[davidchong]

Horses that meet my requierement to bet at 2-1 or better for today:

Aqueduct, Race 6 No.1, Race 7 No. 5.
Gulfstream, Race 2 No. 1, Race 3 No. 6 (Best Value), Race 5 No. 8, Race 7 No. 6 (john velasquez prefers this horse over the 11),
Hawthorne Race 3 No.5, Race 7 No. 2, Race 8 No.2, Race 9 No. 6

There ten bets. lets see waht happend with $2 mythical win bet.

[sharpcat]

then you dont need odds, you need efficiency on your picks.... look into your %..
if youre good you always have a profit.

Your missing the point here I am looking from a mathematical perspective where I feel that anytime I am getting say 10% more value on horse than I expect that whether that is the horse I like in the race or not it is worth betting on because of the large amount of +EV that the horse has. Long term thousands of +EV bets is likely to be profitable.

I have no problem picking horses i like to win and will continue to do so, but I am looking for under valued horses who say are 7-1 when I feel they should be 5-1 although they may not be the horse I expect to win they are stil valuable horses to bet (possibly using a ducting system if you have several under valued horses in a race)

[brettd]

No your predictions should be spot on.... your probability distributions when you frame markets should reflect reality. I.e: a 3/1 rated horse wins a long term average of 25% of the time. If upon analysis, if your method of handicapping does not create a probability distribution that does not reflect that your 3/1 rated horses win close to 25% of the time, then your handiapping method has serious problems.

[brettd]

 
I have no problem picking horses i like to win and will continue to do so, but I am looking for under valued horses who say are 7-1 when I feel they should be 5-1 although they may not be the horse I expect to win they are stil valuable horses to bet (possibly using a ducting system if you have several under valued horses in a race)

+EV is +EV wherever you get it. In terms of theoretically optimal bankroll growth, it is better to back a +EV favourite at $1.40 than a $10.00 roughie.

[sharpcat]

Mate there's no other way to go but framing your own book at 100% and finding +ev bets from there. If you can quantitatively define your handicapping variables into values (and are a +ev handicapper), then going through the framing process is the easy part.

I'm gathering you're having problems because you handicap on a more situational basis, and so thus have problems assigning values to your handicapping process?

Exactly I have no problems picking winners but I have no way of assigning value to them.

If you think the 2 horse is going to win the race how can you conclude that he has a _ _% of winning without having any mathematical say so. You are basically just betting in the blind because all though you think the horse will win you are likely getting negative value on the horse. for a horse to move from 3-1 to 5-1 is only a 5% difference with nothing but guesses you could easily be off on your estimation by 5% not to mention that no matter how good you are if you are just guessing what your odds should be it will be very hard long term to hold any consistency.

[sharpcat]

 
I have no problem picking horses i like to win and will continue to do so, but I am looking for under valued horses who say are 7-1 when I feel they should be 5-1 although they may not be the horse I expect to win they are stil valuable horses to bet (possibly using a ducting system if you have several under valued horses in a race)

+EV is +EV wherever you get it. In terms of theoretically optimal bankroll growth, it is better to back a +EV favourite at $1.40 than a $10.00 roughie.

yes but if the horse at $1.40 should really be at $2.00 and the horse at 10-1 should really be 8-1 than the 10-1 horse is mathematically the better bet.

[brettd]

Exactly I have no problems picking winners but I have no way of assigning value to them.

If you think the 2 horse is going to win the race how can you conclude that he has a _ _% of winning without having any mathematical say so. You are basically just betting in the blind because all though you think the horse will win you are likely getting negative value on the horse. for a horse to move from 3-1 to 5-1 is only a 5% difference with nothing but guesses you could easily be off on your estimation by 5% not to mention that no matter how good you are if you are just guessing what your odds should be it will be very hard long term to hold any consistency.

I totally agree with you, thus you need to determine probabilities for every single horse in the race, in order for you to get the probability of the horse that you like. If your 'picking winners' without accurately inferring true underlying value, then you need to seriously change the way you cap.

It's not about finding the 'winner'. It's about accurately inferring probabilities, and betting any overlays.

Price, and in turn value, is EVERYTHING to be +ev

[sharpcat]

whether you win or lose one individual race does not matter and it does not matter how many races you lose, if you are making positive EV bets consistently it does not matter if you are betting longshots or favorites you will profit.

If you bet 100 10-1 horses and you win 15 of those bets than you have turned a profit not by betting the best horse in the race but because you were clearly getting horses who were better than 10-1.

[sharpcat]

I totally agree with you, thus you need to determine probabilities for every single horse in the race, in order for you to get the probability of the horse that you like. If your 'picking winners' without accurately inferring true underlying value, then you need to seriously change the way you cap.

It's not about finding the 'winner'. It's about accurately inferring probabilities, and betting any overlays.

Price, and in turn value, is EVERYTHING to be +ev

I would like to check out this book you talk about and I appreciate your input Brett

[brettd]

yes but if the horse at $1.40 should really be at $2.00 and the horse at 10-1 should really be 8-1 than the 10-1 horse is mathematically the better bet.

I should have clarified, if both bets were +EV, and both bets had the same overlay, it is better to bet the $1.40 favourite (staking using the Kelly Criterion adjusted for mutually exclusive events).

[brettd]

I have programs that frame an 85% back market and a 115% lay market, and then basically it gets let loose on any given race on various betting exchanges. Any +EV overlays (whether it be backs or lays) has bets inputted according to the kelly criterion for mutually exclusive events. This way, I usually have numerous backs and lays for any given race, on any horse that I perceive has value.

So it's kind of like i'm taking a +EV 'position' on every race I engage in. I think that all mutally exclusive events should be approached in this way on betting exchanges. Find the value in every proposition in the betting event, and build your collective 'position' that is ultimately +EV.

[sharpcat]

I should have clarified, if both bets were +EV, and both bets had the same overlay, it is better to bet the $1.40 favourite (staking using the Kelly Criterion adjusted for mutually exclusive events).

Yeah I did catch that I was just saying that many times your favorite horses to win the race are mathematically not your best bet.

I may like the 2 an 3 horse in race 7 but even though I think they will win, if the 10 horse is 10-1 when he should be 7-1 he is likely the better horse to bet even though he is not the first horse that comes to my mind when I think about who will win the race. On any given day a horse in a 10 horse race really only has a 10% of winning, if he gets a bad start than no matter how great of a horse he is he has 9 other horses that can win.

[sharpcat]

I have programs that frame an 85% back market and a 115% lay market, and then basically it gets let loose on any given race on various betting exchanges. Any +EV overlays (whether it be backs or lays) has bets inputted according to the kelly criterion for mutually exclusive events. This way, I usually have numerous backs and lays for any given race, on any horse that I perceive has value.

So it's kind of like i'm taking a +EV 'position' on every race I engage in. I think that all mutally exclusive events should be approached in this way on betting exchanges. Find the value in every proposition in the betting event, and build your collective 'position' that is ultimately +EV.

100% agree

[Bsims]

Horses that meet my requierement to bet at 2-1 or better for today:

Aqueduct, Race 6 No.1, Race 7 No. 5.
Gulfstream, Race 2 No. 1, Race 3 No. 6 (Best Value), Race 5 No. 8, Race 7 No. 6 (john velasquez prefers this horse over the 11),
Hawthorne Race 3 No.5, Race 7 No. 2, Race 8 No.2, Race 9 No. 6

There ten bets. lets see waht happend with $2 mythical win bet.

Good luck. Just for the heck of it, I entered conditional wagers on these at YouBet. The conditions were at least 2-1 odds at 0 MTP.

By the way, at Gulfstream race 5 #8 and race 7 #6 were rejected because they were scratched.

[Bsims]

Sharpcat,

From you posts I gather you have an idea of assigning points for certain factors and would like to translate these into a morning type line looking for overlays.

I am working with a large database of past races and have the programming and math skills required to do what you'd like to do. I would be willing to work with you to develop a system if you are interested.

I should point out though that it can only be accomplished if the points you wish to assign are strictly numerically derivable from the data in past performances. If human judgment or interpretation is involved, then back testing is not possible.

If you are interested in pursuing this, PM me.

[sharpcat]

Sharpcat,

From you posts I gather you have an idea of assigning points for certain factors and would like to translate these into a morning type line looking for overlays.

I am working with a large database of past races and have the programming and math skills required to do what you'd like to do. I would be willing to work with you to develop a system if you are interested.

I should point out though that it can only be accomplished if the points you wish to assign are strictly numerically derivable from the data in past performances. If human judgment or interpretation is involved, then back testing is not possible.

If you are interested in pursuing this, PM me.

sounds good Sims, let me work up some ideas and I will PM you maybe we can put some things together.

The only problem I must warn you about is this would take some time to perfect and adjust, and who knows may never work for us but if we can come up with something and test over time to see if we are actually finding value it could be nice. Me personally I enjoy things like this and the challenge is the whole reason I wager on sports and racing.

The big thing is that the public overbets favorites therefore creating value in many races on some of the dogs, if we could develop a solid database to spit out consistent % for us that prove to be accurate we will not be winning every single one of our bets but will be able to profit with the odds we are getting.

[sharpcat]

Once again I appreciate the contributions and opinions from everybody on this topic

[jw]

The big thing is that the public overbets favorites therefore creating value in many races on some of the dogs

That's you're key right there.

Only reason I hate this method is that you can't put the mathematical value on the horse that is boxed in and comes with the late run one it gets a break, or on the one that stumbles out of the stalls and drops 15 lengths yet finishes just 2 lengths down. However, if you are also doing your regular handicapping and scanning the field for events like this alongside your mathematical points system. You should be good.

(maybe assign a 10 point "bad beat" value that can be manually entered somewhere)

Good luck guys.

[davidchong]

Good luck. Just for the heck of it, I entered conditional wagers on these at YouBet. The conditions were at least 2-1 odds at 0 MTP.

By the way, at Gulfstream race 5 #8 and race 7 #6 were rejected because they were scratched.

good luck buddy.... then only 8 bets today...

[Bsims]

good luck buddy.... then only 8 bets today...

All 8 wagers met the conditional requirements Of these, 4 were winners for a net of +$15.80 for the day. Thanks and good picking.

[davidchong]

good profit today, was no the best. 4 winners on 8 picks, invest $16 with return more than 30