Allright guys. First of all: I am terribly sorry to have left you hanging here for three weeks.
I simply couldn't get my butt up to get it done when I had some time on my hands ... my apologies!
Nevertheless, it's done when it's done so I added the top/most played 6 decks with:
SKIP-stats (
http://elementscommunity.org/forum/index.php/topic,25609.msg349340#msg349340)
Here is the procedure I ended up using:
- Gods that are below a certain FGei(c)-threshold are designated as skip-gods
- The threshold is determined for each god individually and depends on the number
of games played against that god (see twixys suggestion)
-> the more games played, the tighter the threshold
-> the less games played, the more slack a god gets to "stay in"
- The threshold is relative to the FGei(cn) of the deck against all gods (as in kirchj33s posts)
-> e.g. "This gods FGei(c) is more than 20% below the FG(cn)"
- Thresholds:
# of games against a particular god -> individual threshold for this god
15+ -> 20% below = skip
10+ -> 30% below = skip
5+ -> 40% below = skip
<5 -> 50% below = skip
-> e.g. if the decks FGei(cn) was 5210 before implementing skips, a god against whom
15 or more games have been played would have to beat an FGei(c) of 4168, a god with
10 or more games an FGei(c) of 3647 etc.
- If designated a skip-god, all of the gods games are set to "skip"
- All of the gods times for each game are set to 1 second (see 10mens input)
- Skip-gods that didn't make the threshold but made the next, theoretical, threshold in line are
marked yellow in the god-by-god (but they are still skip-gods of course) - Skip-gods that didn't even make the next threshold in line are marked in red in the god by god
-> This is to somehow indicate the level of inefficiency of each skip-god. There are gods who are just way not
worth your time and nobody should ever play them with this deck (red) and there are those that just barely fell
short of making the playable list by a few percent (yellow). E.g., if you consider 3647-4168 electrum/h "good enough"
in a deck that runs in 5210 electrum on average against all gods, then you should go ahead and play the yellow gods ...
Once again, this is not the most aggressive method of simulating the skips as 10mens "progressive" way to
boost the Fast-Draw-Ghostal-deck has shown:
It would be possible to get even higher FGeis by taking a threshold of 0% below FGei(cn), taking out all the
respective gods and then taking out another set of gods in a second, third, fourth ... round, effectively ending
up with only the very most lucrative gods as being "worth your time".
Then again there is still quite a bit of uncertainty that goes along with samples our size: Some 15 games
against a god are good enough to make a call, but they are not really good enough to make a call that draws
the line along a puny couple hundred electrum/h, especially not when considering what a difference there has been
between the players, their overall performance and their performance against specific gods with the very same deck ...
Concluding, the above procedure more or less gives each god a chance to still run in his share and follows the credo
"a god is worth playing until it's pretty damn sure he obviously isn't"
The top six decks now each have a skip-suggestion for 10-15 gods out of 29 which is, I believe, pretty radical after all.
