• Hi Guest, The software has been updated but I have not had a chance to tweak anything yet.
    It took longer than I had hoped, so I just turned it on and hope everything is OK
    If you spot anything that does not look rigfhyt then please let me know.
    Ark Royal
  • There seems to be a problem with some alerts not being emailed to members. I have told the hosts and they are investigating.
  • Hi Guest, If you are seeing that Lurker has appeared under your name then please take a look here to see why. AR

PAUL STEELE POWER RATINGS.

Delboy99

Gelding
..... the fixtures appear to be cut up into chunks for coming matches, so look out for dates:-
ENGLISH PREMIER LEAGUE.
8th Feb - Everton 10.221 v Crystal P 9.703 .518 1
Brighton & H A 9.419 v Watford 9.916-.497 2
9th Feb - Sheffield U 10.248 v Bournemouth 9.054 1.194 1
Manchester C 10.995 v West Ham U 9.433 1.562 1
14th Feb- Wolverhampton W 10.396 v Leicester C 10.570 -.174 2
15th Feb- Southampton T 10.037 v Burnley 10.076 -.039 X
Norwich C 9.187 v Liverpool 12.074 2.887 2
16th Feb - Aston V 9.140 v Tottenham H 10.517 -1.377 2
Arsenal 10.035 v Newcastle U 9.612 .423 1
17th Feb - Chelsea 10.167 v Manchester U 10.127 .040 X
Two matches there look tight, Southampton T v Burnley and Chelsea v Man U. ;)
 

Delboy99

Gelding
..... the Paul Steele Power Ratings for 22nd/23rd February:-
ENGLISH PREMIER LEAGUE.
Chelsea 9.881 v Tottenham H 10.501 -.620 2
Burnley 10.227 v Bournemouth 9.109 1.118 1
Crystal P 9.549 v Newcastle U 9.196 .353 1
Sheffield U 10.193 v Brighton & H A 9.430 .743 1
Southampton T 9.886 v Aston V 9.156 .730 1
Leicester C 10.580 v Manchester C 11.019 -.439 2
Manchester U 10.413 v Watford 9.885 .528 1
Wolverhampton W 10.387 v Norwich C 9.392 .995 1
Arsenal 10.451 v Everton 10.375 .076 1
Liverpool 11.869 v West Ham U 9.409 2.460 1
The closet match I can see there appears to be Arsenal v Everton. ;)
 

Delboy99

Gelding
Here are this week's Paul Steele Power Ratings:-
ENGLISH PREMIER LEAGUE.
Norwich C 9.173 v Leicester C 10.479 -1.306 2
Brighton & H A 9.574 v Crystal P 9.599 -.025 X
Bournemouth 8.905 v Chelsea 10.052 -1.147 2
Newcastle U 9.146 v Burnley 10.431 - 1.285 2
West Ham U 9.498 v Southampton T 10.014 -.516 2
Watford 9.607 v Liverpool 11.780 -2.173 2
Everton 10.291 v Manchester U 10.691 -.400 2
Tottenham H 10.330 v Wolverhampton W 10.606 -.276 2
Seems to me to be a glut of aways this week. ;)
 

Delboy99

Gelding
..... OK here are this weekend's Paul Steele Power Ratings:-
ENGLISH PREMIER LEAGUE.
Liverpool 11.151 v Bournemouth 9.017 2.134 1
Arsenal 10.535 v West Ham U 9.781 .754 1
Crystal P 9.752 v Watford 10.127 -.375 2
Sheffield U 10.069 v Norwich C 9.430 .639 1
Southampton T 9.835 v Newcastle U 9.275 .610 1
Wolverhampton W 10.728 v Brighton & H A 9.421 1.307 1
Burnley 10.302 v Tottenham H 10.208 .099 1
Chelsea 9.940 v Everton 10.310 -.370 2
Manchester U 10.672 v Manchester C 11.120 -.448 2
Leicester C 10.222 v Aston V 9.028 1.194 1
Not many close calls there, probably Burnley v Tottenham looks tighter. ;)
 

Delboy99

Gelding
.... as the Man City v Arsenal game has been called off for tonight, I can now complete this week's Paul Steele Ratings:-
ENGLISH PREMIER LEAGUE
Watford 9.986 v Leicester C 10.542 -.556 2
Bournemouth 9.190 v Crystal P 9.873 -.703 2
Brighton & H A 9.616 v Arsenal 10.535 -.919 2
Manchester C 10.845 v Burnley 10.259 .586 1
Newcastle U 9.508 v Sheffield U 10.083 -.575 2
Norwich C 9.416 v Southampton T 9.653 -.237 2
Aston V 8.708 v Chelsea 10.455 -1.747 2
West Ham U 9.781 v Wolverhampton W 10.533 -.752 2
Tottenham H 10.251 v Manchester U 10.947 -.696 2
Everton 9.795 v Liverpool 10.798 -1.003 2
The Norwich game looks to be the tightest around this week. ;)
 
Interesting alternative to elo ratings.
Some painstaking calibration is needed though I should think -using past season data.
Also can be done separately for France, Germany, Italy ...
Looks to me this should take three or for months to calbrate with accuracy.
Also you should drop the requirement that they all start level. You could start the teams in August when the league starts from where they were last year. Give to the newcomers the points of those who relegated last season. That looks more like it.

But the bookies also look into the parallel information as well.
If the stadium is empty what happens ?
If a key player is missing (such as Messi from Barcelona) ?
If there is no motive ?
If, if, if ...
There are many situations and motivation is an all imporant factor. To give you an example, the match between Manchester City and Basle last year. City won 4-0 in Switzerland then in the second leg at home they went to sleep - lost 1-2 but it did n't matter. Many more like it.

The bookies employ people all over the world and they share those correspondents presumably.
I used to know one who was responsible for the Greek 2nd division and another one who was responsible for the English-Scottish amateur leagues.
The Greek one was very good. He also maintained a fantastic telephone line for racing tips.
 

Delboy99

Gelding
..... Certainly is a different perspective you have there and I suppose one could go on forever searching for the right mix on what goes into the making of a really good rating system. The method employed here however is gleaned from Paul Steel's book which I talk about in the introduction of this thread, where he experiments with twelve systems and their merits. The system adopted here is from one of those methods and what looked to be the best percentage success rate at the time of reading his book. I think as far as further experimentation on empty stadiums and missing key players is best left to those who have time on their hands and of course the dedication and will power to continue in this vein. Certainly not my cup of tea I'm quite happy churning ratings out weekly in a current routine under normal conditions. ;)
 
..... Certainly is a different perspective you have there and I suppose one could go on forever searching for the right mix on what goes into the making of a really good rating system. The method employed here however is gleaned from Paul Steel's book which I talk about in the introduction of this thread, where he experiments with twelve systems and their merits. The system adopted here is from one of those methods and what looked to be the best percentage success rate at the time of reading his book. I think as far as further experimentation on empty stadiums and missing key players is best left to those who have time on their hands and of course the dedication and will power to continue in this vein. Certainly not my cup of tea I'm quite happy churning ratings out weekly in a current routine under normal conditions. ;)
I always measure percentage success in whatever I 'm doing.
Otherwise it is like chasing the Bismarck without navigational instruments.
What is the "percentage success" and how is it measured in Steel's book ?
I don't mind measuring percentage success with another method to the one I 'm used to as long as we all use the same method.
 

Delboy99

Gelding
.... looking at Steele's book he called this system The Jewel In The Crown. Of the twelve systems tested it came first overall, 1st in predicting home wins, 1st in predicting aways, 2nd in predicting draws. Home wins 73.43%, Aways 55.03% with draws a lowly second at 30.22%. There is an absolute wealth of statistics in the book in batches of three yearly figures 1996-1999 and even goals scored frequencies in the same period. If one has the time and ability and enthusiasm to look back, it's all there! ;)
 
.... looking at Steele's book he called this system The Jewel In The Crown. Of the twelve systems tested it came first overall, 1st in predicting home wins, 1st in predicting aways, 2nd in predicting draws. Home wins 73.43%, Aways 55.03% with draws a lowly second at 30.22%. There is an absolute wealth of statistics in the book in batches of three yearly figures 1996-1999 and even goals scored frequencies in the same period. If one has the time and ability and enthusiasm to look back, it's all there! ;)

Where can one buy this book ?
I 'm going to compute percentages like yours from the sample of matches I last used and I will let you know.
How do you "predict" draws ? Without conversion to probabilities for 1-2-X when is a draw "predicted" ?
I can compute % of first chance wins. Also % of first chance wins when the first chance is a home win, % of first chance wins when the first chance is an away win and % wins when the first chance is a draw (but the last type of situation is unussual).
 

Delboy99

Gelding
..... I just Googled and saw 1 on e-bay at£21.46 but in USA, the other was on Amazon £13.46 so that looks the better bet. It is called "The Essential Football Betting Guide by Paul Steele, and is dated 1/12/99 20 years it has been around! Wow it hurts! Hope this helps. ;)
 
.... looking at Steele's book he called this system The Jewel In The Crown. Of the twelve systems tested it came first overall, 1st in predicting home wins, 1st in predicting aways, 2nd in predicting draws. Home wins 73.43%, Aways 55.03% with draws a lowly second at 30.22%. There is an absolute wealth of statistics in the book in batches of three yearly figures 1996-1999 and even goals scored frequencies in the same period. If one has the time and ability and enthusiasm to look back, it's all there! ;)
The sample of matches I last used consists of 1782 matches from EPL, FRANCE, SCOTLAND, SPAIN, SWEDEN, TURKEY, BULGARIA and BELGIUM.
The percentage of first chances coming true is 54.8%, first chance home wins when the first chance is a home win 56.9%, first chance away wins when the first chance is an away win 49.8%. There were no cases when the draw was the first chance (a "first chance" starts from 34-33-33 up to 100-0-0).
The probabilistIc parameter predictivity is 38.5%.
Steel's figure of 73.43% sounds to me like a physical impossibility (and I don't think in the year 2000 football was so tame !).
 

Delboy99

Gelding
..... seems like a lifetime since I was making notes on here. Still not a lot we can do but wait until the day it is back among us. Let us hope that the wait will be much sooner than later and it is well worth the wait. ;)
 
Top