25 october
• Домашняя • Вверх • Rank Rank 0707 • Rank Rank 0706 • RankRat • Rank Rank 05 • Rank Rank • RankRat • 11 October • 11 October • 1 November • 8 Novenber • 15 November • 22 Novenber • 25 october • 6 December •

 

• WWRR •
• БАСКЕТБОЛ •
• ХОККЕЙ •
• ТЕННИС •
• ФОРМУЛА 1 •
• CF •
• БЕНДИ •
• ЭКОНОМИКА •


Гостевая

 

Рейтинг@Mail.ru

Rambler's Top100 Service

271314@mail.ru

ФУТБОЛ

ШАХМАТЫ

У.Е.НОТА

ХОККЕЙ

ТЕННИС

ЭКОНОМИКА

ОЛИМПИАДЫ

 

College Football Ranking Comparison

Saturday, October 25, 2003 (80 Rankings)


Kenneth Massey has made a remarkable thing - has collected about hundred ranking systems of College Football teams. And having assumed, that as behind everyone ranking authors stand and all of them are equal and are the experts as though armed with mathematics. And time so that is naturally simple to calculate certain average ranking, for example, by summing ranks of teams.

Well and as consequence in his compilation of ranking system are located according to a degree of the consent with average ranking.

I have decided to look who has the best ranking system. The criterion the most simple is better the one who makes mistakes less. I hope that my e-rating certainly the best :)).

The amount of mistakes is shown in the table in column Err.

But mistakes are different. One - when 35-l team loses 36-th and other - when 10-I team lose 90-th team. To estimate WEIGHT of a mistake take a difference of places of teams.

WEIGHT = the RANK of the WINNER - the RANK of the DEFEATED

On the other hand the place in the table matters also where the mistake is accomplished. One - were mistaken at the top of the table for example 5-th a have lost 10-th team and other when 95-th team has lost 100-th.

Let's define the IMPORTANCE as the sum of estimations of rank of teams which we shall define simply the N - Rank.

In a result we shall receive

IMPORTANCE = 2*N - the RANK of the WINNER - the RANK DEFEATED

And than VALUE of MISTAKE = WEIGHT "  IMPORTANCE

Here again we come to a problem of an ESTIMATION of RESULT.

As result for the given object of ranking it is natural to take which number of participants it has outstripped. Or the number of participants a minus a rank of the given participant. As some teams meet teams not included in 1À division number of 1 A teams we increase on 1 - 117+1.

In the elementary case it is possible to count, that the ESTIMATION of RESULT is equal to RESULT.

In this case the formula for VALUE of a mistake will be so

VALUE of MISTAKE = (Result of the Winner + Result Defeated) * (Result of the Winner - Result Defeated)

Result of the winner = Number of Participants + 1 - the Rank of the Winner

Result defeated = Number of Participants + 1 - the Rank Defeated

 

E-Rnk - Ranking by this method

E-RNK Link Massey's
Compilation
ABR Err25 Rnk-Err Est-Err Err18 Err this
week
1 Sports 54 TSR 68 2 133 55 13
2 Rothman 20 RTH 73 7 136 57 16
3 Wilson 45 WIL 68 1 136 59 9
4 Sagarin-Elo 44 SE 76 18 144 56 20
5 Self 14 SEL 70 4 148 57 13
6 Greenfield 34 GRN 74 11 148 55 19
7 E-Rating Soft NO E1000 73 19 164 59 14
7 McMurry 48 MMY 76 18 149 64 12
8 Massey BCS 25 MB 73 7 150 58 15
9 Suni 57 SUN 74 11 151 62 12
10 Bobcat 39 BOB 70 4 153 59 11
11 Montgomery 59 MGY 75 14 158 57 18
12 Ashburn 3 ASH 76 18 158 61 15
13 Massey 6 MAS 83 53 159 61 22
14 Sagarin 23 SAG 84 59 160 67 17
15 Mark 65 MRK 77 29 162 67 10
16 Maas 29 MAA 77 29 162 62 15
17 Dolphin 9 DOL 79 38 164 65 14
18 Massey COMP 0 CONS 83 53 164 62 21
18 E-Rating Hard NO E20 76 58 191 59 17
20 Sauceda 22 SAU 81 42 165 66 15
21 Daniel 53 DAN 77 29 165 64 13
22 Bihl 62 BIH 84 59 166 63 21
23 Coffey 4 COF 81 42 166 67 14
24 ARGH 2 ARG 87 67 167 69 18
25 Reese 46 RSE 76 18 168 59 17
26 Wobus 21 WOB 76 18 168 59 17
27 Craig 28 CRD 73 7 168 64 9
28 Matschke 11 MKE 83 53 169 68 15
29 Whitlock 8 WLK 82 49 169 69 13
30 PerformanZ 31 PRZ 85 65 170 69 16
31 Mease 35 MEA 76 18 170 64 12
32 SportsDoc 13 SD 78 34 170 63 15
33 Claassen 1 CLA 81 42 172 63 18
34 E-Rating Real NO E31 75 14 174 58 17
34 E-Rating 16 ER 75 14 174 65 10
36 Sorensen 10 SOR 76 18 175 63 13
37 Bassett 26 BAS 82 49 175 57 25
38 Howell 5 HOW 80 40 177 69 11
39 Boyer-DeSimone