Flat Wagering vs Progressions

[limit2]

Progressions have to vary according to the amount of average $2 mutual expected. However, flat bets have to do the job and progressions have to fare better. In the case of the longshot player, of which I am one, the progression should be long enough and be in moderate steps to absorb the occasional exhaustion (1 of 6) of 20 steps of 2 selections per race. I am currently experimenting to determine an optimum series of terms for a progression to fit my system. I tried a 15 and a 10 day series of races with the results on paper, and if discipline and consistency rule the day, I believe it can be translated into reality. I have to wonder what the experts of this forum believe? The respective ROI's for the 15 day progression, 15 day flat wagering, 10 day progression and 10 day flat wagering were 136%, 127.5%, 75.8% and 63.5%. The biggest losing streak was 17 races (34 tickets). I must have captured 2 pretty good streaks. But, my intuition tells me that a 33%-40% ROI is attainable in the long run.

[sammy the sage]

Progression...you mean upping the bet after a loser?

If so you'll be LIMITED by handle...as at smaller tracks of course your bet would DISTORT the win or any pools for that matter...

If you're talking about let it ride type bets...DON'T...

[thaskalos]

Progressions have to vary according to the amount of average $2 mutual expected. However, flat bets have to do the job and progressions have to fare better. In the case of the longshot player, of which I am one, the progression should be long enough and be in moderate steps to absorb the occasional exhaustion (1 of 6) of 20 steps of 2 selections per race. I am currently experimenting to determine an optimum series of terms for a progression to fit my system. I tried a 15 and a 10 day series of races with the results on paper, and if discipline and consistency rule the day, I believe it can be translated into reality. I have to wonder what the experts of this forum believe? The respective ROI's for the 15 day progression, 15 day flat wagering, 10 day progression and 10 day flat wagering were 136%, 127.5%, 75.8% and 63.5%. The biggest losing streak was 17 races (34 tickets). I must have captured 2 pretty good streaks. But, my intuition tells me that a 33%-40% ROI is attainable in the long run.
In that case...your biggest problem will be where to store all that money.

[Robert Goren]

If you had a dollar for every progression scheme that anybody has ever come up with, you would need your progression scheme.

[Some_One]

Over the long run, a progressive betting system (i.e. martingale, fibonacci) will not have any higher or lower ROI versus flat betting. All progressive betting systems do is move your individual session results to the extremes (either good win or good loss), versus a flat betting which should have a bell distribution looking probability graph. In the end, the average of both will be the same.

In the end, you're just on the happy side of variance for now, enjoy it, it won't last, unless you truly are flat bet profitable, which I highly doubt.

[raybo]

Correct! If you do not have a flat bet long term positive ROI, no wagering system will produce long term profit, you will just lose your money faster than with a flat bet.

[Dave Schwartz]

I will offer the short version of this:

There are only 3 basic wagering systems. All others are variations of these 3.

1. Increase after a win.
2. Increase after a loss.
3. Flat bet.


You CAN change your results with one of these 3 wagering strategies. However, you CANNOT turn a losing system into a winner or turn a winning system into a loser.


Using the flat bet as a benchmark, you will find that #1 moves your results closer to break even, and #2 moves it away from break even.

That is, if your sample is a winner, increasing after a win will cut down the advantage, while increasing after a loss will raise your advantage.

If the sample is a loser, increasing after a win will cause you to lose less, while increasing after a loss will cause you to lose more.


Betting Percentage of Bankroll is a category 1 proposition that will decrease your advantage if you are a winning player. This is driven by the win-a-bet/lose-a-bet sequence that causes the player to ROI.

If you use this approach as a losing player, you will lower your losses because as you lose and your BR gets smaller, your bets get smaller, too.


If you are a winning player, chasing the money helps "smooth out the bumpy patches." However, the price you pay for it is that while you win more sessions, the wins are smaller and when you lose the losses are huge.


If you are a losing player (i.e. your long-term expectation is negative), chasing the money will not SOLVE your problems. In fact, it will help guarantee that you lose.


Bottom Line: There is no free lunch.

[Nets]

I will offer the short version of this:



Bottom Line: There is no free lunch.

Damn! And I'm hungry

[DRIVEWAY]

My basic approach is to initially flat bet every horse that I like on a run of the mill racing day.

Let's say I've identified 40 flat bets to win. For illustration purposes there are 5 wagers at 8 different tracks.

At track A, I win the first race. I will take back 1/2 the payout and use the remaining 1/2 for additional (push) wagers at this track. I will divide the (push) amount by the (flat) wager amount. Then I will flat wager at Track A up to the next three races. If the (push) amount only covers one or two races than that is the wager. If there are only one or two races left at the track in question then that's what will be considered. If I'm hot and win multiple races in a row, I will continue to (push) into the upcoming races. Some races could have up to three (push) wagers or four times the original flat bet.

When you're hot your hot. You can be ice cold at seven out of eight tracks and still crush for the day by clustering wagers at the HOT track.

Your maximun wager amount exposure is identified at the beginning of the day(sum of the initial flat bets). Your push amount may limit your profit for the day when there is no follow up winner(s) at that track. However, when you're HOT at any given track, you get rewarded.

If you keep accurate records, go over your wager history and see how Cluster Wagering may have affected your results.

[lamboguy]

I will offer the short version of this:

There are only 3 basic wagering systems. All others are variations of these 3.

1. Increase after a win.
2. Increase after a loss.
3. Flat bet.


You CAN change your results with one of these 3 wagering strategies. However, you CANNOT turn a losing system into a winner or turn a winning system into a loser.


Using the flat bet as a benchmark, you will find that #1 moves your results closer to break even, and #2 moves it away from break even.

That is, if your sample is a winner, increasing after a win will cut down the advantage, while increasing after a loss will raise your advantage.

If the sample is a loser, increasing after a win will cause you to lose less, while increasing after a loss will cause you to lose more.


Betting Percentage of Bankroll is a category 1 proposition that will decrease your advantage if you are a winning player. This is driven by the win-a-bet/lose-a-bet sequence that causes the player to ROI.

If you use this approach as a losing player, you will lower your losses because as you lose and your BR gets smaller, your bets get smaller, too.


If you are a winning player, chasing the money helps "smooth out the bumpy patches." However, the price you pay for it is that while you win more sessions, the wins are smaller and when you lose the losses are huge.


If you are a losing player (i.e. your long-term expectation is negative), chasing the money will not SOLVE your problems. In fact, it will help guarantee that you lose.


Bottom Line: There is no free lunch.there was a free lunch Dave, that was when the race books in Nevada were booking their own wagers. Pari mutuel wagering in Nevada was the very worst thing that ever happened to me in 40 years of betting horses.

[limit2]

The replies I get are very instructive and I pay heed to them. I am somewhat convinced that flat wagering and progressions will come to an equality over time. However, the progression process puts in motion what the wagers will be while the flat wagers are unknown beforehand to establish equality. It is necessary to know the wagers of the progression over time to determine the correct unit of wagers for the flat bet process. Thus, I might give favorability to limited progressions................................When I say progressions I mean for 2 selections per race a series of 20 terms that look like 5-$4's, 5-$6's, 5-$8's and 5-$12's. $150 outlay. I am small change and very much unable to DISTORT the handle. Let's take a closer look: Hits $50.20 (@ $2), $121.20 (@ $2), $28.20 (@ $3) and $64 (@ $6). Ten (10) days.....29 races.....cost $172.....return $406.10......profit $234.10....ROI= 136 %. The method requires me to take a look at the various ML's available and take a downward direction with the appropriate overlays (2) and stop play at 1 hit a day.

[pondman]

A serious player/gambler would have an estimate of the #of plays he would be making per season/sitting, and therefore would have an estimate of the needed bankroll. You could run it as a business. You have a reasonable cash flow expectation. I believe it's easier to allocate x amount of money to a season/sitting, rather than say I'll double up if I lose or win. And then periodically adjust your % of bank up or down.

If your hit rate is lower, but you have a positive ROI, you'll have a wide roller coaster of a ride if you have a progressive betting strategy. I believe this is a game that needs to be played with consistency. And if you have a longshot method that needs to be bet consistently to be profitable, then do that. Don't have $2 on a horse that wins, and $100 on a horse that loses.

[thaskalos]

Doesn't the progressive style of betting practically guarantee that our losing bets will be larger than our winning ones?

Is that a desirable effect?

[classhandicapper]

Has anyone ever run any computer simulations on this kind of thing?

Some things that are a mathematical certainty to happen eventually, happen so rarely, perhaps they are unlikely within a single lifetime.

[sammy the sage]

Don't have $2 on a horse that wins, and $100 on a horse that loses.

Been there...done that... :lol: :bang:

[Dave Schwartz]

Doesn't the progressive style of betting practically guarantee that our losing bets will be larger than our winning ones?

Is that a desirable effect?

If you mean, "progress after a loss," exactly the other way around.

[thaskalos]

If you mean, "progress after a loss," exactly the other way around.
Only if we can be sure that a winner is forthcoming... :)

[Some_One]

The Wizard of odds has a great write up on this (Again it assumes a casino game where you have a negative expectation, if your handicapping has a flat profit, then the rules don't apply to you)

http://wizardofodds.com/gambling/betting-systems/

[Dave Schwartz]

That was an excellent link.

Everyone who has ever thought that a money management scheme would save them from a negative expectation should read this.

[jfdinneen]

Dave,

You may wish to review this article, Generalized Martingale Betting Strategy (http://www.math-cs.ucmo.edu/~mjms/2009.3/neal091509.pdf)!

John

[CincyHorseplayer]

Set a % of bankroll goal like you do a bet.Say 25%.Win 25% of your current bankroll then readjust your basic bet %.If you lose that 25% readjust your bet %.You'll still have a compounding effect over time.

[Dave Schwartz]

JF,

That is a well-known ForEx system, I believe.

There is also a fellow who has a variation on this who lives south of me in Dayton, NV. He has always claimed great things. Truth is that he came up with a nice twist on this system.

Correct me if I am wrong but the original system works like this:

Imagine you are playing craps and betting "Pass."

You start with a $1 bet. If you win the first bet, the game is over because you are $1 ahead.

If you lose, you continue wagering $1 until you win once, at which time you wager enough money to get back all that you have lost plus $1. Thusly...

$1 L -1
$1 L -2
$1 L -3
$1 W -2
$3 W +1

Now, if that $3 wager had been a loser, the wager would continue at $3 until you won again, at which time you would bet enough to wind up $1 ahead.

$1 L -1
$1 L -2
$1 L -3
$1 W -2
$3 L -5
$3 L -8
$3 W -5
$6 W +1

Again, if the $6 wager was lost, then the wagers would continue at the $6 level.

One could describe this strategy as "Betting the entire (session) bankroll vs $1 that I can win 2 bets in a row before I go broke."

The ForEx version (which Dick Schmidt introduced me to several years ago) made the original bet a percentage of bankroll.

What is scary about this strategy is that, at some point the bets begin to escalate to a point where you can be betting very large (relative) sums.

Example of escalation:
$1 L -1
$1 L -2
$1 L -3
$1 W -2
$3 L -5
$3 L -8
$3 W -5
$6 L -11
$6 L -17
$6 W -11
$12 L -23
$12 L -35
$12 L -47
$12 L -59
$12 W -47
$48 L -95
$48 L -143
$48 W -95
$96 L -181

You will note that every W-L combination causes a dramatic increase in the continuing bet size!


The variation that the guy in Dayton came up with was really quite ingenious. What it did (as I understood) was begin cycling the bets downward with each losing bet.

Example of escalation with a switch to the "Dayton" version at the $96 mark:
$1 L -1
$1 L -2
$1 L -3
$1 W -2
$3 L -5
$3 L -8
$3 W -5
$6 L -11
$6 L -17
$6 W -11
$12 L -23
$12 L -35
$12 L -47
$12 L -59
$12 W -47
$48 L -95
$48 L -143
$48 W -95
$96 L -181
$48 L -229
$24 L -253
$12 L -265
$6 W -260
$261...

It does not relieve one of the pressure building in a protracted losing session (with a few wins making it all the worse) because ultimately, you must fire that HUGE bet to catch up.


Regards,
Dave Schwartz

PS: Did I describe the correct system?

[Dave Schwartz]

Wow! Kept reading. There is a lot more in there!

[classhandicapper]

The one I liked the best was the 1 -2 -3 system.

It requires that you hit more than 1/3 of your wagers at even money or better.

You start off with 1 - 2 - 3 and bet the sum of the first and last number in the sequence. (bet $4)

If you lose, you put the losing amount at the end of the sequence and again sum the first and last numbers in the sequence as your bet.

1-2-3-4 (bet $5)

If you win, you eliminate the first and last numbers from the sequence and again sum the remaining first and last numbers in the sequence as your bet.

2-3 (bet $5)

When you finally get rid of all the numbers, you will be profitable and can start over.

[limit2]

One can really go to town on progressions. I like to think I can get a hit within 20 races most of the time. Say,83% of the time. At 5-$2'sx2, 5-$3's x 2, 5-$4's x 2 and 5-$6's per race I believe the effort is worthwhile. Get the hit for the day and start the progression again the next day. An approach I hope has merit.

[HUSKER55]

I think everyone has their own betting program that works for them. I am a firm believer of staying within your "comfort zone".

The one I like the best is to bet no more than the square root of your bankroll. When you lose the wagers go down and when you hit the wagers go up.

But that is my comfort zone which may or may not work for anyone else.

[raybo]

I, and Cincy obviously, still think the best wagering system is a combination of percentage of bankroll, and percentage of profit, setting a target profit level/percentage in the very beginning. Bet increases only take place once you meet or exceed your profit amount for that stage. Bet decreases only happen if you lose your previous profit amount.

The amount of bet increase, upon reaching your target profit amount, is a function of both percentage of current bank, and the same percentage of the target profit amount you just made. So, in effect, you are increasing bet size by the compounding of the percentage of current bankroll and the percentage of the recently attained target profit amount.

Decrease in bet size, after losing your last profit amount, is to the bet amount previous to the last bet size increase.

So, in effect, your bet size never increases until you prove you are beating the game, to the tune of your target profit amount/percentage. If you never reach that profit target then you never increase your bet size. If you reach your target then you are only putting that profit at risk, at the higher bet size, rather than your whole bankroll.

One would have to endure a "tremendous" losing streak to ever get back into risking any of your original bankroll, using this method.

NOTE: This method, or any other method, will not work if you do not have a long term edge in the first place.

[Robert Goren]

If you have good numbers, Kelly is the way to go. If you think you have good numbers but are only pretty sure then some % of Kelly is probably the way to go.

I believe Pondman someplace on this site explained how he did it with the systems or methods he uses. It seem pretty reasonable when I read it. Somebody better with search could probably find it, but I can't locate right now.

[raybo]

If you have good numbers, Kelly is the way to go. If you think you have good numbers but are only pretty sure then some % of Kelly is probably the way to go.

I believe Pondman someplace on this site explained how he did it with the systems or methods he uses. It seem pretty reasonable when I read it. Somebody better with search could probably find it, but I can't locate right now.

"True" Kelly requires that you know your edge for each individual bet. The nature of racing does not allow that. Risk of ruin is a constant threat using Kelly or fractional Kelly betting.

[classhandicapper]

In practice I try to stick with flat bets.

Trying to determine my probability of cashing, my exact edge, how much I should bet, how to construct the bet etc.. just adds to the things I have to think about when handicapping. Sometimes I am trying to figure out how the track is playing. I want to focus on that because it's a real time issue that could change my opinion on who to bet. Worrying about the bet size can become distracting to me.

So I start with a flat bet and then just decide which pool to go into.

I push my flat bet higher if I am feeling confident and lower it if I am feeling less confident. It's not always correlated to near term success either. Sometimes my handicapping goes through changes. My bet size will usually decrease when I make changes and then get bigger again when I sort it all out.

I stop pushing it higher when the size starts causing me to be mentally afraid to pull the trigger on lower probability bets. I used to just lower the bet in those situations, but those kinds of fluctuations in bet size almost always cause worse results for me. I wind up betting too little on my biggest edges and vice versa. So I stay with what I feel comfortable with at that time and go from there.

[Sapio]

I, and Cincy obviously, still think the best wagering system is a combination of percentage of bankroll, and percentage of profit, setting a target profit level/percentage in the very beginning. Bet increases only take place once you meet or exceed your profit amount for that stage. Bet decreases only happen if you lose your previous profit amount.

The amount of bet increase, upon reaching your target profit amount, is a function of both percentage of current bank, and the same percentage of the target profit amount you just made. So, in effect, you are increasing bet size by the compounding of the percentage of current bankroll and the percentage of the recently attained target profit amount.

Decrease in bet size, after losing your last profit amount, is to the bet amount previous to the last bet size increase.

So, in effect, your bet size never increases until you prove you are beating the game, to the tune of your target profit amount/percentage. If you never reach that profit target then you never increase your bet size. If you reach your target then you are only putting that profit at risk, at the higher bet size, rather than your whole bankroll.

One would have to endure a "tremendous" losing streak to ever get back into risking any of your original bankroll, using this method.

NOTE: This method, or any other method, will not work if you do not have a long term edge in the first place.

Hi Raybo,

This sounds interesting. Is it possible for you to provide a sequence of bets?

Thomas Sapio

[classhandicapper]

So, in effect, your bet size never increases until you prove you are beating the game, to the tune of your target profit amount/percentage. If you never reach that profit target then you never increase your bet size. If you reach your target then you are only putting that profit at risk, at the higher bet size, rather than your whole bankroll.



I like this a lot.

[raybo]

Hi Raybo,

This sounds interesting. Is it possible for you to provide a sequence of bets?

Thomas Sapio

I did that once, rather complicated. Let me see if I can find the thread. BTW, the bet sizing app is posted on this site somewhere, it's a stand-alone Excel workbook. Probably in the "Software" section.

[raybo]

Hi Raybo,

This sounds interesting. Is it possible for you to provide a sequence of bets?

Thomas Sapio

Here's the thread: http://www.paceadvantage.com/forum/showthread.php?t=100881&highlight=sizing

The betting example is a Word document in RTF file format. It is in post #32

The latest version of the Excel workbook is in post #48, along with the VBA that runs it.

[Sapio]

Here's the thread: http://www.paceadvantage.com/forum/showthread.php?t=100881&highlight=sizing

The betting example is a Word document in RTF file format. It is in post #32

The latest version of the Excel workbook is in post #48, along with the VBA that runs it.

Hi Raybo,

Thank you for your time and effort. You are very generous.

Thomas Sapio

[fmolf]

the method i have found to be the best is fixed % of bankroll.Say you start with $100 and use 10% as your fixed base bet and hit a 2/1 shot, the end result is you bet $10 and won $20. boosting your bankroll to $120 so your next win bet would be $12.....Had you lost the first $10 bet your bankroll would have been $90 and the win bet would have been $9....This method allows for larger bets when on a hot streak and winning, and smaller bets when cold and losing.

[jfdinneen]

Dave,

While in no way recommending a martingale approach, the Neal and Russell (2009) analysis is, principally, of interest in terms of deriving two values - namely, loss multiplier (m) and number of recovery bets (n) for a given payoff (a), win rate (p), and risk of ruin (r) (see example in section 11).
Incidentally, if one were to stick diligently to the key parameters in terms of session trading, a Monte-Carlo analysis of the quoted example suggests a median RoI of approximately 8%!

John