That would mean a lot of additional data to store, it's just simpler to calculate it in real time
Comment has been collapsed.
wouldn't it be less? From what I have seen in the BL/WL charts each data point is a specific indicator. For example, if user A removes me from their WL, the chart drops the WL data point from when user A added me, as if it never happened. That would mean (to my feeble mind) the chart is tracking each specific piece of data and not just the "total."
Of course, this all assumes the same algorithm is used for WL/BL charts as contributor.
Comment has been collapsed.
Like you said it would be like it never happened, because the website just look up how many blacklists you are currently on and when you were added on them. To see wl/bl you have been removed from, we would need to store the info for every user, every day.
Same with level, it just look at when your giveaway ended and their current value. It's all data that was already there before the stats page was created. To show all various level fluctuations that can occur (game not added to the bundle list yet, price glitch, ...), we would have to store the level for each user everytime time there's a change.
Comment has been collapsed.
I don't really get what you mean, it's not that it's hard to make the graph go down, it's just that info about what level you were on a given day doesn't exist.
Comment has been collapsed.
This is the data available :
Based on that, you get an extrapolation of your level evolution. But there is no actual record of your level at a given day. Your level will be calculated based on the current conditions. So if you gave away a bundle game before it was added to the list, or a game had a price drop, it will not be taken into account, so that's why the graph only goes up.
Comment has been collapsed.
Simply put, if it can calculate what lower levels you were at a given point, it's almost negligible to get it to show you what higher points you were, too.
That's the way the tool being used (http://www.highcharts.com) was designed to work.
Comment has been collapsed.
If I give away a $50 CV game, and the price changes to $40 (via Steam API, therefore my CV changes to $40), there is no longer a data point that I gave away a $50 CV game, it is a $40 CV data point after the price change. This is what TempeteJoachim is explaining.
Comment has been collapsed.
There is one technical way to do this without storing additional database data, and that is to make database queries based on dates (which are already stored) and then cross-reference some sort of historical price API -- I think the data may exist in the Steam API game changelog area. I'm not sure what the impact to server performance or site optimization would be to cross-reference this data, but that would be one way to alleviate the issues with fluctuating CV without revamping the database table structure.
Comment has been collapsed.
156 Comments - Last post 18 minutes ago by NebraskaJones
395 Comments - Last post 56 minutes ago by wigglenose
39 Comments - Last post 1 hour ago by Foxhack
284 Comments - Last post 1 hour ago by Wok
8 Comments - Last post 5 hours ago by TheLimeyDragon
1,247 Comments - Last post 9 hours ago by WaxWorm
82 Comments - Last post 10 hours ago by GarlicToast
79 Comments - Last post 1 minute ago by Ottah
1,372 Comments - Last post 1 minute ago by Vasharal
1 Comments - Last post 3 minutes ago by Melliuc
17 Comments - Last post 5 minutes ago by Nayden
28 Comments - Last post 7 minutes ago by juhaszandor
2,827 Comments - Last post 39 minutes ago by ozo2003
778 Comments - Last post 41 minutes ago by Vampus
The contributor level chart adjusts it's metric to reflect your level instead of keeping the same metric and showing the drop on the line like would be expected. It'd be nice if it actually tracked your level, not just the increases.
Comment has been collapsed.