#1251378 - Sat Nov 30 2019 05:00 AM
Re: ATTN: TEAM VS TEAM STATS
|
Mainstay
Registered: Fri Sep 01 2006
Posts: 711
Loc: Florida USA
|
Some more information about game scoring: Since the number of possible scoring players per league is less than 200 now (A has a max of 160, B has 150 and C has 165) and we want to make sure that quality trumps quantity, the points assignment has been changed a bit, primarily using a top 100 scale:
- The top 10 players for each game (per division) receive a bonus score for their team that is slightly increased from a linear "1 to 100" progression:
1st - 120 points 2nd - 117 points 3rd - 114 points 4th - 111 points 5th - 108 points 6th - 105 points 7th - 102 points 8th - 99 points 9th - 96 points 10th - 93 points.
- From 11th to 100th place, the progression is one point per rank. 11th is worth 90, 12th is 89, 13th is 88, and so on down to 100th which is worth 1 point.
- Everyone else who makes the "Heroes" ranks for their team (top 4, 5, 6 or 8, depending on division) but ranks 101st and beyond still contributes 1 point towards the team standings.
Thus:
- It is ALWAYS strictly better to play than not to play. Using up your full scoring allocation will increase your points, even if only by 1 per player. - The bonus points available at the top make outstanding performances worth going for but not dominant. Compared to a linear allocation up to 100, the extras only make about 2% of the total available daily points. But in a tight race, every point counts, so you want your best players to give it all! - There are NO more games where playing and having a low score can bring your team's scores down. Any such fears (and any "withdraw from the game instead of submitting" strategies) are a thing of the past. We want players to play, enjoy the games and contribute. I think this should be revisited for the coming new year. Rather than "only" getting 160 players, which is the maximum who can get points, and only 99 score more than 1 point, we're getting something like 300 people playing Team Heroes and the Daily Game just in League A, meaning 200 people are getting just one or no points from playing. And I will point out an additional anomaly vis-a-vis the old scoring system of top 200... under the old system, the #1 person would get 200 points, the 21st person 180, making a difference of 10% in their scores. Under the new system, the #1 person gets 120 points, the 21st person just 80, a difference of 33%. The difference in scores yesterday in Piece of Cake, for example, between the #1 scorer and the 21st scorer wasn't 33%, it was less than 1%. So I suggest that you use the actual number of players to determine the number of points, so Daily would get 300 points at the top but the 21st person would get 280, while a game that doesn't count and is gold member like GMM yesterday, the #1 person would get 50 points, the 21st just 30. Alternatively, if you'd like the scores to reflect actual differences in ability, you could just use the raw scores as points. So someone who was 1st in Cake being only 1% better than #21 in Cake would only get 1% more of a score, while someone who is 1st in Obscurity being 38% better than #21 in Obscurity would have the team scores reflect that. This would yield wildly different scores in different games (Mind Melt maxes out at 3000 points to Cake's 1000), but they would be internally consistent from day to day, anyway. None of this would require any change in the number of games or number of players being scored, so League A would stay eight games and eight players, etc. Just the scoring system would change to more accurately reflect ability levels.
Now that I've written it out, I think I like the second option better, just use raw scores for team points rather than rank order.
|
|
Top
|
|
|
|
#1251482 - Sun Dec 01 2019 11:44 AM
Re: ATTN: TEAM VS TEAM STATS
|
Mainstay
Registered: Sat Sep 04 2010
Posts: 722
Loc: Alberta Canada
|
Final Standings for Team vs Team competition are in place but the placement named on the Team pages is out of whack. Mossbacks finished 1st but it reads 19th on their page. Worldwide Wizards were second but it says 7th...which is where the Funsters ended up. Devious Demons says 1st place on their page. It's still connected to the old TvT game rather than the monthly standings.
|
|
Top
|
|
|
|
#1251499 - Sun Dec 01 2019 03:15 PM
Re: ATTN: TEAM VS TEAM STATS
|
Administrator
Registered: Thu Sep 04 2008
Posts: 7583
Loc: Germany
|
HairyBear: For the last time, the scoring table has been chosen and set with a purpose. That purpose is not to reward a specific team or punish any specific team but to reward quality over quantity and provide a good game balance with the current participation figures and those we expect to see over the next few years.
100 out of 160 (assuming a full League A table) is more than 62%. Or, to use the converse, a performance in the bottom 38% of the table will net you only a participation score. Being average (80th) or better will net you at least a 20.
The extra points for top 10, compared to a linear count to 100, add up to 110 (+20 for 1st, +18 for 2nd, and so on down to +2 for 10th). The sum of the base 1 to 100 is 5,050. That means the top 10 bonus is a mere 2.13% of the overall points awarded to the top 100 players. This is a small bonus, mostly intended to provide a little extra motivation to make multiple attempts at the hourlies instead of stopping after getting a score possibly good for 25th or 30th place on your first try "because it won't make a real difference anyway".
We have duly and repeatedly noted that you would prefer a different balance. However we are currently happy with this aspect of the system and see no reason to change it. The old scores which were used in the Team of the Year standings were scaled to a much larger player base and, as a result of declining numbers, rewarded raw player count over actual skill. In particular, being unable to fill your maximum number was practically synonymous with not being able to compete for top spots - in some games, especially the Gold Member ones, participation was down to around 100, meaning that any nonzero score was good for half of what the top player would receive and thus indirectly adding a very undesirable and unintended "pay to win" component to the lists.
As to the bug reports, Terry will get to them when he has the time. Thanks to samak for the observation that the discrepancy only showed on the last day - this might be very helpful in finding the reason.
Edited by WesleyCrusher (Sun Dec 01 2019 03:26 PM)
_________________________
FunTrivia Editor (Hobbies and Sci/Tech) and Administrator Guardian of the Tower
|
|
Top
|
|
|
|
#1251557 - Mon Dec 02 2019 07:51 AM
Re: ATTN: TEAM VS TEAM STATS
|
Enthusiast
Registered: Thu May 10 2012
Posts: 346
Loc: Norfolk England UK
|
As a team we don’t have a problem with the new system. Sometimes odd members complain about a game they are normally not very good at, but everyone has a game they either don’t like or score badly in, I must admit I pulled a face when I saw “Mind Melt” & “Word Wizard” included this month. My view is it’s only for that month and we are all here for a bit of fun.
_________________________
I am so clever that sometimes I don't understand a single word of what I am saying.
|
|
Top
|
|
|
|
#1251563 - Mon Dec 02 2019 10:14 AM
Re: ATTN: TEAM VS TEAM STATS
|
Mainstay
Registered: Thu Jan 15 2009
Posts: 680
Loc: New York USA
|
Final Standings for Team vs Team competition are in place but the placement named on the Team pages is out of whack. Mossbacks finished 1st but it reads 19th on their page. Worldwide Wizards were second but it says 7th...which is where the Funsters ended up. Devious Demons says 1st place on their page. For the good of the system and humanity, I vote that those awards stand as they are. 
_________________________
"We mock what we are to become"
|
|
Top
|
|
|
|
#1251911 - Thu Dec 05 2019 12:31 PM
Re: ATTN: TEAM VS TEAM STATS
|
Head Honcho
Registered: Wed Dec 31 1969
Posts: 21449
Loc: USA
|
|
|
Top
|
|
|
|
#1251917 - Thu Dec 05 2019 12:38 PM
Re: ATTN: TEAM VS TEAM STATS
|
Head Honcho
Registered: Wed Dec 31 1969
Posts: 21449
Loc: USA
|
|
|
Top
|
|
|
|
|
|