I was wondering if there are any ideas or plans to deal with Astro single target card feeding. FFLogs obviously is way more than the dps leaderboards but it kinda defeats the purpose of competitive dps leaderboards with some people having >50% card uptime (especially since your all star points are heavily influenced by it).
I don’t think making a leaderboard that excludes Astro parses and such (like @Jinzhu mentioned in the past) would work but maybe make the actual dps gain by cards go towards an extra tab like limitbreak in FFLogs instead of someones personal dps? That might even enforce smart card usage and using your cards wisely as you could measure an Astros performance by the cards dps gain as well since e.g. a Blackmage would benefit a lot more from an arrow than a Ninja.
Unfortunately, even though your idea doesn’t sound too bad it’s rather idealistic. In the same way a lot of things are exploited in life, this isn’t very different unfortunately. It ruins the idea of logs in my opinion, but it’s pretty much normal parsing: All the matters really is the number as long as it was achieved within the legal parameters of the game.
I do find it meaningless when peoples allstar points are based off their top parses since the best way to find the true quality of someones dpsing skill is to look at their median (though this could easily be skewed by having a few awful parses if you get unlucky). This is because the top dps is a easily manipulated number whereas the median takes into account the bulk data.
It doesn’t technically ruin the idea of competitive dps charts since everyone can do it, just not everyone is able to do it (no astro, raid leader doesn’t organise parse runs etc.) Every dps has it’s ideal comp (some are more practical than others, as some require one dps to compromise their dps e.g. mch using magic hypercharge on single target) and even without the astro BS people will always find ways to pad, such as leaving adds to one dps, or slowing down dps or stopping completely so someone gets more time to dps.
They get a mean of all their parses, and that number is their ‘true dps’ for that fight while another allstars for their ‘top dps’ is included. The problems is the larger the sample, the more likely to have outliers and the smaller the sample, it will be easily skewed. Also the gearing up and first few clears will obviously be lower due to circumstance, though the first clear could be excluded or maybe it only takes from the recent 10 clears. The most obvious problem though is having have a comp that doesn’t work in your favour. Same as current allstars, if you’re seriously aiming for the top you either have to be exceptional, or have a comp that works for you. If you have neither then well your SOL.
I do get what you mean but then again you could easily inflate your median by not uploading bad parses or simply uploading good parses only.
There will always be padded parses (e.g. only 1 person being allowed to aoe etc.) and I guess there is no reliable way to detect those but I figured it might be different for single target Astro cards or catered comps like double Bard for example.
You might be right tho that at the end of the day it’s rather unrealistic.