Data on clustering of longshots?

[lansdale]

I'm curious about the extent to which there's any statistically significant clustering of longshots. I say this, having experienced this recently (check out the late HOL card from yesterday for an example) a couple of times, while still harboring bitter memories of the long stretch of chalk at SAR about which Dave Schwartz posted. All part of my ongoing quest to play better and fewer races. Thanks for any help.

Cheers,

lansdale

[dav4463]

I could improve my betting a thousand percent if I had a clue as to when the favorites are going to dominate.

Normally I play longshots, but many times the races I expect the favorite to win are the ones where I hit with a big longshot.

When I think a race is wide-open, it seems like the inevitable 4.40 bet down winner romps home.

[RichieP]

Speaking specifically to this thread TITLE:

It has virtually nothing to do with the actual race you are looking at today (including volatility, chaos or perceived vulnerability of the favorite).

Yup you heard that right. It flies in the face of "reality" yet it is very real.

I have been working on this for months and 10 days ago began working live races with a user of Ted Craven's RDSS package. I explained to her the concept, told her what to look for and then kept quiet.

To say she now "sees" what will happen PRICE WISE before things go down(unquestionable majority of the time btw) is quickly becoming an understatement.

Research/small bets on this concept is ongoing and please remember I am speaking specifically to this threads title ONLY.

Merry Christmas and Happy Holidays
http://www.youtube.com/watch?v=USN2mKMT9Zs&feature=rec-HM-fresh+div

[the_fat_man]

I play a lot of longshots. It's puzzling to me that one would post about the clustering of longshots specifically. Why would only longshots cluster or why would only the clustering of longshots be of importance. Longshots are just like any other bet and, for me, their tendencies are similar to those of the others:

1) I can't pick my nose
2) I pick the right horse and it loses (bad ride, poor setup, etc.)
3) The horse wins

2 and 3 far exceed 1 in terms of frequency.

The above, of course, depends on how strong my opinion is. The stronger my opinion, the higher the likelyhood of 2 and 3. And, it's no secret that the strength of my opinion depends on what I see in a given race (based on the way I play).

I learned the hard way that expecting 'clustering' or any such thing is the way to the poor house. Each play needs to be made INDIVIDUALLY, with no residual baggage or expectations.

[riskman]

I learned the hard way that expecting 'clustering' or any such thing is the way to the poor house. Each play needs to be made INDIVIDUALLY, with no residual baggage or expectations.

You have that right. But I am sure someone will come back and start talking about biases which in itself is usually short lived.

[markgoldie]

You have that right. But I am sure someone will come back and start talking about biases which in itself is usually short lived.

Add many horses first over the track esp. when the track change is from dirt to poly. On Sat.'s opening card at Turfway, the average $2 super payout on 8 races was over $18,000 and the one pent was unhit.

[ranchwest]

Add many horses first over the track esp. when the track change is from dirt to poly. On Sat.'s opening card at Turfway, the average $2 super payout on 8 races was over $18,000 and the one pent was unhit.

Most people can't pick Turfway from the Charts.

[lansdale]

Speaking specifically to this thread TITLE:

It has virtually nothing to do with the actual race you are looking at today (including volatility, chaos or perceived vulnerability of the favorite).

Yup you heard that right. It flies in the face of "reality" yet it is very real.

I have been working on this for months and 10 days ago began working live races with a user of Ted Craven's RDSS package. I explained to her the concept, told her what to look for and then kept quiet.

To say she now "sees" what will happen PRICE WISE before things go down(unquestionable majority of the time btw) is quickly becoming an understatement.

Research/small bets on this concept is ongoing and please remember I am speaking specifically to this threads title ONLY.

Merry Christmas and Happy Holidays
http://www.youtube.com/watch?v=USN2mKMT9Zs&feature=rec-HM-fresh+div

Hi Richie,

Best of luck with your research on this valuable area. I probably wasn't clear enough, but what I was asking about was less individual races, than a general overview. Let me explain. A number of tracks I don't even look at, since I know the ratio of chalk is usually high - I think you yourself cited Suffolk as one example, for me it's the Maryland circuit. At others, I know that the opportunities will be there - the first six weeks of GP are usually good, Keeneland, SAR (usually), and, to a degree, Churchill. Many tracks fall somewhere in the middle. But it has often seemed to me that longshots run in cycles at many tracks, and I'd like to know how much validity there is to this notion, and if it does hold water, what the correlated factors might be, since handicapping time is finite.

Much thanks for your input, and for the powerful tenor of Di Stefano. You probably know better than I that he was the key influence on the singing of Pavarotti.

Cheers,

lansdale

[valueguy]

Longshots as far i am concerned are not predicable .They are a random happening.(Thank Goodness) I had an epiphany when i started making an odds
line ala the 80% contender method by Barry Medows .Even when i made my
favorte 8/5 and the tote agreed with me .I figured it would be a solid favorite .If i made the mistake of passing the race instead of going with my 6/1 overlay whom the tote is making 12/1 then sure as hell the longshot
would come in.The odds line results changed my way of thinking.I went from
trying to predict the winner to an overlay based betting method .It didn,t make
me an over night millionaire but at least i now make enough to buy a premium
beer at the racebook.The secret i found ,was you don,t try to avoid a loss, but you must commit yourself to making money.You will be wrong 75% time but the ROI on the overlays will cover your mistakes.

[Backstretch Pirate]

No doubt, longshots do come in clusters. Track conditions, and maintenance are the main reasons for this.

[therussmeister]

No doubt, longshots do come in clusters. Track conditions, and maintenance are the main reasons for this.

Randomness also comes in clusters.

[cmoore]

This is for any of you database guys or gals..Is there a way for you to look at maiden win prices early in the year compared to later in the year. I'm thinking first timers here. It seems that the early stock of 2 year olds pay less to win compared to 2 year olds that had to wait an extra 4-8 months before they make thier first start. Just an idea I thought of. I could be all wet.

[Backstretch Pirate]

Randomness also comes in clusters.
Randomness? Why play the track if you believe this?

[RaceBookJoe]

I am interested in your results. Lou Holloway wrote about this a long time ago ( even though he used roulette as an example ), he said that longshots came in clusters, he couldnt say why they did, just that his extensive research showed him thatit happened. Read his book "Full Time Gambler"...he talks a little about it in there. rbj

[46zilzal]

Randomness? Why play the track if you believe this?
Unless one accepts the reality that randomness plays a part in the outcome of many races, one is simply fooling themselves

[Backstretch Pirate]

Unless one accepts the reality that randomness plays a part in the outcome of many races, one is simply fooling themselves

Horse racing isn't random, not even pseudo-random.
Just because your mind can't phathom what has taken place in a horse race doesn't make it a random event.

[citygoat]

maybe early in the meet longshots huddle but as the meet goes on prices appear to smooth out.I go with the assumption that every horse has a chance to win before the gate opens.Usually at least one horse does win every race.

[garyoz]

Randomness? Why play the track if you believe this?

Don't confuse randomness with a normal distribution. Yes, red on the roulette wheel does come up mulitple times in a row--and that is probably within one or two standard deviations of expectations. It is dangerous to assume causation between what are independent events. If #4 wins a race, does that effect the likelihood of #4 winning the next race?

Sometimes longshots do occur for a reason--bias, etc., but bettors usually adjust to this--the market is pretty efficient--and it corrects. Or, some tracks do not lend themselves to orthodox handicapping strategies--particularly the artificial surfaces--in which case perceived randomness (e.g., more longshots) may seem apparent.

IMHO, each race is an array of statistical probabilities associated with each horse winning--and given the vig, in order to be successful, you have to find the "inefficiencies" between the odds board and the "true" probablility of winning (which you are smart enough to ascertain (yeah, sure)), and then make the right bet to leverage that "inefficiency" (e.g., exacta, tris, pick-4, win)

[Backstretch Pirate]

Don't confuse randomness with a normal distribution.

I agree with you. I think your point was meant for 46 Zizal.

[46zilzal]

There is a lot more to randomness than is usually perceived.
http://www.amazon.ca/Drunkards-Walk-Randomness-Rules-Lives/dp/0375424040

[Vinnie]

Don't confuse randomness with a normal distribution. Yes, red on the roulette wheel does come up mulitple times in a row--and that is probably within one or two standard deviations of expectations. It is dangerous to assume causation between what are independent events. If #4 wins a race, does that effect the likelihood of #4 winning the next race?

Sometimes longshots do occur for a reason--bias, etc., but bettors usually adjust to this--the market is pretty efficient--and it corrects. Or, some tracks do not lend themselves to orthodox handicapping strategies--particularly the artificial surfaces--in which case perceived randomness (e.g., more longshots) may seem apparent.

IMHO, each race is an array of statistical probabilities associated with each horse winning--and given the vig, in order to be successful, you have to find the "inefficiencies" between the odds board and the "true" probablility of winning (which you are smart enough to ascertain (yeah, sure)), and then make the right bet to leverage that "inefficiency" (e.g., exacta, tris, pick-4, win)

Love your Post garyoz:

How often do you watch a race and at its conclusion hopefully it worked out for you as you perceived it would and you bet it accordingly? :) If you listen to an "expert" you may hear them say something along the lines of "That race was run very oddlyl" and the end result was pretty unusual because of this weird way the race went. You are saying to yourself, "what the heck is that guy talking about"? The race went just the way that you thought that it would in your minds eye and through the application of your own handicapping process or methods that you employ. Can any number of a million things occur in a race that aren't favorable to your preferred outcome? Of course! :) I have always heard that in any excellent software the winner can usually be found in say the TOP 5 oddsline approximely 88% of the time. I am not sure if this is entirely true or not. I believe that figure could in all likelihood be pretty accurate. The other 12% may be of the head scratcher variety or the randomness factor that is spoken of in this thread. Once again, superb post.