Mafia Statistics (Super Cool Table)

[Callid]

So according to the SAMT, I should never be killed or lynched on the first Night/Day, right?  Right?

Well, except they think starting at the bottom is also a good idea...

[TheBlind]

Or you can do what I did and ask Abs., the club manager, to let you be part of it. .

What? You're a great club manager.

My 6 BO partners last Round are located in the last 10 positions

The most dangerous people are on the bottom. Just Abs. and Sakina alone are a tip off that the lower half of the list is the awesome half. .

[Tanner-kun]

Or you can do what I did and ask Abs., the club manager, to let you be part of it. .

What? You're a great club manager.

My 6 BO partners last Round are located in the last 10 positions

The most dangerous people are on the bottom. Just Abs. and Sakina alone are a tip off that the lower half of the list is the awesome half. .

What about us all stars at the top.

[Akonyl]

Rather than your current weighting system (which seems to keep the exact same ranking ordering from win amounts, aside from tie-breaking), I'd suggest something like win amount * win percentage, seems like it gets the best of both worlds imo.

I had another idea, based on yours:
(Win percentage)² x (Number of games played (including GMing))^0.5

Then it would be:


1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Schillok  
pofa  
Akonyl  
Callid  
conia  
Blufberg  
Xcommando  
Beastly  
mangaluva  
c-square  
PhoenixTears  
Rohoph  
Kleene Onigiri  
Commi-Ninja  
Holmes  
Detective Tommy  
Elika12  
Ctu  
Bluekaitou1412  
xpon  
Laurell  
nomemory  
ranger  
xGinx  
KainTheVampire  
Paix672  
Sebolains  
TheBlind  
Abs.  
Chekhov MacGuffin  
James Rye  
Sakina  
1,94381729  
1,837327299  
1,778781184  
1,567673435  
1,530612245  
1,414213562  
1,349873118  
1,138419958  
1,125  
1  
1  
1  
0,923568041  
0,863918795  
0,790569415  
0,769800359  
0,769800359  
0,624647439  
0,612372436  
0,612372436  
0,5  
0,357770876  
0,353553391  
0,353553391  
0,19245009  
0,19245009  
0,19245009  
0,19245009  
0,089442719  
0  
0  
0  


Another idea would be
(Win percentage)² x ((Number of games played (including GMing))^0.5-1);
to deal with the One-Game-One-Win-Guys.

OK, the very new method
(Win percentage)² x (((Number of games played/GMed)-1)^0.5)
would create the following order:

1. Schillok
2. pofa
3. Akonyl
4. conia
5. Callid
6. Xcommando
7. Beastly
8. Blufberg
9. mangaluva
10. Kleene Onigiri
11. Commi-Ninja
12. Holmes
13. Detective Tommy
14. Elika12
15. Ctu
16. Bluekaitou1412
17. xpon
18. Laurell
19. nomemory
20. ranger
21. xGinx
22. KainTheVampire
23. Paix672
24. Sebolains
25. TheBlind
26. Abs.
Chekhov MacGuffin
c-square
James Rye
PhoenixTears
Rohoph
Sakina

Any one who has participated in only one game is not rated (as the second half of the formula becomes zero).

still number 3... still too high imo.

[Sebolains]

Hmm, I'm surprised you didn't simply go for wining percentage. But true, those who've played only one-two rounds would rank pretty high then.
On the other hand, players (like me) who simply aren't able to play in every single game due to RL rank lower than they would otherwise. For example, my winning ratio is much higher than Holmes', but because I missed four games I'm "not as good" as (s)he is. Also, GMing will cause your rank to fall as well, cause you'll also miss a game, despite GMing being much more difficult than playing. Perhaps you could think of a way to do something about those points, like counting GMing as a win or increasing the importance of the winning percentage?

See, the thing is that I wanted to prize playing. In the end, our games are about participation, and I liked the idea that only through participation could one go up in the table. What I mean is that it is impossible to pass other people that are playing, regardless of their poor participation, when one is not playing. That is the main reason why I chose this weighted ratio. However, I do acknowledge that it is far from a perfect system, and I would love for the ultimate decision as to what the WR should be to be community based. I do take very much into consideration your input, as with other inputs, and I'll do my best to find a better system. Right now, I think I'll compile all of the suggested modifications to the WR and will decide via a pole. I would ask from anyone who thinks they have a good idea as to what the WR should be to suggest it in this thread. Previously stated ideas need not be restated. I'll go through them.

Summary: I know my system is very flawed, but I reached that weighted ratio because I like rewarding participation, not only results. However, I'm more than open to changes.

Note: I am really opposed to the idea of including GMed games on the W.R. Why? Because I really don't think it has anything to do with what we are aiming for here. The SAMT is about playing mafia, not GMing it; they're two completely different things. I am extremely grateful to each and every person who ever GMed a mafia game, and I do understand the tremendous effort they made to be able to do that. The SAMT does show who GMed and how many times. I have not ignored your wonderful contribution to our community. However, I do think that the WR should have anything to do with that. If I see that most people disagree to this point, I will change that, I'm just stating my opinion.

Note's Summary: GMing is awesome and I thank you for that, but my opinion is that it should have nothing to do with the WR.

[Abs.]

Obviously GMing has nothing to do with the win ratio, etc.  Because the GM can never win, or lose, a game they GMed.

[xpon]

Dont you agree.. that now we have more than 4 good player?

in earlier round.. i heard someone say kill akonyl or kill schillok because they are experience!.. but now.. almost half the palyer is great deductor and some are good newbie..

only xpon that dont change ... still act like a clown!

[Callid]

Note: I am really opposed to the idea of including GMed games on the W.R. Why? Because I really don't think it has anything to do with what we are aiming for here. The SAMT is about playing mafia, not GMing it; they're two completely different things. I am extremely grateful to each and every person who ever GMed a mafia game, and I do understand the tremendous effort they made to be able to do that. The SAMT does show who GMed and how many times. I have not ignored your wonderful contribution to our community. However, I do think that the WR should have anything to do with that. If I see that most people disagree to this point, I will change that, I'm just stating my opinion.

Note's Summary: GMing is awesome and I thank you for that, but my opinion is that it should have nothing to do with the WR.

You're completely right there, but if the number of games played affects the ratio, and GMing is not counted into them (cause you only count wins and losses), then GMing will affect your ratio in the same way as if you hadn't played. However, the GM did participate, and as you said yourself, this should be considered (or even rewarded), which is why I think including GMing is important, but not as a loss or win, but in a different way (in my suggestion, for example, GMing will change the number of played rounds, but not the winning percentage, so someone who was just as good as you before the round you GMed will be better as you if he won and not as good as you if he lost).

[Sebolains]

I think including GMing is important, but not as a loss or win, but in a different way.

I agree with this statement, but I sort of think it's impossible.

Think about it: Regardless of the way you end up making the GMed games modify the WR, in the end they're always going to increase it, right? I mean, we're not gonna let GMing hurt one's WR. So, if GMing is gonna affect the WR in any way, it will be positive. This is where I disagree. As much as I respect people who have GMed (and as much as I understand that these are all amazing mafia players, undoubtedly amongst the best) I believe it should be left out of the WR, because the latter should only represent the ability to play, not the willingness to GM.

I ask you to please not take any offence to this. I know you've been a GM before (and if I may say so, one one the best  8) ) but this is still my opinion. And let's face it, I didn't choose this WR system to have a higher rank; my rank is terrible either way. I did it because I thought it was fair. Of course, (not to reiterate, but oh well) I do understand that this is not mine nor yours decision to make, but the community's. In the end, it will be be them who will choose the system.

Are you currently set on (Win percentage)² x (((Number of games played/GMed)-1)^0.5) ? Or do you think we should come up with a completely new WR from scratch? I definitely wouldn't be opposed to that. I think you're a smart and great to work with individual.

[Akonyl]

If you're really worried about having a serious-business WR, it's not very hard to incorporate GMing as a game played imo, just coutn it as a game played that doesn't effect your win/loss ratio.

For example:
So, (wins/games played) * (games played + games GMed)

so, let's say I play 5 games, lose 2 and GM one. Then, my WR would be (5/7)*(8), or 5.7.
If, next game, I win: (6/8)*(9), or 6.75.
If, next game, I lose: (5/8)*(9), or 5.625.
If, next game, I GM: (5/7)*(9), or 6.42.

Effectively, it makes games GMed count as much as your average game, thus if you have 100% wins, and you GM a game, you'll get just as much out of it as if you'd actually played in it and won again. And, if you always lose, and you GM a game, it'll act like you just entered a game and lost again. And, if you win some, lose some, like in the example, it counts as much as your "average" game would have gone.

[Sebolains]

If you're really worried about having a serious-business WR, it's not very hard to incorporate GMing as a game played imo, just coutn it as a game played that doesn't effect your win/loss ratio.

For example:
So, (wins/games played) * (games played + games GMed)

so, let's say I play 5 games, lose 2 and GM one. Then, my WR would be (5/7)*(8), or 5.7.
If, next game, I win: (6/8)*(9), or 6.75.
If, next game, I lose: (5/8)*(9), or 5.625.
If, next game, I GM: (5/7)*(9), or 6.42.

Effectively, it makes games GMed count as much as your average game, thus if you have 100% wins, and you GM a game, you'll get just as much out of it as if you'd actually played in it and won again. And, if you always lose, and you GM a game, it'll act like you just entered a game and lost again. And, if you win some, lose some, like in the example, it counts as much as your "average" game would have gone.

I understand that there are multiple ways of incorporation the GMed games into the WR. My point is that, whichever way you end up choosing to do that, the GMed games will affect the WR positively, and I don't like that being the case. I don't want the GMed games to affect the WR because I don't think they have anything to do with each other. But, oh well, if most people think they SHOULD affect the WR, I'll change that.

[Akonyl]

I understand that there are multiple ways of incorporation the GMed games into the WR. My point is that, whichever way you end up choosing to do that, the GMed games will affect the WR positively, and I don't like that being the case. I don't want the GMed games to affect the WR because I don't think they have anything to do with each other. But, oh well, if most people think they SHOULD affect the WR, I'll change that.

it wouldn't really affect it positively, though. Sure, the number goes up, but because it's only counted as an "average game", and the fact that in the current system, WR inflation exists, all the GMed-game inclusion does is it keeps the player from falling behind due to inflation. If everyone else on the table plays, and one player GMs, on average the GM will stay the same ranking.

Though, as I said I don't really care much how the WR is done or if it's there at all. I'm just saying, it's not impossible to include GMed games, imo.

[Sebolains]

I understand that there are multiple ways of incorporation the GMed games into the WR. My point is that, whichever way you end up choosing to do that, the GMed games will affect the WR positively, and I don't like that being the case. I don't want the GMed games to affect the WR because I don't think they have anything to do with each other. But, oh well, if most people think they SHOULD affect the WR, I'll change that.

it wouldn't really affect it positively, though. Sure, the number goes up, but because it's only counted as an "average game", and the fact that in the current system, WR inflation exists, all the GMed-game inclusion does is it keeps the player from falling behind due to inflation. If everyone else on the table plays, and one player GMs, on average the GM will stay the same ranking.

Though, as I said I don't really care much how the WR is done or if it's there at all. I'm just saying, it's not impossible to include GMed games, imo.

Wow, I really liked your take on this. That actually makes a lot of sense. 

[Callid]

If you're really worried about having a serious-business WR, it's not very hard to incorporate GMing as a game played imo, just coutn it as a game played that doesn't effect your win/loss ratio.

For example:
So, (wins/games played) * (games played + games GMed)

so, let's say I play 5 games, lose 2 and GM one. Then, my WR would be (5/7)*(8), or 5.7.
If, next game, I win: (6/8)*(9), or 6.75.
If, next game, I lose: (5/8)*(9), or 5.625.
If, next game, I GM: (5/7)*(9), or 6.42.

Effectively, it makes games GMed count as much as your average game, thus if you have 100% wins, and you GM a game, you'll get just as much out of it as if you'd actually played in it and won again. And, if you always lose, and you GM a game, it'll act like you just entered a game and lost again. And, if you win some, lose some, like in the example, it counts as much as your "average" game would have gone.

That's exactly what my method does, I've only weighted them differently, and now settled for this: (Win percentage)² x (((Number of games played or GMed)-3)^0.5+3). This means al players who have played three or less games are not taken into account, as they create an "math error" (cause there is no square root of negative numbers). However, this is not a bad idea anyway.


I've also thought of the following, complicated new formula (although it bases on my above, it's much better  ;)):

((((W_T/P_T)*(P_T/P_∑)/(G_T/G_∑)) + ((W_B/P_B)*(P_B/P_∑)/(G_B/G_∑)) + ((W_L/P_L)*(P_L/P_∑)/(G_L/G_∑)))/3)² * (((P_∑+P_GM)-3)^0.5+3)

W_T is the number of times the player Won as a Townie.
P_T is the number of times the player Played a Townie.
P_∑ is the number of times the player Played altogether (∑).
G_T is the number of Games the Town won.
G_∑ is the number of Games that were played altogether (∑).
W_B is the number of times the player Won as BO.
P_B is the number of times the player Played BOe.
G_B is the number of Games the BO won.
W_L is the number of times the player Won as a Lover.
P_L is the number of times the player Played a Lover.
G_L is the number of Games the Lovers won.
P_GM is the number of times the Player GMed a game.

So, now I'll explain that monstrous formula. W_T/P_T is the winning quote of the player as a townie. This is then multiplied with P_T/P_∑, the percentage of games the player was a townie. Then the whole thing is divided by G_T/G_∑, witch is the likeliness of the town to win. This terminus is meant to create a number showing how well the player did as a townie weighted by how much of all his games he was a townie (or, you could say, the importance of his town games in relation to his other games) relative to how difficult it was to win as a townie. The same is then done with both BO and Lover games. However, as a always winning player who played all games

Spoiler: special other variant

or whose games had as much Town/BO/Lover-wins relative to the games played as there were in all games relative to all games

would get a "1" for each of the three parts, the total number is divided by three, giving a weighted winning percentage. This can excel 100% (or 1), if a player played once and won, let's say, the only lover win game out of fifteen games that were played, he'd get 500%.
Therefore, we still need to take the number of played games into account, and I decided to pick my weighted variant from above (which doesn't include players who played/GMed less than three times), giving us a really work-out ratio.

Note:
Lover Wins are Lover-only wins! So a lover+townie win is a town win!

[Akonyl]

To your first equation:
Yeah it's basically the same as mine, but I just don't see much of a need for the +/-/sqrt on the # games played. The sqrt I guess, to make # of games not matter as much after a point, but I don't really see the point of the +3-3. It's not like anyone's going to complain about data being skewed at the bottom of the table.

To your second equation:
Yeah admittedly, the chance to win as a townie vs the chance to win as a BO is different, however this is largely inconsequential imo. Not to mention, this equation fails to take into account (as it's really impossible for it to) the effectiveness of town makeups when people win, or whatever variation on the ruleset the GM decides to play. Maybe all the times I play town, the BO is loaded with worthless roles and the town is loaded with good ones, while when you play the BO has gin/vodka/vermouth/tequila/chianti, and the town just has detective boys, kazuha, eri, and a bunch of other "less useful" roles? You would certainly have to play as a "better townie" than I would, but this owuldn't be taken into account. Certainly it's an impressive formula, but I just think it's overly complicated for what small benefit (if any) it would give.