systemst pick3Prev Topic Next Topic

New Topic New Poll

hypersoniq

Pennsylvania
United States
Member #1,340
April 6, 2003
2,939 Posts
Offline

Dec 29, 2024, 6:41 am

ARIMA does not work. It always puts the results in the 4.8 to 5.2 range... you can't play 4.x or 5.x, and 4s and 5s are not always in the winning combos. I tried that with both ARIMA and SARIMA, which is seasonally adjusted ARIMA.

SARIMA found no seasonal trends either.

Have an EXCELlent day!
dr san

bgonÃ§alves
Brasil
Member #92,558
June 9, 2010
4,155 Posts
Offline

Dec 29, 2024, 7:29 am

hello hyper=

It is common to encounter difficulties when applying ARIMA and SARIMA models, especially when the results are concentrated in a specific range and do not reflect the expected variability of the data. Let's look at the possible causes and solutions for the problems you are facing:

Why are ARIMA results concentrated in a range?
Stationary data: If your data is already very stationary (little variability), the model may have difficulty capturing more complex patterns and generating predictions with a wider range.
Lack of variability in the data: The absence of significant events or changes in the data may limit the model's ability to learn broader patterns.
Incorrect parameters: The values of p, d, and q (and P, D, Q for SARIMA) may not be appropriate for your data. An inappropriate choice can lead to an overfitting or underfitting model.
Uncaptured seasonality: If there is seasonality in your data and SARIMA did not identify it correctly, the predictions may be biased.
Other relevant variables: There may be other variables that influence the results and that are not being considered in the model.
What can be done to improve the results?
Exploratory Data Analysis (EDA):

Visualization: Create line graphs, histograms, boxplots, and autocorrelation graphs to understand the distribution of data, the presence of trends, seasonality, and outliers.
Descriptive statistics: Calculate mean, median, standard deviation, and coefficient of variation to assess data variability.
Stationarity tests: Apply tests such as Augmented Dickey-Fuller (ADF) to verify whether the data is stationary.
Data preprocessing:

Transformations: Apply logarithmic or power transformations to stabilize the variance and linearize the relationship between variables.
Outlier removal: Identify and remove outliers that may significantly influence the model.
Creation of new variables: Generate new variables that capture relevant information, such as trends, seasonality, or interactions between variables.
Parameter selection:

Automatic methods: Use functions such as auto.arima in R to find the best p, d, and q values.
Information criteria: Compare models with different parameter combinations using criteria such as AIC or BIC.
Cross-validation: Split the data into training and test sets to evaluate model performance.
Consider other models:

Exponential models: If your data shows exponential growth, models such as Holt-Winters may be more appropriate.
Regression models: If there are relevant explanatory variables, linear or nonlinear regression models can be used.
Machine learning models: For more complex data, algorithms such as recurrent neural networks (RNNs) or random forest models can be explored.
Residual analysis:

Check for stationarity: Residuals should be approximately independent and identically distributed (i.i.d.).
Normality test: Check whether the residuals follow a normal distribution.
Autocorrelation: Check for autocorrelation in the residuals, which would indicate that the model did not capture all the information.
Additional tips:
Refer to the documentation: Read the documentation of the functions you are using to understand the available options and parameters.
Explore different packages: There are several packages in R (forecast, statsmodels in Python) that offer different implementations and functionalities for ARIMA models.
Participate in communities: Discuss your questions in online forums such as Stack Overflow or R communities to get help from other users. Example in R (using the forecast package):

Code snippet

library(forecast)
# Assuming your data is in an object called 'data'

# Finding the best parameters automatically
autoarima_model <- auto.arima(data)

# Making predictions
forecasts <- forecast(autoarima_model, h = 12) # Predicting the next 12 periods

# Plotting the predictions
plot(forecasts)
Remember: Choosing the best model depends on the nature of your data and the purpose of your analysis. Experiment with different approaches and adjust the parameters until you find a model that fits your data well and generates accurate predictions.
Pick3geniu$

New York
United States
Member #203,701
January 14, 2020
380 Posts
Offline

Dec 29, 2024, 3:51 pm

Quote: Originally posted by dr san on Dec 29, 2024
hello hyper=

It is common to encounter difficulties when applying ARIMA and SARIMA models, especially when the results are concentrated in a specific range and do not reflect the expected variability of the data. Let's look at the possible causes and solutions for the problems you are facing:

Why are ARIMA results concentrated in a range?
Stationary data: If your data is already very stationary (little variability), the model may have difficulty capturing more complex patterns and generating predictions with a wider range.
Lack of variability in the data: The absence of significant events or changes in the data may limit the model's ability to learn broader patterns.
Incorrect parameters: The values of p, d, and q (and P, D, Q for SARIMA) may not be appropriate for your data. An inappropriate choice can lead to an overfitting or underfitting model.
Uncaptured seasonality: If there is seasonality in your data and SARIMA did not identify it correctly, the predictions may be biased.
Other relevant variables: There may be other variables that influence the results and that are not being considered in the model.
What can be done to improve the results?
Exploratory Data Analysis (EDA):

Visualization: Create line graphs, histograms, boxplots, and autocorrelation graphs to understand the distribution of data, the presence of trends, seasonality, and outliers.
Descriptive statistics: Calculate mean, median, standard deviation, and coefficient of variation to assess data variability.
Stationarity tests: Apply tests such as Augmented Dickey-Fuller (ADF) to verify whether the data is stationary.
Data preprocessing:

Transformations: Apply logarithmic or power transformations to stabilize the variance and linearize the relationship between variables.
Outlier removal: Identify and remove outliers that may significantly influence the model.
Creation of new variables: Generate new variables that capture relevant information, such as trends, seasonality, or interactions between variables.
Parameter selection:

Automatic methods: Use functions such as auto.arima in R to find the best p, d, and q values.
Information criteria: Compare models with different parameter combinations using criteria such as AIC or BIC.
Cross-validation: Split the data into training and test sets to evaluate model performance.
Consider other models:

Exponential models: If your data shows exponential growth, models such as Holt-Winters may be more appropriate.
Regression models: If there are relevant explanatory variables, linear or nonlinear regression models can be used.
Machine learning models: For more complex data, algorithms such as recurrent neural networks (RNNs) or random forest models can be explored.
Residual analysis:

Check for stationarity: Residuals should be approximately independent and identically distributed (i.i.d.).
Normality test: Check whether the residuals follow a normal distribution.
Autocorrelation: Check for autocorrelation in the residuals, which would indicate that the model did not capture all the information.
Additional tips:
Refer to the documentation: Read the documentation of the functions you are using to understand the available options and parameters.
Explore different packages: There are several packages in R (forecast, statsmodels in Python) that offer different implementations and functionalities for ARIMA models.
Participate in communities: Discuss your questions in online forums such as Stack Overflow or R communities to get help from other users. Example in R (using the forecast package):

Code snippet

library(forecast)
# Assuming your data is in an object called 'data'

# Finding the best parameters automatically
autoarima_model <- auto.arima(data)

# Making predictions
forecasts <- forecast(autoarima_model, h = 12) # Predicting the next 12 periods

# Plotting the predictions
plot(forecasts)
Remember: Choosing the best model depends on the nature of your data and the purpose of your analysis. Experiment with different approaches and adjust the parameters until you find a model that fits your data well and generates accurate predictions.

dr san, these all seem like AI generated results that you always post. Have any of these supposed AI results worked for you thus far or how well have they worked for you? It does not need to be that complicated or involved at all with ARIMA models, SARIMA models, visualization, transformations, Exponential models, Regression models, and so on to make predictions. How about a KISS model for predicting lottery draws?
dr san

bgonÃ§alves
Brasil
Member #92,558
June 9, 2010
4,155 Posts
Offline

Dec 29, 2024, 4:33 pm

not predicting,

why a KISS model doesn’t work for lotteries: Intrinsic Randomness: Lotteries are designed to be games of chance, based on random draws. There are no historical patterns, number sequences, or any other factors that can be used to accurately predict the numbers that will be drawn.Independence of Events: Each draw is an independent event. The outcome of a previous draw has no influence on the next one in any way.Large Number of Possible Combinations: In a lottery like Pick 3, the number of possible combinations is enormous. This makes it virtually impossible to analyze all the combinations and identify a pattern.Probability Theory: Probability theory tells us that every number has an equal chance of being drawn. There are no “hot” or “cold” numbers in the long run.Why complex models don’t work either: Noise: Even complex models, which use machine learning algorithms and large data sets, tend to find patterns in noise. That is, they can identify spurious correlations that have no real meaning. Overfitting: Models that are too complex can overfit the training data, losing the ability to generalize to new data.What you can do: Manage Expectations: It’s important to understand that there is no magic formula for winning the lottery. The probability of hitting all the numbers is extremely low.Have Fun: The lottery can be a fun way to participate in a game of chance, but it’s crucial not to spend more than you can afford to lose.Learn Probability: Understanding the basics of probability can help you make more informed decisions about gambling.
Brock Lee

us
Member #231,828
May 4, 2023
1,630 Posts
Offline

Dec 29, 2024, 4:52 pm

Quote: Originally posted by dr san on Dec 29, 2024
not predicting,

why a KISS model doesn’t work for lotteries: Intrinsic Randomness: Lotteries are designed to be games of chance, based on random draws. There are no historical patterns, number sequences, or any other factors that can be used to accurately predict the numbers that will be drawn.Independence of Events: Each draw is an independent event. The outcome of a previous draw has no influence on the next one in any way.Large Number of Possible Combinations: In a lottery like Pick 3, the number of possible combinations is enormous. This makes it virtually impossible to analyze all the combinations and identify a pattern.Probability Theory: Probability theory tells us that every number has an equal chance of being drawn. There are no “hot” or “cold” numbers in the long run.Why complex models don’t work either: Noise: Even complex models, which use machine learning algorithms and large data sets, tend to find patterns in noise. That is, they can identify spurious correlations that have no real meaning. Overfitting: Models that are too complex can overfit the training data, losing the ability to generalize to new data.What you can do: Manage Expectations: It’s important to understand that there is no magic formula for winning the lottery. The probability of hitting all the numbers is extremely low.Have Fun: The lottery can be a fun way to participate in a game of chance, but it’s crucial not to spend more than you can afford to lose.Learn Probability: Understanding the basics of probability can help you make more informed decisions about gambling.

thanks for that explanation of the randomness of lotteries, chad g petey. unfortunately, the man you are impersonating, aka dr san, believes that lotteries can be predicted by identifying patterns. since dr san cannot read english very well, he doesn't know that what you wrote directly contradicts everything he has posted here the last 14 years.

i demand you untie him and restore his ability to post gibberish and beg people to write programs for him. dr san is like dr pepper, the imitations just don't hit the spot.
Pick3geniu$

New York
United States
Member #203,701
January 14, 2020
380 Posts
Offline

Dec 29, 2024, 9:23 pm

Quote: Originally posted by dr san on Dec 29, 2024
not predicting,

why a KISS model doesn’t work for lotteries: Intrinsic Randomness: Lotteries are designed to be games of chance, based on random draws. There are no historical patterns, number sequences, or any other factors that can be used to accurately predict the numbers that will be drawn.Independence of Events: Each draw is an independent event. The outcome of a previous draw has no influence on the next one in any way.Large Number of Possible Combinations: In a lottery like Pick 3, the number of possible combinations is enormous. This makes it virtually impossible to analyze all the combinations and identify a pattern.Probability Theory: Probability theory tells us that every number has an equal chance of being drawn. There are no “hot” or “cold” numbers in the long run.Why complex models don’t work either: Noise: Even complex models, which use machine learning algorithms and large data sets, tend to find patterns in noise. That is, they can identify spurious correlations that have no real meaning. Overfitting: Models that are too complex can overfit the training data, losing the ability to generalize to new data.What you can do: Manage Expectations: It’s important to understand that there is no magic formula for winning the lottery. The probability of hitting all the numbers is extremely low.Have Fun: The lottery can be a fun way to participate in a game of chance, but it’s crucial not to spend more than you can afford to lose.Learn Probability: Understanding the basics of probability can help you make more informed decisions about gambling.

You had dismissed patterns referring to "apophenia" and you downplay them in some posts like in this one, yet you speak about identifying patterns in other posts. Either you believe in patterns that naturally arise in the game or you don't. It is true that there is no way to accurately predict what the winning numbers will be. However, it's more about management of how and when you play and about mitigating losses. But, what you fail to realize or to accept is that the intrinsic randomness you speak of works to your benefit in that it is this inherent randomness that naturally gives rise to patterns that you can use. The KISS model is always what's best to go with. There's no reason to make this more complicated or more involved then it needs to be.
BobP

Dump Water Florida
United States
Member #380
June 5, 2002
3,678 Posts
Offline

Dec 30, 2024, 4:35 am

The secret to Pick-3 is only play Straight, Positional Wheeling and finding a place that pays better then half the odds.

BobP

If you're glowing, see a doctor, get it checked out.
dr san

bgonÃ§alves
Brasil
Member #92,558
June 9, 2010
4,155 Posts
Offline

Dec 30, 2024, 4:42 am

hello brock lee=

Luck vs. Probability: Unraveling the Mystery of Lotteries in Light of Laplace's Theory
The distinction between luck and probability is fundamental to understanding games of chance, such as lotteries. Although they are often used as synonyms, they have distinct and complementary meanings.

Luck is a subjective concept, linked to the idea of chance, destiny and unpredictable events. It is the feeling that something positive or negative happens randomly, without an apparent cause.

Probability, in turn, is a mathematical concept that quantifies the chance of an event occurring. It is an objective measurement, based on calculations and statistical analysis.

Laplace's Theory and Lotteries
Laplace's theory, a milestone in the history of probability, provides us with the tools to calculate the probability of random events, such as the result of a lottery draw.

How Laplace's theory applies to lotteries:

Calculating Probabilities: Using combinatorial formulas, we can calculate the exact probability of getting the numbers right in a lottery draw. For example, in Mega-Sena, the chance of getting all six numbers right is extremely low, which demonstrates the random nature of the game.

Understanding Chance: Laplace's theory helps us understand that the lottery is a game of pure chance. There are no strategies or systems that can guarantee victory, since the outcome is determined randomly.

Managing Expectations: By knowing the probability of winning, we can have more realistic expectations and avoid the illusion that we can control the outcome.

Luck in the Lottery

Although probability is the basis for understanding lotteries, luck still plays an important role. After all, winning the lottery is a rare event, and those who win often attribute their victory to luck.

However, it is important to emphasize:

Luck is random: There is no magic formula for being lucky.

Probability cannot be manipulated: No matter what we do, the probability of winning the lottery will remain the same.

Luck is a psychological factor: Belief in luck may motivate people to play, but it does not increase their chances of winning.

In summary:

Laplace's theory allows us to calculate the probability of random events, such as lottery draws. Luck, on the other hand, is a subjective concept linked to chance. Although luck can play a role in life, in lotteries probability is the determining force.

To illustrate this better, think of a die:

Probability: The chance of rolling a specific number (for example, 6) is 1/6. This is a fixed probability, calculated mathematically.

Luck: If you roll the die several times, you may roll the number 6 several times in a row or not at all. This is considered good or bad luck, but it does not change the probability of rolling a 6 on each roll.

In conclusion:

Laplace’s theory offers us a clearer and more rational view of games of chance, such as lotteries. By understanding probability, we can make more informed decisions and avoid falling into false expectations. However, the thrill and hope of winning, linked to the concept of luck, remain strong motivators for many gamblers.
dr san

bgonÃ§alves
Brasil
Member #92,558
June 9, 2010
4,155 Posts
Offline

Dec 30, 2024, 5:01 am

hello bobp and block lee=

Creating a Complex System for Pick 3 or Hundreds: Frequency Analysis and Patterns
Understanding Complexity

When seeking a more complex system for Pick 3 or Hundreds, we are exploring strategies that go beyond simply analyzing individual numbers. This approach involves identifying deeper patterns and trends in historical draw data, such as the frequency of groups of numbers, the distance between numbers drawn (jumps or deltas), and specific patterns in each position.

Key Elements of the Complex System:

Frequency Analysis:

Groups of numbers: Identify groups of numbers that tend to appear together frequently. For example, if the numbers 1, 2, and 3 appear together frequently, it may be interesting to create combinations with these numbers.
Jumps or deltas: Analyze the distance between numbers drawn in each position. Patterns of large or small jumps can indicate trends.
Patterns in each position: Note which numbers are most frequent in each position (first, second, and third digit).
Filter Creation:

Frequency filters: Filter data to identify numbers that meet specific frequency criteria. For example, numbers that have appeared at least X times in the last Y draws.
Pattern filters: Create filters to identify combinations that follow specific patterns, such as ascending or descending sequences.
Model Building:

Statistical models: Use statistical models to identify correlations between different variables and make predictions.
Artificial intelligence: Use machine learning algorithms to identify complex patterns in data.
Combination Generation:

Pattern-based combinations: Generate combinations that follow the patterns identified in the analysis.
Random combinations within filters: Create random combinations, but always respecting the defined filters.
Practical Example:

Imagine that, after analyzing historical lottery data, you identify the following patterns:

Frequent groups: {1, 2, 3}, {5, 6, 7}
Common jumps: 2 or 3 positions
Patterns by position: The number 7 appears frequently in the first position, and even numbers tend to occupy the last position.
Based on this data, you can create the following combinations:

Combinations based on groups: 123, 567
Combinations with jumps: 135, 246
Combinations with patterns by position: 7XY, XYZ (where X and Y are even numbers)
Important Considerations:

Complexity vs. Simplicity: More complex systems can generate a large number of combinations, which can increase the cost of betting. It is important to find a balance between complexity and feasibility.
Adaptability: Patterns can change over time, so it is essential to update the system regularly.
Human Factor: Data analysis and the creation of complex systems require knowledge in statistics, programming and, often, the use of specialized tools.
Luck: Even with a sophisticated system, the lottery is still a game of chance.
Useful Tools:

Spreadsheets: Excel or Google Sheets can be used to analyze data and create filters.
Programming Languages: Python, R and MATLAB are excellent options for more complex data analysis and model building.
Specialized Software: There are specific software programs for lottery analysis that offer tools to identify patterns and generate combinations.
Conclusion

A complex system for Pick 3 or Hundred can offer a competitive advantage by identifying patterns and trends that may not be apparent in a superficial analysis. However, it is important to remember that there is no magic formula for winning the lottery. The combination of data analysis, system creation and a little luck is the key to increasing
Pick3geniu$

New York
United States
Member #203,701
January 14, 2020
380 Posts
Offline

Dec 30, 2024, 11:49 am

Quote: Originally posted by Brock Lee on Dec 29, 2024
thanks for that explanation of the randomness of lotteries, chad g petey. unfortunately, the man you are impersonating, aka dr san, believes that lotteries can be predicted by identifying patterns. since dr san cannot read english very well, he doesn't know that what you wrote directly contradicts everything he has posted here the last 14 years.

i demand you untie him and restore his ability to post gibberish and beg people to write programs for him. dr san is like dr pepper, the imitations just don't hit the spot.

Hi Brock. So, that's not dr san? What's going on here or where is the real dr san?
Brock Lee

us
Member #231,828
May 4, 2023
1,630 Posts
Offline

Dec 30, 2024, 11:53 am

Quote: Originally posted by Pick3geniu$ on Dec 30, 2024
Hi Brock. So, that's not dr san? What's going on here or where is the real dr san?

dr san used to get a lot of grief for his nonsensical posts written in badly translated english from portuguese. so he got "wise" and started using ai to generate his posts in flawless, verbose english. unfortunately dr san doesn't understand a word of what he's posting now. that's my best guess as to what's going on here.
Brock Lee

us
Member #231,828
May 4, 2023
1,630 Posts
Offline

Dec 30, 2024, 11:55 am

olá dr san=

seus escritos no fórum eram mais interessantes quando você não usava inteligência artificial para escrevê-los.
AllenB

California
United States
Member #167,120
June 27, 2015
1,777 Posts
Offline

Dec 30, 2024, 11:56 am

Quote: Originally posted by Pick3geniu$ on Dec 30, 2024
Hi Brock. So, that's not dr san? What's going on here or where is the real dr san?

I'm worried. AI is going to replace All Humans, Starting with Dr. Jack.

Link removed due to language. Sorry for that. I should have watched it first.
dr san

bgonÃ§alves
Brasil
Member #92,558
June 9, 2010
4,155 Posts
Offline

Dec 30, 2024, 1:12 pm

hello,allenb= AI is a great tool for anyone who wants to use it, even in lore. It's like electricity, you have to respect its characteristics, conduction and insulation. Everyone here knows that the lottery is random, but let's give luck a chance, looking at patterns in the Gauss curve, using filter blocks, trying to establish references, etc.
dr san

bgonÃ§alves
Brasil
Member #92,558
June 9, 2010
4,155 Posts
Offline

Dec 30, 2024, 1:33 pm

hello, block lee ai is present like computing, the internet, etc. there is no going back, and it will evolve, it will help to give better opportunities to luck in winning the lottery, how will lottery administrators deal with this?

New Topic New Poll

Subscribe to this topic

systemst pick3Prev TopicNext Topic

systemst pick3Prev Topic Next Topic