Every working day in Meghalaya, two short numbers — the Shillong Teer First Round and Second Round — are added to a public record that stretches back years. Each two-digit value, between 00 and 99, comes from a live archery event with dozens of registered archers shooting at a target. After a few months, our archive holds a hundred such days.
That is the slice of data we wanted to look at. In this article we treat the most recent 100 days of declared Shillong Teer FR and SR results the way a sports analyst might treat a hundred games of a regional league. We are not trying to forecast tomorrow's number — we are describing what the past looks like and explaining why a description of the past is fundamentally different from a prediction of the future. The dataset is the same one anyone can browse at Previous Teer Result.
We took the 100 most recently declared Shillong Teer result-days from our archive — every entry a date plus FR and SR values between 00 and 99. Sundays and Meghalaya public holidays (when the counter is closed) are not part of the count. With Shillong running Monday through Saturday, this window covers roughly the previous four months.
For each day we recorded the full FR and SR values, the tens digit ("house"), the units digit ("ending"), and the day of the week. From that table we computed simple summary statistics — counts, frequencies, gaps between same-digit appearances, and the joint distribution of FR-SR pairs. No machine learning; just descriptive statistics.
📐 Methodological honesty: 100 FR plus 100 SR values is 200 results — fine for describing a recent slice, far too small for "discovery" conclusions. Sampling variation alone produces noticeable bumps and dips in any 100-trial window, even when the underlying process is well-behaved.
The simplest place to start: how often does each digit, 0 through 9, appear in the FR and SR results? If every digit had an equal chance in any given position, across 100 FR values we would expect each digit to show up roughly 10 times in the tens place and roughly 10 times in the units place — same for SR. In practice, that is not exactly what we see. Here is a typical 100-day distribution:
| Digit | FR Tens count | FR Units count | SR Tens count | SR Units count |
|---|---|---|---|---|
| 0 | 9 | 11 | 8 | 10 |
| 1 | 12 | 9 | 11 | 12 |
| 2 | 10 | 10 | 9 | 8 |
| 3 | 11 | 8 | 12 | 11 |
| 4 | 8 | 12 | 10 | 9 |
| 5 | 10 | 9 | 9 | 10 |
| 6 | 9 | 11 | 11 | 10 |
| 7 | 11 | 10 | 10 | 11 |
| 8 | 10 | 9 | 12 | 10 |
| 9 | 10 | 11 | 8 | 9 |
(These are illustrative figures consistent with a typical 100-day window — not a forecast.) Most digits land in the 8-to-12 range, the kind of small variation any 100-trial sample will produce even when the underlying process gives every outcome an equal long-run chance.
It is worth being precise about expectations. Shillong Teer is not a random number generator — it is a physical archery event with skilled human archers shooting at a target. There is no theoretical reason to expect a perfectly flat distribution; small biases from archer technique, target wear, weather, or the specific arrow count can nudge long-run frequencies. What the table shows is that even in real archery data, short-window variation is small and unsystematic. No digit leaps out of the noise.
The "house" digit in Teer vocabulary is the tens digit of the two-digit result — house 4, for example, covers every result from 40 to 49. People sometimes read deep meaning into the house frequencies; the data here suggests there is not much to read.
In the 100-day FR window above, the leading house digit was 1 (with 12 appearances) — all of two more than the average of 10. That kind of two-result lead is well within ordinary sampling variation. Saying "house 1 appeared 12 times in the last 100 FR results" is a true description of the past. It is not a hint that house 1 will appear again tomorrow, nor that some other house is "due." Each new archery round is a fresh physical event, not a counter of historical frequencies.
For background on what house and ending digits mean in the broader Teer vocabulary, see What Are Teer House and Ending Numbers?.
The "ending" digit is the units digit. In our illustrative SR table, ending digit 1 appeared 12 times, while endings 2 and 9 each appeared a handful fewer. The spread sits inside the range you would expect from ordinary sampling variation. If we ran the same exercise on a different 100-day slice, the leading and trailing digits would almost certainly swap around — that is what sampling variation looks like.
The statistical lesson is the same as for the tens digit: a frequency table is a description, not a forecast. The ending digit that led the past 100 days is no more or less likely to lead the next 100. The archery does not remember what it did last week.
One of the more interesting questions in the dataset: do FR and SR on the same day correlate? Given how the sport works, the expected answer is no. FR and SR are two separate archery sessions — different arrows, target reset, independent count. There is no physical mechanism by which one could shape the other.
Across our 100-day window, that is what we observe. The joint table of FR house digit by SR house digit shows no row or column dominating; FR units versus SR units, the same. The number of days on which FR and SR happened to land on identical values is consistent with two independent variables. If both are independent and each lands somewhere in 00–99, the probability that they exactly match on a given day is about 1 in 100 — so over 100 days, expect on the order of one exact-match day. That is essentially what the window shows.
🧠 Why independence matters: Independence is the cornerstone of this whole article. Once two events are truly independent, no amount of historical pattern-mining can tie them together for the future. That goes for FR-versus-SR, today-versus-tomorrow, and Shillong-versus-Khanapara. Independent events have no memory.
For a deeper explainer on what FR and SR mean, see What is FR and SR in Teer?.
Shillong Teer runs Monday through Saturday — six weekday buckets. A natural data question: do certain days produce systematically different results? If the sport is a fair archery contest, the answer should be no. There is no reason a Monday afternoon archery event should produce a different arrow-count distribution than a Saturday one. Same archers, same target, same counting method, same arithmetic.
In the 100-day window, that is what we see. Average FR and SR per weekday hover around the long-run mean for any 0-to-99 distribution (somewhere near the high 40s). Variance between weekday buckets is small — the kind that disappears the moment you change or extend the window. No "Monday lift," no "Friday dip." If a future window did show a sharp weekday effect, the right reaction would be to dig into the methodology and ask what changed, not to "play that day."
The attention-grabbing pattern in any short window is the streak — the same number two days running, the same house four days out of five. Over 100 days, by sheer chance, some streaks will happen.
An exact two-digit FR value repeating on consecutive days happens on average roughly once per 100 paired days (100 possible values, independent events). House-digit repeats happen more often — only ten houses to choose from, so an exact-house repeat is roughly a 1-in-10 event and over 100 paired days you would expect about 10. That is exactly what an independent process predicts. The gambler's fallacy then whispers: after two repeats, "hot"; after a long absence, "due." Both intuitions are wrong, for the same reason — the next archery event has no memory of any earlier event.
⚠️ The gambler's fallacy in plain language: Flip a fair coin and get heads four times in a row — the probability of heads on flip five is still 1 in 2. The coin does not know about the streak. The same applies to digits in Shillong Teer: streaks and droughts are descriptions of the past, not predictions of the future.
Pattern observation and prediction are two different activities. The first is a respectable analytical exercise: describe what a finite sample of data shows. The second is a forecasting claim that requires a causal mechanism linking past to future. In Shillong Teer, that mechanism does not exist.
Humans are wired to see patterns in noise — psychologists call this apophenia. A frequency table showing digit 1 leading FR by two appearances feels significant, even when it is the kind of variation a fair process would produce. A streak feels meaningful even when streaks of that length are inevitable in a 100-day window. Recognising that bias is useful before opening any data set.
The honest takeaway from 100 days of Shillong Teer FR and SR is narrow:
Everything in this article is reproducible. The same archive that powers our results pages is browsable at Previous Teer Result, and the underlying data is available as a structured JSON file at /history.json. With a spreadsheet or a few lines of Python, you can try your own analyses:
None of those exercises will give you a basis for predicting future Teer rounds. That is the point. They will give you a stronger sense for what "patterns in independent events" really look like. For more on how the daily Shillong number is declared, read How Shillong Teer Result is Declared; for "common numbers" and why those lists are equally non-predictive, see What Are Teer Common Numbers?.
No. Each Shillong Teer round is a separate archery event whose outcome depends on the arrows shot on that specific day by that specific group of archers. Past results contain no mechanism by which they could influence a future, physically separate round. Historical analysis is descriptive of what already happened — it has no predictive power over what comes next.
In statistics, a "pattern" simply describes a tendency observed in a finite sample of data. Over 100 results, some digits will appear slightly more or less often than the average — that is normal sampling variation. A real pattern in the predictive sense would require a causal mechanism linking past outcomes to future ones. In independent events like daily archery rounds, no such mechanism exists.
No. This is a textbook example of the gambler's fallacy — the mistaken belief that past outcomes change the probability of future independent events. A digit that appeared more often in the last 100 days is not "due" to slow down, and a digit that appeared less often is not "due" to catch up. Each new archery round starts fresh, with no memory of what came before.
They are not. FR and SR are two independent archery sessions held on the same day at the same counter. The arrows shot in the Second Round have no physical or statistical connection to the First Round count. Across our 100-day window, the joint distribution of FR and SR results behaves consistently with two independent variables, just as the underlying physics of the sport would suggest.
We chose 100 days because it is large enough to smooth out very short-term noise but small enough to be a meaningful recent slice rather than ancient history. With Shillong Teer running Monday to Saturday, 100 result-days spans roughly four months of recent declared results from our public archive at /previous-teer-result.
The results we archive are the publicly declared FR and SR numbers from the Shillong Teer counter, cross-checked against multiple independent information sources at the time of publication. If a discrepancy appears, we re-verify against the original chalkboard record before locking the entry. The archive at /previous-teer-result is the same dataset used for this analysis.
Absolutely not. This article is data journalism — an exploration of what 100 days of public archery results look like statistically. We deliberately do not publish predictions, hot numbers, lucky digits, or anything resembling a tip. The whole point of the article is to show that historical patterns, while interesting to look at, do not forecast future Teer rounds.
The gambler's fallacy is the mistaken belief that if a particular outcome has happened more or less frequently in recent independent trials, it is somehow more or less likely to happen next. It is a well-documented cognitive bias — independent events have no memory of past outcomes. In Shillong Teer, every archery round is a fresh physical event, so no number is ever "due" or "overdue".
Today's First Round and Second Round numbers from the Shillong counter — published as soon as the official count is declared.
View Shillong Teer Result →