Interesting Article

[traynor]

"We played as a team so that we could pool our funds to place higher bets, and so the natural losses that players suffer along with wins - what gamblers call fluctuation - was evened out into a steady, marginal gain.

If people really knew about fluctuation before they decided to become players they would give up on the idea. It's possible to play correctly and lose for a grotesque amount of time. It might sound strange, but part of what makes a professional gambler is an ability to lose and lose and lose without going bonkers."

http://www.bbc.com/news/magazine-38103674

While this might play into the beliefs of some, it is well to realize she is referring to blackjack players--NOT horse race bettors. There is not much "fluctuation"--a random distribution of events within a tightly constrained set--in horse race betting that can (even with wishin' and hopin' on steroids) "reward" the stoic loser with an (approximately even) distributions of wins.

Horse racing is not blackjack. There is no "statistical pressure" that guarantees that long strings of losses may be somehow balanced out by wins if one can only keep slogging on long enough. There may be long strings of losses in blackjack because of normal distribution of events within a fixed set of possible outcomes. There is no equivalent in horse race betting. Those long strings of losses in horse race betting have one primary cause--the bettor bet on the wrong horse(s). Accepting such as "normal" does not seem a particularly wise long-term strategy.

[ReplayRandall]

Originally Posted by traynor

Horse racing is not blackjack. There is no "statistical pressure" that guarantees that long strings of losses may be somehow balanced out by wins if one can only keep slogging on long enough. There may be long strings of losses in blackjack because of normal distribution of events within a fixed set of possible outcomes. There is no equivalent in horse race betting. Those long strings of losses in horse race betting have one primary cause--the bettor bet on the wrong horse(s). Accepting such as "normal" does not seem a particularly wise long-term strategy.

The link was a good read. However, I disagree with your above paragraph, and here's why. A long-term winning horse-player goes through similar variances/fluxes on a regular basis. Using exactly the same methodologies continuously will have, for example, good runs of up to 3 weeks, followed by horrible runs for 3 weeks. This is NORMAL when betting the T-breds, as it's almost like someone turning on a light switch, and coming back later, at an undetermined time, to switch it OFF. There is absolutely no rhyme or reason why this happens on such a regular basis, but it's all part of the game in the long run. BTW, pro's in poker go through exactly the same runs, some much more severe than pro horse-players.....

[CincyHorseplayer]

Good read T.

I can go a pretty lengthy bit of time before I hit the wall. Then stop for a week and I'm fresh again. The break while no incoming dollars is worth it by lack of losses. Then it's as if I never stepped away. I bet more races these days precisely because I know the variance evens out after about 100 bets or more. The more cycles the better. Passing on low value, knowing when to stop briefly, only as a means to play at the highest level one is capable of for as long as possible is all I've worked on in addition to other non handicapping research the last few years. It's been the most interesting and fun I've had with this game in years. I embrace battling the variance!

[traynor]

Originally Posted by ReplayRandall

The link was a good read. However, I disagree with your above paragraph, and here's why. A long-term winning horse-player goes through similar variances/fluxes on a regular basis. Using exactly the same methodologies continuously will have, for example, good runs of up to 3 weeks, followed by horrible runs for 3 weeks. This is NORMAL when betting the T-breds, as it's almost like someone turning on a light switch, and coming back later, at an undetermined time, to switch it OFF. There is absolutely no rhyme or reason why this happens on such a regular basis, but it's all part of the game in the long run. BTW, pro's in poker go through exactly the same runs, some much more severe than pro horse-players.....

Most of the bettors I know use computer models. An intrinsic component of those models is (predicted) distribution. In essence, some (types of) wagers consistently produce scattered results--long strings of losses, long strings of choppy results, and (not nearly often enough) relatively tight clusters of wins. I think bettors should understand when such are "normal" and when they are the result of inconsistent or inadequate selection tactics. Other types of wagers result in fairly consistent results (with very few long stretches up or down). I much prefer the latter.

[traynor]

Originally Posted by CincyHorseplayer

Good read T.

I can go a pretty lengthy bit of time before I hit the wall. Then stop for a week and I'm fresh again. The break while no incoming dollars is worth it by lack of losses. Then it's as if I never stepped away. I bet more races these days precisely because I know the variance evens out after about 100 bets or more. The more cycles the better. Passing on low value, knowing when to stop briefly, only as a means to play at the highest level one is capable of for as long as possible is all I've worked on in addition to other non handicapping research the last few years. It's been the most interesting and fun I've had with this game in years. I embrace battling the variance!

Again, I think it is important for a bettor to understand when the results are "normal" (whatever that might be for the type(s) of wager(s) made) and when they are not. Every model I use has a built-in analysis of frequency distribution--exactly how many (and how severe) ups and downs existed in the base population (and in subsequent sample populations used for verification) used to create the model. I have no problem whatsoever taking breaks from betting if the results are not what I expect them to be. If the model(s) are any good, they perform the same whether I bet on them or not. I would rather miss a few juicy payoffs than chase rainbows. There will always be more of those juicy payoffs. Chasing rainbows does not seem to be a good strategy (at least for me--YMMV).

[thaskalos]

Didn't I see a recent run of harness plays of yours on this site...where you ran up losses in excess of $400 by wagering a mere $4 per bet? Were you using a "model" for those plays...and, did this run qualify to be called a "long string of losses"?

I apologize in advance if I misunderstood your harness wagers.

[CincyHorseplayer]

Originally Posted by traynor

Most of the bettors I know use computer models. An intrinsic component of those models is (predicted) distribution. In essence, some (types of) wagers consistently produce scattered results--long strings of losses, long strings of choppy results, and (not nearly often enough) relatively tight clusters of wins. I think bettors should understand when such are "normal" and when they are the result of inconsistent or inadequate selection tactics. Other types of wagers result in fairly consistent results (with very few long stretches up or down). I much prefer the latter.

I'm going to have to pick this up later today. This conversation hits on a lot of the philosophical experiences I have had as a player. The relationship between predictability/profitability. Not just tracks or surfaces but the worth of any model. I am new to this but have discovered very quickly it's value in my research. And as you said also the near lack of variance in certain as opposed to other bets. Starting out with Beyer's Winning Horseplayer I learned the game exotics first. But over time realized I was squandering a great deal of my abilities by neglecting to bet to win. And eventually realized using a 3 unit method of win and win-place that it could be self contained and profitable to a degree that would surprise some. Then later adding the model concept mainly to biased tracks but especially turf courses and their stark contrast to dirt and each other that dutch betting was highly profitable and could even rival exotics betting in profit potential. The entirety of the last 18 months I have spent basically trying to invent better techniques in contender identification and race recognizability for exotics potential while cutting down on pure methodical exotics play. Being more efficient in race selection for those bets. This is these days an inherently backwards concept. But I see baseball parallels. With baseball now in it's post steroid era(for the most part) there exist hitters who hover around the Mendoza line but can bang HR's without having many other skills that translate into runs scored(profit potential) like OBP and RBI ability of which the best skill is to be able to hit the ball for average(hit % in horses). Many players share similar stories of swinging for the fences while the ROI never attains great heights. It seems that a more well rounded "Betting" game would deal with the profitability issue on multiple fronts just like a player's overall batting line makes him a more intrinsically valuable player within the concept of a team. AVG(win of 5% bank or better)/OBP(cash rate)/SLG(ROI)/OPS(Horseplayer Sabermetric of cash rate+ ROI). Each betting approach would have it's inherent strengths and weaknesses that it addresses in the equation. All the concepts are fascinating. Much of this bet psychology also. Anyway good stuff all. Got my wheels turning and have to be up in 2 hours! Thanks!

[barn32]

Originally Posted by traynor

If people really knew about fluctuation before they decided to become players they would give up on the idea. It's possible to play correctly and lose for a grotesque amount of time. It might sound strange, but part of what makes a professional gambler is an ability to lose and lose and lose without going bonkers.

http://www.bbc.com/news/magazine-38103674

Thanks for posting that link. It brings back memories.

I can't remember every playing with Cat, but I probably did. I played with David Heyden quite a lot and had many conversations with him. Very impressive.

David was a lifelong stud player until no-limit took over and he switched games.

In fact, he was one of the regular players in Larry Flynt's 1000-2000 stud game (staked I'm sure) that went on for quite some time.

The best female players I've ever played with were Betty Carey, Barbara Samuelson (nee Gold), and Cissy Bottoms.

Betty Carey's story is fascinating and mysterious at the same time. She was an excellent player, especially at no-limit. No one really knows what happened to her or why, but rumors abound. She was good friends with Mike Sexton, and he probably knows, but one story was that her life was threatened by none other than Bob Stupak so she disappeared.

Barbara Gold played in all the high stakes stud, razz and hold'em games and did quite well.

Cissy Bottoms, well she was just a medium limit machine.

I'm reminiscing now. I was playing in a $1000 satellite for the main event. Betty Carey, Stu Ungar and Ken (what a player) Smith were in the game.

Stuey kept going on and on about how he was a two to one favorite to win a hand that (he tried to steal), and Betty kept telling him that he would get it back. Just hang in there, you'll get it back.

This was 1984, and Stuey went on to win the main event that year in Lake Tahoe.

I didn't see Stuey for a half a dozen years after that, and I barely recognized him when I did. His face was a mess. His nose was practically caved in from the cocaine. What a waste.

He defeated Ralph Morton in that 1984 heads up showdown, who was subsequently beaten within an inch of his life over a drug deal (robbery) in a cheap motel about a mile from the Commerce Club, and was lucky to survive.

I ran into Ralph a half a dozen years after that and we got to talking. He said they beat him so badly (they broke off a sink and were crushing his head with it), that he lost 75% of his mental facilities. It took him quite a long time to get back on his feet. His memory was shot.

I reminded him about that last hand he played against Stuey in the 1984 main event (QQvs99), and he said, "I don't remember much about my past but I do remember that hand! You've got a pretty good memory."

[traynor]

Originally Posted by thaskalos

Didn't I see a recent run of harness plays of yours on this site...where you ran up losses in excess of $400 by wagering a mere $4 per bet? Were you using a "model" for those plays...and, did this run qualify to be called a "long string of losses"?

I apologize in advance if I misunderstood your harness wagers.

A good example, and the (paper) losses were up to $700 or so before I stopped posting (and betting) that particular model. After more than a year and a half of fairly consistent positive returns, the Yonkers reverse exacta model (using specific criteria for selections) went belly up big time. I continue(d) monitoring the results (a routine process) and there has been no "reversal of fortune positive swing to compensate for the (thankfully shortened) string of losses" in that case. The year and a half of profitable selections was not representative of a "normal distribution" that could be depended on to continue.

My only regret (other than pining for what I (unfortunately) boasted of as the "easy profit" available using reverse exactas at Yonkers) is that I didn't catch the downturn/cessation of profitability earlier.

[traynor]

Originally Posted by CincyHorseplayer

I'm going to have to pick this up later today. This conversation hits on a lot of the philosophical experiences I have had as a player. The relationship between predictability/profitability. Not just tracks or surfaces but the worth of any model. I am new to this but have discovered very quickly it's value in my research. And as you said also the near lack of variance in certain as opposed to other bets. Starting out with Beyer's Winning Horseplayer I learned the game exotics first. But over time realized I was squandering a great deal of my abilities by neglecting to bet to win. And eventually realized using a 3 unit method of win and win-place that it could be self contained and profitable to a degree that would surprise some. Then later adding the model concept mainly to biased tracks but especially turf courses and their stark contrast to dirt and each other that dutch betting was highly profitable and could even rival exotics betting in profit potential. The entirety of the last 18 months I have spent basically trying to invent better techniques in contender identification and race recognizability for exotics potential while cutting down on pure methodical exotics play. Being more efficient in race selection for those bets. This is these days an inherently backwards concept. But I see baseball parallels. With baseball now in it's post steroid era(for the most part) there exist hitters who hover around the Mendoza line but can bang HR's without having many other skills that translate into runs scored(profit potential) like OBP and RBI ability of which the best skill is to be able to hit the ball for average(hit % in horses). Many players share similar stories of swinging for the fences while the ROI never attains great heights. It seems that a more well rounded "Betting" game would deal with the profitability issue on multiple fronts just like a player's overall batting line makes him a more intrinsically valuable player within the concept of a team. AVG(win of 5% bank or better)/OBP(cash rate)/SLG(ROI)/OPS(Horseplayer Sabermetric of cash rate+ ROI). Each betting approach would have it's inherent strengths and weaknesses that it addresses in the equation. All the concepts are fascinating. Much of this bet psychology also. Anyway good stuff all. Got my wheels turning and have to be up in 2 hours! Thanks!

A sound approach. Find stuff that works. Use it. If it stops working, dump it. Rinse and repeat. There is no Holy Grail. There is only stuff that works for awhile. Whatever you discover in the process, it is relatively safe to assume that a whole pack of very aggressive, very serious bettors have discovered (or will shortly discover) the same thing. Take advantage of it while the opportunity exists.

The same situation exists in most other areas of human endeavor, especially business.

[traynor]

Originally Posted by barn32

Thanks for posting that link. It brings back memories.

I can't remember every playing with Cat, but I probably did. I played with David Heyden quite a lot and had many conversations with him. Very impressive.

David was a lifelong stud player until no-limit took over and he switched games.

In fact, he was one of the regular players in Larry Flynt's 1000-2000 stud game (staked I'm sure) that went on for quite some time.

The best female players I've ever played with were Betty Carey, Barbara Samuelson (nee Gold), and Cissy Bottoms.

Betty Carey's story is fascinating and mysterious at the same time. She was an excellent player, especially at no-limit. No one really knows what happened to her or why, but rumors abound. She was good friends with Mike Sexton, and he probably knows, but one story was that her life was threatened by none other than Bob Stupak so she disappeared.

Barbara Gold played in all the high stakes stud, razz and hold'em games and did quite well.

Cissy Bottoms, well she was just a medium limit machine.

I'm reminiscing now. I was playing in a $1000 satellite for the main event. Betty Carey, Stu Ungar and Ken (what a player) Smith were in the game.

Stuey kept going on and on about how he was a two to one favorite to win a hand that (he tried to steal), and Betty kept telling him that he would get it back. Just hang in there, you'll get it back.

This was 1984, and Stuey went on to win the main event that year in Lake Tahoe.

I didn't see Stuey for a half a dozen years after that, and I barely recognized him when I did. His face was a mess. His nose was practically caved in from the cocaine. What a waste.

He defeated Ralph Morton in that 1984 heads up showdown, who was subsequently beaten within an inch of his life over a drug deal (robbery) in a cheap motel about a mile from the Commerce Club, and was lucky to survive.

I ran into Ralph a half a dozen years after that and we got to talking. He said they beat him so badly (they broke off a sink and were crushing his head with it), that he lost 75% of his mental facilities. It took him quite a long time to get back on his feet. His memory was shot.

I reminded him about that last hand he played against Stuey in the 1984 main event (QQvs99), and he said, "I don't remember much about my past but I do remember that hand! You've got a pretty good memory."

Way back, I saw two books in a bookstore window, and bought both. I chose the direction of the first--Playing Blackjack as a Business--rather than the second (with a more engaging title)--Play Poker, Quit Work, and Sleep 'til Noon.

Your post almost makes me wish I had chosen the direction of the second book.

[traynor]

Originally Posted by thaskalos

Didn't I see a recent run of harness plays of yours on this site...where you ran up losses in excess of $400 by wagering a mere $4 per bet? Were you using a "model" for those plays...and, did this run qualify to be called a "long string of losses"?

I apologize in advance if I misunderstood your harness wagers.

As I have mentioned before, I often increase wagers after losses. Using such a process requires explicit knowledge of the frequency distribution of specific models--in short, what is "normal" and what is "not normal." Not an easy task, when using literally dozens of different models for a variety of wagers. Fortunately, most of the process is automated.

[thaskalos]

Originally Posted by traynor

Most of the bettors I know use computer models. An intrinsic component of those models is (predicted) distribution. In essence, some (types of) wagers consistently produce scattered results--long strings of losses, long strings of choppy results, and (not nearly often enough) relatively tight clusters of wins. I think bettors should understand when such are "normal" and when they are the result of inconsistent or inadequate selection tactics. Other types of wagers result in fairly consistent results (with very few long stretches up or down). I much prefer the latter.

In parimutuel games like horse racing and poker...the "safer" a playing style is, the less profitable it becomes long-term. The players who embrace these safer playing styles are typically the ones whose temperament and personality can't cope with the high degree of variance associated with the more aggressive styles of play.

In no-limit holdem...the competent "loose-aggressive" players are the most dangerous and the most profitable...but they also suffer the most turbulent financial swings. The "nits", who always wait for a big hand before committing to a pot, aren't nearly as dangerous or long-term profitable...but they can scrape up a profit in the smaller games without the stress of the prolonged drawdowns that the bolder, more profitable players experience. For most players...a compromise between these two extreme betting styles would probably be the best thing.

In horse racing...the competent bettor who restricts his wagers to presumed overlays, and keys these overlays in win-bets and exotics, will have a chance at the highest profit levels that the game offers...but he will also have to pay the significant price of experiencing the biggest fluctuations to his bankroll. That's why your 2-horse exacta-box wagers resulted in the significant drawdown that I referenced earlier...even thought the method itself could very well have been highly profitable overall.

The horseplayer who opts to "spread out" more in his wagering, does so in an attempt to "smooth-out" the high degree of variance experienced by the more aggressive bettor...but by doing so, he sacrifices some long-term profitability. There is no "free lunch"...in life OR in gambling.

To summarize...the degree of bankroll fluctuations that horseplayers experience have less to do with the competency of their methods...and more to do with the aggressiveness of their betting styles. It all comes down to how much "gamble" the bettor has in him...and how much "risk" his temperament could effectively tolerate. As a good friend of mine likes to say..."you have to wear the clothes that suit you".

By the way...I think this thread is a much better fit in the horse racing section of the site.

[ReplayRandall]

Originally Posted by traynor

As I have mentioned before, I often increase wagers after losses.

After a win, what do you do with wager size?

[traynor]

Originally Posted by CincyHorseplayer

... that dutch betting was highly profitable and could even rival exotics betting in profit potential.

An interesting area. It might be that the opportunities exist because the races that favor dutching may be more competitive, with outcomes not as clear cut as (many) other races. I have been doing a lot of study in the last few months that indicates more (selections) may be better. Specifically, races in which the selection process I use for dutching indicates three potential winners are more profitable (betting all three) than races it indicates (only) two potential winners (and betting both). In some instances, dutching four entries is productive.

That is not "every race, every track, every day." That is only the very specific races that fit the pre-determined profile of that particular model. However, the obvious benefit of being able to spread wagers over several entries (with less effect on the odds of the individual entries) should not be underestimated.

Still in the "research" phase. One thing I need to study is the effect of using conditional wagers when dutching multiple entries (for example, restricting minimum odds to 2/1 for a three entry dutch, 3/1 for a four entry, or whatever). Easy to say, hard to study, given the last minute odds fluctuations.