Forum
Mazaman wins pennant in December baseball-style standings.
|
skrumgaer wrote
at 2:36 PM, Sunday January 2, 2011 EST
The baseball-style standings are explained on my Wall.
GB Player --- Mazaman 06 Gauntlet15 07 gilbe 07 Fatman_x 08 greygoo 09 boogybytes 12 E. T. 14 cool g 15 kostur 16 Azet 17 MB238 20 mreinwald 22 sertach 24 auTs4iDa 24 dirtyrolls 24 Fonias 25 ovbogaert 25 toms 25 pizza_the_hutt 25 clarkebar 25 chooks204 26 jf220 27 caparzo 28 MsCat 28 ghj luigi 28 Mitsi the cat 28 FackFuce 29 NIke 3 29 lindstrom 29 kdiceplaya! 30 what_up23 30 musetel 30 Honyo 31 rocka09 31 Az_Balu 31 kokobongo 32 Leroy_ 32 CuteKittens 32 Marius_1987 32 caesar-blue 33 Brendan Smith 34 hokansoc 36 Known2Own 37 PRESlDENT 37 Wandigo 37 rednttautology 38 wzzrd 38 wenig geil 38 AlphaQupNdown 39 JETS1 41 glasssource 42 Yetihenry 42 Robertass 44 vashthestabde 45 miguelz262 48 EddyB 48 Easy A 48 /wanted 48 Duraj 49 SumSar 50 greekboi 50 Tate73 50 moulue 50 Loobee 50 Gurgi 53 crystal4444 55 habit1 56 22-Apr 57 Tourney Champ 58 snmlmz 58 WarPig11 60 Zoidinator 61 borowka 62 nodice4u 63 Jew Jon 64 yellowfin 64 aismo 66 montecarlo 66 phuze 66 olkainry38 68 trendz 68 Oranges the 1st 72 Tazy 77 Kdice_CPR 77 Kdice_First Aid 81 Connan The Barba 83 kakku man 86 sung2323 90 Big BaII PauI 90 Bleached 91 PRESIDENT N° 1 94 Shevar 94 Andro66 95 its really chase 97 EInegro 99 GreGGwar 105 MadWilly 107 ji-jo 109 jurgen #DIV/0! CCSKAOT |
|
kakku man wrote
at 6:41 AM, Tuesday January 11, 2011 EST im wrote a greasemonkey script that calculates tazd tapl and all that stuff. and i use the formula i posted earlier. wandigo was just an example. i get different numbers for everybody.
for example Gauntlet15 you say 13376. but i get: sqrt(163) * ((((22 - 7.2)^2) / 7.2) + (((22 - 7.9)^2) / 7.9) + (((17 - 10.2)^2) / 10.2) + (((9 - 14)^2) / 14) + (((8 - 17.5)^2) / 17.5) + (((9 - 20.3)^2) / 20.3) + (((9 - 22.9)^2) / 22.9)) = 1 044.24398 |
|
|
skrumgaer wrote
at 7:12 AM, Tuesday January 11, 2011 EST 1. The TAPL and TAZD have multipliers of 1000.
2. If you use whole numbers for percentages (example, 10.2% is shown as 10.2), you DIVIDE the result by the square root of the number of games instead of multiplying. |
|
|
kakku man wrote
at 7:37 AM, Tuesday January 11, 2011 EST ok new try, now with the correct amount of games:
(1 000 * ((((22 - 7.2)^2) / 7.2) + (((22 - 7.9)^2) / 7.9) + (((17 - 10.2)^2) / 10.2) + (((9 - 14)^2) / 14) + (((8 - 17.5)^2) / 17.5) + (((9 - 20.3)^2) / 20.3) + (((9 - 22.9)^2) / 22.9))) / sqrt(262) = 5 053.09582 |
|
|
skrumgaer wrote
at 8:01 AM, Tuesday January 11, 2011 EST Sorry I messed up. If you use percentages (10.2%) multiply by the square root of the number of games. If you use the NUMBER OF GAMES in each parenthesis (17 x 262 -10.2% x 262)^2, divide by the square root of the number of games. If you use percentages x 100 (10.2), multiply by the square root of the number of games and multiply the total by only a factor of 10 instead of 1000.
|
|
|
kakku man wrote
at 8:22 AM, Tuesday January 11, 2011 EST (10 * ((((22 - 7.2)^2) / 7.2) + (((22 - 7.9)^2) / 7.9) + (((17 - 10.2)^2) / 10.2) + (((9 - 14)^2) / 14) + (((8 - 17.5)^2) / 17.5) + (((9 - 20.3)^2) / 20.3) + (((9 - 22.9)^2) / 22.9))) * sqrt(262) = 13239.111 (against your 13376)
close enough or still wrong ? |
|
|
skrumgaer wrote
at 8:33 AM, Tuesday January 11, 2011 EST I am not at my spreadsheet computer to check, but I might also have a corrector for the number of games based on the sum of the player's percentages. Because Ryan truncates the numbers, it usually is less than 100%. In Gauntlet's case it is 96%. So his number of games is underrepresented. Try dividing your result by the square root of 0.96.
|
|
|
kakku man wrote
at 8:41 AM, Tuesday January 11, 2011 EST ((10 * ((((22 - 7.2)^2) / 7.2) + (((22 - 7.9)^2) / 7.9) + (((17 - 10.2)^2) / 10.2) + (((9 - 14)^2) / 14) + (((8 - 17.5)^2) / 17.5) + (((9 - 20.3)^2) / 20.3) + (((9 - 22.9)^2) / 22.9))) * sqrt(262)) / sqrt(.96) = 13512.1111
almost same difference.. |
|
superxchloe wrote
at 10:46 AM, Tuesday January 11, 2011 EST There's usually a small difference between when I calculate the tazd and skrum's published tazd, which I've always written off to rounding when skrum posts his zero datum (ie skrum is using more accurate "expected" values than I am). For Gauntlet, with the correct formula, I also found 13239.11 so the formula isn't the problem kakku.
|
|
|
superxchloe wrote
at 10:47 AM, Tuesday January 11, 2011 EST It's less than a 2% difference. I wouldn't worry about it.
|
|
|
skrumgaer wrote
at 11:54 AM, Tuesday January 11, 2011 EST Winter storm warning in effect and I might not get to my computer today. Try changing the 22.9%'s to 22.0% for seventh place and running it. This was a problem I identified last year but I might have forgotten to get it fixed.
|
