Example: the 1st pattern—which repeats a digit from the previous result—appears 60% of the time (this is just an example; another one, the 4th pattern on the list, appears 25% of the time). The goal is to play based on the highest percentages (ideally covering 100% of the probability distribution); you can organize this better in columns and include other patterns. Essentially, the objective is to identify the highest percentages and the patterns that are "due" (lagging).
Hi, can you help? Take a look at this! Given a list of 800 draws, what is the appearance percentage for each Pick 3 number?
- Thread starter jack
- Start date
Olá. Para esclarecer o objetivo: trata-se de analisar padrões — especificamente, focando nas maiores taxas de ocorrência e nos "intervalos" (desde a última aparição) — e manter as mesmas escolhas em vários sorteios, em vez de jogar aleatoriamente. Em essência, o objetivo é alinhar a estratégia com os resultados estatísticos reais e as frequências dos sorteios, mesmo que o jogo em si seja aleatório.
Frank
Member
Jack. I’m surprised that somebody somewhere in all the years pick3 has existed hasn’t already done this. It’s probably been done a hundred times already. Pick 3 isn’t a lottery we have in the UK. So I have had no incentive to look into this. However, I needed to refresh my experiences with Excel as I’m now in my 80th year and need to keep my mind and memory alive. So, more for my own benefit rather than yours I have been working on this. I’m still debugging it and rectifying some errors. I’ll be posting something soon.
Olá Frank, obrigado por conferir a postagem. Como eu disse, quando eu ganhar um grande prêmio, você será recompensado, já que sempre me ajudou. Sobre a planilha, fique à vontade para adicionar suas próprias ideias, ok? O objetivo é jogar com base nas maiores porcentagens para cada padrão "atrasado" — ou seja, padrões que estão atrasados em termos de colunas, posições, pares gêmeos ou combinações específicas. Por exemplo, se um padrão envolvendo três dígitos diferentes estiver atrasado, eu o jogarei em vários sorteios. Fique à vontade para contribuir com suas ideias.
Hi Frank, thanks for checking out the post. As I’ve said, when I win a big prize, you’ll be rewarded since you’ve always helped me out. Regarding the spreadsheet, feel free to add your own ideas, okay? The goal is to play based on the highest percentages for each "overdue" pattern—meaning patterns that are lagging behind in terms of columns, positions, twin pairs, or specific combinations. For example, if a pattern involving three different digits is overdue, I’ll play it across several draws. You’re welcome to contribute your ideas.
Frank
Member
I just popped in to give a progress update. Since your post I have made some changes to the original spreadsheet including creating a separate sheet to actually generate all the various patterns as a check on my formulas and to give feedback to the original 800 draw sheet highlighting all the results with that chosen ( generated) pattern within the 800, I thought this might be useful. The problem is my analysis gives too much information covering :-
3 odd
3 even
2 odd I even
2 even 1 odd
all digits different
all different digits and odd
sll different digits and even
ascending pairs ( one different) in the same draw
descending pairs ( one different) in the same draw
repeated digits to the next draw
displaying percentage appearance, average skip, maximum skip, most recent skip from which overdue can be calculated.
All this can be overwhelming, so I’m trying to design a summary sheet to assemble only the current skips or delays as you call them and display the overdues alongside the expected delays ( based on the average) and possible maximum values in each case , and your chosen overdue limit which an be adjusted by you.
Im quite busy doing other things, I’ll post it as soon as it’s ready.
3 odd
3 even
2 odd I even
2 even 1 odd
all digits different
all different digits and odd
sll different digits and even
ascending pairs ( one different) in the same draw
descending pairs ( one different) in the same draw
repeated digits to the next draw
displaying percentage appearance, average skip, maximum skip, most recent skip from which overdue can be calculated.
All this can be overwhelming, so I’m trying to design a summary sheet to assemble only the current skips or delays as you call them and display the overdues alongside the expected delays ( based on the average) and possible maximum values in each case , and your chosen overdue limit which an be adjusted by you.
Im quite busy doing other things, I’ll post it as soon as it’s ready.
Okay, Frank, thanks for checking out the post. There’s no rush—feel free to handle it however you like; even basic analysis helps a lot. With Pick 3, we’re dealing with digits 0 through 9, so looking at "lags" (how long it's been since a number appeared) combined with patterns of high frequency should give us a good idea of what to expect—since the range is small (only 10 digits to track). I’ll look for some tips to help out, but feel free to use them only if you need to.
hello,FRANK=
800 Avg. gap Hist. max. Current gap Gauss center Limit Status
3 odd automatic automatic automatic automatic automatic 10 automatic
3 even automatic automatic automatic automatic automatic 10 automatic
2O + 1E automatic automatic automatic automatic automatic 10 automatic
1O + 2E automatic automatic automatic automatic automatic 10 automatic
3 distinct automatic automatic automatic automatic automatic 10 automatic
3 distinct odd automatic automatic automatic automatic automatic 10 automatic
3 distinct even automatic automatic automatic automatic automatic 10 automatic
Ascending automatic automatic automatic automatic automatic 10 automatic
Descending automatic automatic automatic automatic automatic 10 automatic
Repetition automatic automatic automatic automatic automatic 10 automatic
The number 10 is just an example. The limit is located in a configuration cell for you to change.
📈 AND THEN THE GAUSSIAN CURVE COMES IN
This is an important point.
For each pattern, we will store the historical gaps:
Pattern: 3 ODD
gap
1
2
3
2
4
5
3
2
6
4
...
Based on these 800 results, or the set of actually observed occurrences, the spreadsheet calculates:
PATTERN CURVE
│
▼
┌─────────────┐
│ distribution│
│ of gaps │
└──────┬──────┘
│
┌──────┼──────┐
▼ ▼ ▼
lower CENTRAL upper
│ │ │
└──────┼──────┘
▼
current gap
│
▼
comparison with the
central range
Thus, each pattern will have its own curve, rather than a general curve mixing different things.
🔵 THE MOST IMPORTANT PART: LAG
For each pattern:
Current gap
= number of draws since the last occurrence. For example, just to illustrate:
Last occurrence ........ draw 792
Last draw ........... 800
CURRENT LAG = 8
The spreadsheet then compares:
3 ODD NUMBERS
CURRENT LAG
8
│
┌─────────┼─────────┐
▼ ▼ ▼
average central maximum
5 6 12
And shows:
Current = 8
Average = 5
Maximum = 12
Configured limit = 10
This allows you to see where the pattern stands relative to its own history.
🔴 BUT THERE IS AN IMPORTANT CAVEAT
I wouldn't call the central range the "number that is due to appear."
It is a historical statistical range for the pattern.
For example:
Pattern X exhibited a specific lag behavior across the 800 records.
This is different from stating:
"It's going to appear now."
The spreadsheet should simply show the pattern's current state relative to its history.
This aligns perfectly with your idea of using lags as a filter after the analysis.
🌳 SECOND TREE — FINAL FILTER
After all patterns have been calculated:
800 DRAWS
│
▼
ALL PATTERNS
│
▼
INDIVIDUAL CURVE
FOR EACH PATTERN
│
▼
CURRENT LAG
│
┌───────────┼───────────┐
▼ ▼ ▼
NORMAL CENTRAL HIGH
│ │ │
└───────────┼───────────┘
▼
DEFINED LIMIT
│
▼
SELECTED PATTERNS
│
▼
┌──────────────────────┐
│ FINAL CROSS-REFERENCING │
│ OF PATTERNS │
└──────────┬───────────┘
│
▼
OPEN LAGS
│
▼
FINAL TABLE
And here is the major advantage
The spreadsheet doesn't need to display the hundreds of data points your analysis produces.
It can have:
SHEET 1 — BASE_800
The 800 results. SHEET 2 — PATTERNS
All classified patterns.
SHEET 3 — SKIPS
Average, maximum, and current skip for each pattern.
SHEET 4 — CURVES
Individual central curve for each pattern.
SHEET 5 — DELAY_SUMMARY
Only:
Pattern | % | average | maximum | current | center | limit | status
SHEET 6 — TREE
The decision flowchart.
⭐ And I would add a very useful column:
"DISTANCE FROM CENTER"
For example:
CURRENT SKIP = 8
CENTER = 6
DISTANCE = +2
This way, you don't have to look at the entire curve.
The table could end up looking like this:
Pattern % Avg. Max. Current Center Distance Limit
3 Odd — — — 8 6 +2 10
3 Even — — — 3 5 -2 10
2O+1E — — — 7 6 +1 10
1O+2E — — — 4 5 -1 10
This makes reading the 800 draws much cleaner.
And, most importantly, it preserves your original idea: first analyze the pattern individually and its central curve; then look at the delays; and only then cross-reference the patterns.
800 Avg. gap Hist. max. Current gap Gauss center Limit Status
3 odd automatic automatic automatic automatic automatic 10 automatic
3 even automatic automatic automatic automatic automatic 10 automatic
2O + 1E automatic automatic automatic automatic automatic 10 automatic
1O + 2E automatic automatic automatic automatic automatic 10 automatic
3 distinct automatic automatic automatic automatic automatic 10 automatic
3 distinct odd automatic automatic automatic automatic automatic 10 automatic
3 distinct even automatic automatic automatic automatic automatic 10 automatic
Ascending automatic automatic automatic automatic automatic 10 automatic
Descending automatic automatic automatic automatic automatic 10 automatic
Repetition automatic automatic automatic automatic automatic 10 automatic
The number 10 is just an example. The limit is located in a configuration cell for you to change.
📈 AND THEN THE GAUSSIAN CURVE COMES IN
This is an important point.
For each pattern, we will store the historical gaps:
Pattern: 3 ODD
gap
1
2
3
2
4
5
3
2
6
4
...
Based on these 800 results, or the set of actually observed occurrences, the spreadsheet calculates:
PATTERN CURVE
│
▼
┌─────────────┐
│ distribution│
│ of gaps │
└──────┬──────┘
│
┌──────┼──────┐
▼ ▼ ▼
lower CENTRAL upper
│ │ │
└──────┼──────┘
▼
current gap
│
▼
comparison with the
central range
Thus, each pattern will have its own curve, rather than a general curve mixing different things.
🔵 THE MOST IMPORTANT PART: LAG
For each pattern:
Current gap
= number of draws since the last occurrence. For example, just to illustrate:
Last occurrence ........ draw 792
Last draw ........... 800
CURRENT LAG = 8
The spreadsheet then compares:
3 ODD NUMBERS
CURRENT LAG
8
│
┌─────────┼─────────┐
▼ ▼ ▼
average central maximum
5 6 12
And shows:
Current = 8
Average = 5
Maximum = 12
Configured limit = 10
This allows you to see where the pattern stands relative to its own history.
🔴 BUT THERE IS AN IMPORTANT CAVEAT
I wouldn't call the central range the "number that is due to appear."
It is a historical statistical range for the pattern.
For example:
Pattern X exhibited a specific lag behavior across the 800 records.
This is different from stating:
"It's going to appear now."
The spreadsheet should simply show the pattern's current state relative to its history.
This aligns perfectly with your idea of using lags as a filter after the analysis.
🌳 SECOND TREE — FINAL FILTER
After all patterns have been calculated:
800 DRAWS
│
▼
ALL PATTERNS
│
▼
INDIVIDUAL CURVE
FOR EACH PATTERN
│
▼
CURRENT LAG
│
┌───────────┼───────────┐
▼ ▼ ▼
NORMAL CENTRAL HIGH
│ │ │
└───────────┼───────────┘
▼
DEFINED LIMIT
│
▼
SELECTED PATTERNS
│
▼
┌──────────────────────┐
│ FINAL CROSS-REFERENCING │
│ OF PATTERNS │
└──────────┬───────────┘
│
▼
OPEN LAGS
│
▼
FINAL TABLE
And here is the major advantage
The spreadsheet doesn't need to display the hundreds of data points your analysis produces.
It can have:
SHEET 1 — BASE_800
The 800 results. SHEET 2 — PATTERNS
All classified patterns.
SHEET 3 — SKIPS
Average, maximum, and current skip for each pattern.
SHEET 4 — CURVES
Individual central curve for each pattern.
SHEET 5 — DELAY_SUMMARY
Only:
Pattern | % | average | maximum | current | center | limit | status
SHEET 6 — TREE
The decision flowchart.
⭐ And I would add a very useful column:
"DISTANCE FROM CENTER"
For example:
CURRENT SKIP = 8
CENTER = 6
DISTANCE = +2
This way, you don't have to look at the entire curve.
The table could end up looking like this:
Pattern % Avg. Max. Current Center Distance Limit
3 Odd — — — 8 6 +2 10
3 Even — — — 3 5 -2 10
2O+1E — — — 7 6 +1 10
1O+2E — — — 4 5 -1 10
This makes reading the 800 draws much cleaner.
And, most importantly, it preserves your original idea: first analyze the pattern individually and its central curve; then look at the delays; and only then cross-reference the patterns.