Penetration Ratings

[cavello]

Below are the top 5-6 in 3 divisions based on my pure penetration ratings. These are based on the last 6 home and away games and are closely related to the method developed by Paul Steel. I call them pure because I don't factor in goals and base the ratings on shots and corners. The data is available from http://www.365stats.com/ :ok :clap . This site cannot be praise too highly:clap . Stacks of useful information particularly the shots and corner info which allows deeper insight into a teams ability or form.

PREMIER LEAGUE PENETRATION RATINGS<?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" />

TEAM

ABS H

ABS H

RATING

LIVERPOOL

1.98

1.64

362

ARSENAL

2.38

0.4

278

MAN UNITED

1.06

1.56

262

CHELSEA

1.42

0.38

180

MAN CITY

0.86

-0.16

70

PRIMERA LEAGUE PENETRATION RATINGS

TEAM

ABS H

ABS A

RATING

A MADRID

1.4

0

140

SEVILLA

1.08

0.08

116

MALLORCA

0.86

0.16

102

BARCELONA

0.72

0.2

92

DEPORTIVO

0.5

0.06

56

BUNDESLIGA

PENETRATION RATINGS

TEAM

ABS H

ABS H

RATING

B MUNICH

1.46

0.6

206

NURNBERG

1.06

0.1

116

STUTTGART

0.94

0.18

112

W BREMEN

0.44

0.68

112

HAMBURG

1.2

-0.18

102

SCHALKE

0.9

-0.08

82

Obviously goals decide matches but these rating may show teams experiencing a degree of bad luck or lacking in the striker department. Apart from being able to produce a league I can enter the relevant data for 2 teams and get a picture based on pure penetration of the relative strengths of both sides. Useful for leagues where I can't be bothered to produce League tables, but want to analyze a certain match. Does anyone here apart from Paul, develop or use ratings and how do you relate these to the odds?

[AndyR]

Re: Penetration Ratings Hi, Played around with these last season using Paul Steele's method (modified slightly). He used the following calculation to work out the number of goals a team SHOULD have scored and I put these into the spreadsheet developed by mronemore (http://www.punterslounge.com/forum/showthread.php?t=12655&highlight=metric). The calculation was ((shots off target + corners)+2(shots on target))/10 plus add an extra goal if the team wins. Putting the penetration scores from each game into the spreadsheet gave power ratings, which I then compared against a table/graph put up by Muppet in the same thread for power ratings based on actual goals scored. This was pretty accurate in predicting correct home odds (yield was 6%), and betting on discrepancies appeared to work well, but I couldn't get it to work for away scores (I didn't care about draws, they are just too damn hard), this could be because I retained the 0.4 goal home advantage in the spreadsheet. I wasn't too confident that it wasn't purely luck that I made a profit. When I tried to increase the number of games the spreadsheet covered it crashed my computer, so I couldn't back test it over entire seasons. Maybe you could do something with it?! GL Andy