IronAxe
Auto-translated
Let me quote how the third option sounds:
...
I would like to draw your attention to the fact that there is not a word there about HOW to account for the results of all players. After all, we are most likely talking about the arithmetic mean, and not some "Chinese" sum of all results... :)
That is exactly how you understood me correctly!!! =) Of course, I didn't want to just sum up the ratings of all the clan players )))
1) Cumulative rating:
AllRating_new = AllRating_old + (P1 + P2 +...Pn)
where: AllRating_new is the rating after the tournament; AllRating_old is the rating before the tournament; P1, P2,...Pn are the ratings of players participating in the tournament.
i.e. the larger the clan and the more often it participates in tournaments, the higher the rating.
---
AmberSoler and CrasH, what do you mean "not stated how it will be accounted for"?! Kyle's formulas were published a long time ago. Look at the "Chinese" formula above :) Where does it say anything about the arithmetic mean? Huh?
---
2) Average rating:
MeanRating = AllRating_new/N
where: MeanRating is the average rating; N is the number of clan members nominated in this version of heroes (Thanks to AlexeyPank)
i.e. the presence of "dead souls" in the clan is discouraged.
---
Aha, thanks :) Here the Chinese are finally backing down!
Note that the average rating here is calculated not by the number of participants from the clan, but by ALL clan members nominated in this version of heroes—those who played and those who didn't. But just because a person didn't play doesn't mean they've died :) What "dead souls" are we talking about? My aging by 10% compared to what's here is just child's play :)
My opinion is that ONE formula is needed instead of these two!
AllRating_new1 = AllRating_old + (P1 + P2 +...Pn)/N
where N is the number of participants from the clan who ACTUALLY took part in the tournament for this version of heroes.
In this form of the formula, if anyone feels like it, they can add the zeros of those who missed it too :)
Moreover, it will work regardless of whether option 2 or 3 is adopted. The first one definitely doesn't pass. Kyle cleverly split the voting into three points, although many here understand that points 2 and 3 are almost the same; in option 2, only the number of P is limited. Combine these two points into one, and it's immediately obvious that option 1 is out the window :)
However, there is one big downside to this formula. Weak players, even if they aren't kicked out of the clan, will simply be afraid to play in tournaments and send their results, fearing they will ruin the clan rating earned by the clan's strong players in the tournament!
Therefore, I suggest keeping the Chinese formula as well:
AllRating_new2 = AllRating_old + (P1 + P2 +...Pn)
AllRating_new1 - average power of the clan for the version of heroes
AllRating_new2 - total power of the clan for the version of heroes
If anyone wants to think about how to combine the two parts and calculate the GENERAL clan rating, then this is perhaps where we should talk about the oscillator and AlexPank's coefficients. The refinement should be aimed at encouraging mass participation in tournaments from the clan. That is, the more people from the clan come to the tournament, the larger the bonus they should receive for their clan GENERAL rating! The bonus should be reasonable. Perhaps this is exactly where AlexPank's clever bonus will work. AmberSoler and AlexPank—this is your chance to shine :)
Just don't distort the personal ratings of players, which they receive via the non-linear system :)
For example: General = ( AllRating_new1 + AllRating_new2 ) / 2
Will this be very skewed or not? I would like to hear AmberSoler's opinion on this.
My thinking is this: a small, strong clan will have a good average power, but the total might turn out to be less than that of a large clan, if the latter shows corresponding activity. However, a large clan might have a low average power due to the inconsistency of their players. What will the oscillator show? It seems to me that this is exactly where it will work. :) It actually works everywhere, doing what it's supposed to do, BUT it's not logically justified everywhere. I'm curious—in this case—is it justified or not?
---
3) Clan Activity:
ActivityRating = (ActivityTournament *100)/AllTournament
Where ActivityRating is the clan activity (measured in percent...
---
Let AmberSoler comment on the third formula, I'm lost already :)