Using Outliers

[Robert Fischer]

Traynor posted a wonderful primer on Eliminating Outliers.

I thought it might be interesting to do a brainstorm thread/topic on using Outliers.

By "Outliers" I want to start with focus on the "windfall", "jackpot", small-bet/BIG-return wagers.

These are not mutually exclusive from eliminating outliers from your database(you can eliminate outlier statistics, while shooting for the moon on some wagers at the same time). *Later we can focus more on database stats if the wager side discussion stagnates.

Thoughts? Strategies? On Using these windfall Outliers ??

[Robert Fischer]

I want to post an excerpt I was reading from "Taleb-The Black Swan".

He mentions a strategy.
His "Barbell Strategy" in essence attempts to eliminate middle-range risk, while splitting the focus between Conservative Wagers(90%) and Windfall Outliers(10%).

I am not sure I agree with it, but it is a very interesting starter to this topic.

Anyway, here is an excerpt from the book:


Barbell Strategy
I am trying here to generalize to real life the notion of the "barbell" strategy
I used as a trader, which is as follows. If you know that you are vulnerable
to prediction errors, and if you accept that most "risk measures"
are flawed, because of the Black Swan, then your strategy is to be as hyperconservative
and hyperaggressive as you can be instead of being mildly
aggressive or conservative. Instead of putting your money in "medium
risk" investments (how do you know it is medium risk? by listening to
tenure-seeking "experts"?), you need to put a portion, say 85 to 90 percent,
in extremely safe instruments, like Treasury bills—as safe a class of
instruments as you can manage to find on this planet. The remaining 10 to
15 percent you put in extremely speculative bets, as leveraged as possible
(like options), preferably venture capital-style portfolios.* That way you
do not depend on errors of risk management; no Black Swan can hurt you
at all, beyond your "floor," the nest egg that you have in maximally safe
investments. Or, equivalently, you can have a speculative portfolio and insure
it (if possible) against losses of more than, say, 15 percent. You are
"clipping" your incomputable risk , the one that is harmful to you. Instead of having medium risk, you have high risk on one side and no risk on the other. The average will be medium risk but constitutes a positive exposure
to the Black Swan. More technically, this can be called a "convex" combination.
Let us see how this can be implemented in all aspects of life."

* Make sure that you have plenty of these small bets; avoid being blinded by the
vividness of one single Black Swan. Have as many of these small bets as you can
conceivably have. Even venture capital firms fall for the narrative fallacy with a
few stories that "make sense" to them; they do not have as many bets as they
should. If venture capital firms are profitable, it is not because of the stories they
have in their heads, but because they are exposed to unplanned rare events.

[Magister Ludi]

I want to post an excerpt I was reading from "Taleb-The Black Swan".

He mentions a strategy.
His "Barbell Strategy" in essence attempts to eliminate middle-range risk, while splitting the focus between Conservative Wagers(90%) and Windfall Outliers(10%).

I am not sure I agree with it, but it is a very interesting starter to this topic.

Anyway, here is an excerpt from the book:

Taleb calls the world where Gaussian statistics prevail "Mediocristan". The world that uses a distribution with fatter tails than the lognormal he calls "Extremistan". I believe that the North American racetrack betting market is in the world of "Mediocristan". I don't think that we'll be experiencing any 25-sigma events at Mediocristan Racecourse. Neither is it necessary nor advisable to implement his so-called "Barbell Strategy".

[DeltaLover]

Taleb calls the world where Gaussian statistics prevail "Mediocristan". The world that uses a distribution with fatter tails than the lognormal he calls "Extremistan". I believe that the North American racetrack betting market is in the world of "Mediocristan". I don't think that we'll be experiencing any 25-sigma events at Mediocristan Racecourse. Neither is it necessary nor advisable to implement his so-called "Barbell Strategy".

completely agree :ThmbUp:

[Robert Fischer]

Looks like you are familiar with the book/concepts. :ThmbUp:

It seems like horse racing does have some relatively big windfall opportunities.

For example, the Rainbow Six. The right $2 outlay could have brought you 3.5Million dollars.

Taleb calls the world where Gaussian statistics prevail "Mediocristan". The world that uses a distribution with fatter tails than the lognormal he calls "Extremistan". I believe that the North American racetrack betting market is in the world of "Mediocristan". I don't think that we'll be experiencing any 25-sigma events at Mediocristan Racecourse. Neither is it necessary nor advisable to implement his so-called "Barbell Strategy".

[Magister Ludi]

Looks like you are familiar with the book/concepts. :ThmbUp:

It seems like horse racing does have some relatively big windfall opportunities.

For example, the Rainbow Six. The right $2 outlay could have brought you 3.5Million dollars.

I believe that Taleb’s argument is more nuanced than “watch out for fat tails”. I think that a case may even be made that it has less to do with small probabilities and more to do with false certainties. Many of us seem to have serious flaws in our logic by which we erroneously extrapolate common data ("all swans to date have been white") to believe a false conclusion ("no future swans will be black"). We seem to be surprised that our long chains of dominos fall over. We should learn not to set up our dominos in long chains.

[Robert Fischer]

I believe that Taleb’s argument is more nuanced than “watch out for fat tails”. I think that a case may even be made that it has less to do with small probabilities and more to do with false certainties. Many of us seem to have serious flaws in our logic by which we erroneously extrapolate common data ("all swans to date have been white") to believe a false conclusion ("no future swans will be black"). We seem to be surprised that our long chains of dominos fall over. We should learn not to set up our dominos in long chains.

Well said.

Even his case for the Barbell Strategy in the quoted excerpt seems to rely on the medium risk bets being "incomputable".

If your medium range bets are not any sort of issue for you as a horseplayer, what you are basically left with is whether or not allocating a portion of your bankroll to the randomness of "long shots" is a good idea.

[dnlgfnk]

In other words, know your percentages, like any good poker player can advise in simple street language.

I don't have the past performances, but in Saturday's 5th at Gulfstream, these were the odds:

# Odds %Win
1 10 7
2 20 4
3 7 10
4 Even 41
5 8 9
6 7 10
7 13 5
8 9/2 14

The field can be separated into "conservative"...#4 = 41%
"middle-risk"...#'s 3,6,8 = 34%
and "outlier"...#'s 1,2,5,7 = 25%

Finding a "false certainty" in our "conservative" (he wins less than half the time, and engaged in a three way duel for the lead, possibly anticipated?) leaves us with the body of the field which will win more often than not. And if we can shift one of the "middle-risk" propositions (#3 for instance, since we correctly anticipated #6 being involved in the destructive speed duel) to the "outlier" group, that group now holds a 35% to 24% edge, in addition to the increased chances assimilated from the underlaid favorite. And that group should be collectively scrutinized for any positive handicapping attributes, likely having a strong overlay or two within it.

With all due respect, the reach into other fields to better describe what the parimutuel investor should be doing is merely dressed up language to describe the process most of us here are already aware of...the search for overlays.