- This topic has 98 replies, 6 voices, and was last updated 14 years, 6 months ago by
.
- Posts
[attachment=0:d1uawrmm]FFOS LAS_11-07-2011.jpg[/attachment:d1uawrmm]
Attached are my computer for Ffos Las today. Forecast 1 is first choice and Forecast 2 is second.
Lets see how well it does later…
Right database sorted.
My program downloads all the data for the days races then produces the forecasts by looking at the distance/going/course and horse/jockey/trainer results from the past few months results. The % column is the "score" for each horse.
Different rules are applied to different courses based on past results (this is an ongoing process).
My aim is not to predict winners but to identify horses that will get placed so I next look at the days forecasts and concentrate on the courses with the highest percentage of horses placed, then look for odds over 5/1 and put an E/W patent (or sometimes a yankee) on the three selected horses. Somedays I do mot bet at all, others I may have just one bet and others I may have more – it all depends on what the odds are in the forecasts and I do tend to go for races with 8 or more horses but do choose thoses in smaller races when the % for the first choice is a lot bigger than the second one.
Until today the "Forecast Stats" were a bit wrong as old forecasts (produced before some algorithm changes) are included, I have just cleaned up this data.
Below the two top courses for today.
[attachment=1:3g3uvr7g]KILLARNEY (IRE)_14-07-2011.jpg[/attachment:3g3uvr7g][attachment=0:3g3uvr7g]LEOPARDSTOWN (IRE)_14-07-2011.jpg[/attachment:3g3uvr7g]
Ineresting thread, thought you would be better using it for the placepot good luck with it,i like the computers picks have a few old spectrum ones myself
Ineresting thread, thought you would be better using it for the placepot good luck with it,i like the computers picks have a few old spectrum ones myself
Nice idea, I hadn’t thought about the placepot. Just looked back through my past results and best I have done is get 5 placed horses in my first selection on cards with 6,7 or 8 races.
EDIT: Just looked at my prediction for Southwell 12th July and perming selections 1 and 2 would have won the place pot. Will have to add an analysis program to predict where my selections would have won the placepot when permed. Thanks. Edit ends.I am using my selections for a tipping contest on another forum and am in pocket placing e/w bets though I have to admit backing Wychwoods Mist at Southwell helped a lot.
Just looked at my prediction for Southwell 12th July and perming selections 1 and 2 would have won the place pot.
I’ve just looked at your predictions for Southwell 12th July as well.
Sorry to have to draw this to your attention, but you wouldn’t have won the Placepot as your selections in the 7:40 (Leg 4), Majestic Millie & Two Bucks More were both unplaced.
As I said, sorry to have to point this out to you but I’m a bit of a stickler for detail.
Just looked at my prediction for Southwell 12th July and perming selections 1 and 2 would have won the place pot.
I’ve just looked at your predictions for Southwell 12th July as well.
Sorry to have to draw this to your attention, but you wouldn’t have won the Placepot as your selections in the 7:40 (Leg 4), Majestic Millie & Two Bucks More were both unplaced.
As I said, sorry to have to point this out to you but I’m a bit of a stickler for detail.
Yep you got me there – I should I made sure the forecast I posted was the same as what I have here.
And I also need to lern to count – good job I can get the computer to add up.Just looked at my prediction for Southwell 12th July and perming selections 1 and 2 would have won the place pot.
I’ve just looked at your predictions for Southwell 12th July as well.
Sorry to have to draw this to your attention, but you wouldn’t have won the Placepot as your selections in the 7:40 (Leg 4), Majestic Millie & Two Bucks More were both unplaced.
As I said, sorry to have to point this out to you but I’m a bit of a stickler for detail.
Hi discretedata, Can you please double check my forecast for Leopardstown today regarding perming the selections for the placepot?
I had quite a nice patent on some of the placed horses.
Ta.
As well as considering place pot also worth looking at the quad pot? I know returns nowhere near as good but perhaps worth assessing what the "past performance" would have been?
yorkiedips
[attachment=1:mwo1gr21]HAMILTON_15-07-2011.jpg[/attachment:mwo1gr21][attachment=0:mwo1gr21]HAYDOCK_15-07-2011.jpg[/attachment:mwo1gr21]
[attachment=1:1quxetxh]HAMILTON_15-07-2011.jpg[/attachment:1quxetxh][attachment=0:1quxetxh]HAYDOCK_15-07-2011.jpg[/attachment:1quxetxh]
May ask 1 question does the PROGRAM only give 2 selections per RACE no matter how many RUNNERS
I also run a SOFTWARE program on here that I’ve USED ON AND OFF since 1995 but relised it does far BETTER IN CLASS 1-3 races and if one as THE PATIENCE THE OUTSIDERS SHOW VERY GOOD PROFITS providing one does not mind LONG LOSING RUNS.
When winners come along do you record RACE CLASS AND TYPE these come in as you may find over A PERIOD OF TIME some may out shine THE REST and make decent profits compared with every RACE SHOWN.
I’ve a very OLD BASIC PROGRAM that in my mind is not much better than picking them with a pin YET THIS CAN HAVE FANTASTIC DAYS and from TOP THREE IN HANDICAPS FIND SOME VERY LARGE F/CAST [ 100 + RETURNS] but would I still gets as many by placing all the RUNNERS IN A BAG and picking 3 out ???
I also have what one would call a very intense 1 were I can set all my weightings and with everything ELSE INCLUDED can have up to 18 different imputs IF I WISH but how well this works may WELL DEPEND on how I weight whats used.
If yours WORKS for how you want to use it all well and good and I wish you well with the thread and at LEAST YOURS GIVES PREDICTIONS EVERY DAY were mines VERY LIMITED depending on CLASS
% MAN
My program predicts (or attempts to anyway) the position of every horse in the race. The forecast cards only show the top two choices but as I have written the program myself I can look at past predictions and look to see if there are any trends. I record race class, type, weights, distance, age, going etc so I can also look for trends here.
I also get the program to pick two horses at random for every race and its hit rate is lower.
At the end of the day as you say how well it works depends how I weight everything.
I haven’t started to look a race class yet as I have not enough data yet. I am able to run a forecast on any race over the last 10 months but as I am spending 8 hours a day developing the program I just don’t have the time to make any forecast data for bygone races.
The weird thing is that my program does very well on some courses but is miserable on others. Do you find that with yours? Maybe it depends on what colour of knickers the head waitress behind the bar is wearing?
My program predicts (or attempts to anyway) the position of every horse in the race. The forecast cards only show the top two choices but as I have written the program myself I can look at past predictions and look to see if there are any trends. I record race class, type, weights, distance, age, going etc so I can also look for trends here.
I also get the program to pick two horses at random for every race and its hit rate is lower.
At the end of the day as you say how well it works depends how I weight everything.
I haven’t started to look a race class yet as I have not enough data yet. I am able to run a forecast on any race over the last 10 months but as I am spending 8 hours a day developing the program I just don’t have the time to make any forecast data for bygone races.
The weird thing is that my program does very well on some courses but is miserable on others. Do you find that with yours? Maybe it depends on what colour of knickers the head waitress behind the bar is wearing?
I commend you for writing your own SOFTWARE PROGRAM and no doubt in some respects work in progress.
As I was never brought up with computers but always interested in stats or ratings 2 of the PROGRAMS I have are dedicated to ratings and used both many times over the years but the one I USE HERE is a lot easier and less work than the other one and I HAVE NO SAY WHATSOEVER IT WHATS PICKED but just LIKE WHAT IT USES
As for courses I dont take much notice to be honest but ONLY USE IT IN CLASS 1 – 3 RACES AND UNLESS A MAJOR MIDWEEK MEETING THAT FITS THE BILL the likely updates could be FRIDAY but mainly SATURDAY and in my eyes as well as others not the easiest days to predict winners never mind OUTSIDERS
After saying that maybe with the RACES BEING MORE COMPETATIVE always a chance of an OUTSIDER being found and were the SOFTWARE EXCELS ITSELF AND WHY I USE IT
Although I show mainly 3 selections per race the SOFTWARE never picks more than 3 in HANDICAPS no matter how many runners anyway 4 in non/h.
Also its the only program I have that I dont now how it works IN HANDICAPS but getting nearer at predicting what it will pick BUT ALWAYS THROWS 1 UP I’m miles away.
It works for me but I’m not interested in FAVS or low priced horses anyway even though it could still pick them.
Also the program as no IDEA of the prices of any horse but just WORKS OF 4 MAIN RATINGS which in my mind are THE RIGHT ONES ANYWAY as well as asking all the right questions like going form c or d winner as shown in POSTDATS which maynot always be an actual COURSE winner but must have a TICK in that box.
Due to many personal things this year it not had every UPDATE thats been shown elsewhere SO MANY WINNING OUTSIDERS MISSED but hope from now on it will show every 1.
% MAN
As I wanted ease of use when run my program downloads all the results for the days it last downloaded them to keep upto date. If a forecast has been made it will combine it with the results to keep a record.
To make a forecast the program downloads all the days cards automatically then does its number crunching. It will also check for non-runners and delete them which is very handy as having one non runner in a patent can make a hell of a difference.
Although the odds are shown they are only the mornings expected start prices and as they aren’t used in the calcs I just keep them for record. When the results are downloaded the actual SPs are recorded so I can see which courses are giving I then have to look online to check the SPs before placing a bet.
[attachment=1:2fd7z0y8]HAYDOCK_16-07-2011.jpg[/attachment:2fd7z0y8][attachment=0:2fd7z0y8]RIPON_16-07-2011.jpg[/attachment:2fd7z0y8]
And I also need to lern to count – good job I can get the computer to add up.
- You must be logged in to reply to this topic.