions dominate the listings, so interested users (around half the user base is not north-american) can't find them. In any event I don't really want to prevent them from settling, because then the users that have shown an interest will be put off the site and we will be unable to grow.

In a way the disincentive they provide by letting people make large amounts of money is no more than the powerful disincentive to play that comes to many users from seeing the leaderboard dominated by day traders who are not restricted to playing for a short time each day.

Once we have the infrastructure in place, and have reached a critical mass of users in a given market, we should be able to allow some user settlement of questions, which will help us in the area of questions we don't understand.

In the end I think more segmentation, into trade types, national leagues, and friends or group play, combined with better-directed question listings, is likely to allow us to expand whilst still providing an interesting game for all the players.

Hi

The friends functionality is not actually related to the caching of the current predictions value - the change just happened at the same time. At the moment it stores the value for a maximum of 2 minutes (or should do). I'm planning to add in code to at least make it update when you buy or sell a prediction. I may eventually be able to make it update on any change in price, though this is a whole world more difficult.

The blog server is currently located in the UK rather than the US, though that really shouldn't cause a problem. try www.groopit.com (which is on the same server for now). Otherwise I guess it's something like proxying. Also, I think the site may come up blank if javascript is disabled.

Congratulations on having a working Vista installation though - our windows users just gave up and went back to XP...

Sorry. the cutoff date for the question was set too early. I've fixed this for you