Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Scan for outdated or missing drivers - takes under a minuteDriver Scan →Repair Windows errors before they cause bigger problemsFix Now →ChatGPT can give you a number when you ask for one, but that answer is not evidence that it sampled fairly or followed a distribution you specified. A 2026 evaluation of 11 language models found widespread failures on statistical sampling tasks; its headline results apply to the tested models and protocols, not to ChatGPT alone. So the viral claim that ChatGPT “always” picks 17 is a funny anecdote, not a measured rule.
Why does ChatGPT “always” pick 17?
A community member reported receiving “17, 23, Saturn, triangle” after asking ChatGPT for random items, and said other users had similar experiences. That account is a useful example of how a response can feel oddly predictable, but it is one person’s anecdote—not evidence that ChatGPT always chooses 17 or that the number has a population-wide bias. Read the community post.
The available evidence does not establish why a particular number might recur. It does not show whether repetition comes from training data, the prompt, decoding settings, or another cause. The more important distinction is between an answer that sounds random and a process that reliably samples according to defined statistical rules.
Can ChatGPT generate a genuinely random number?
You can ask ChatGPT to choose a number, but you should not assume its conversational answer is a uniform random draw—or a draw from any other specified distribution. The 2026 Association for Computational Linguistics evaluation by Minda Zhao, Yilun Du, and Mengyu Wang tested 11 language models against 15 probability distributions. It examined both batches of 1,000 samples generated in one response and 1,000 stateless independent requests. The paper’s abstract reports a median pass rate of 7% across the tested models for batch generation; that is not ChatGPT’s individual score. See the ACL paper abstract and bibliographic record.
#1 Best Overall
- Roll A Random Number 1 to 10000!
- 4 Dice Set (UNIT, TENS, HUNDREDS, THOUSANDS)
- Great for Random Numbers & Loot in RPGs
- The Dungeon Master's Friend
What the independent-request result means
In the study’s independent-request protocol—1,000 stateless calls across its distribution suite—10 of the 11 evaluated models passed none of the tested distributions. That result is limited to those models, tests, and pass criteria; it does not establish a score for every ChatGPT version or every way of asking for a number.
Why task and protocol matter
The ACL abstract reports that sampling fidelity declined as distributions became more complex and as the sample horizon grew. It also reports downstream failures when models were used to balance multiple-choice answer positions and to meet demographic targets in constrained image-prompt synthesis. The abstract does not provide a ChatGPT-only breakdown, and the results should not be generalized beyond the tested setup.
Rank #2
- Each set includes 10 pieces of 10 sides dice.
- Standard size 16MM 10 Sides cube with etched number 1-10, it is beautiful color of translucent blue.
- With every number sits very well, the die will randomly give you a number between 1 and 10.
- It is a great addition to any dice game arsenal.
When comparing claims about AI randomness, check which model and version were tested, whether the task asked for one number or repeated samples, whether samples came in a batch or through independent prompts, how many samples were collected, how complex the distribution was, and what statistical test and pass criterion were used.
Does that conflict with earlier research on ChatGPT-3.5?
No. A preliminary 2024 study adapting a human random-number-generation test reported that ChatGPT-3.5 avoided repetitive and adjacent-number patterns more effectively than the human comparison group. That was a different task from checking whether a model faithfully samples a specified probability distribution. Performing well on a human-style sequence task does not demonstrate statistically faithful sampling. Read the 2024 study.
Free tools Windows power users keep installed
One-click scans. No signup required.
Rank #3
- 6 sided dice, each side is numbered 1-6
- Sturdy plastic
- Supports hands-on number and operations activities
- Dice measure 3/4"
- Set contains 3 red, 3 yellow, 3 blue, 3 green dice (Set of 12)
Was a ChatGPT bug responsible for repeated numbers?
OpenAI documented a separate incident in February 2024 in which incorrect inference-kernel results on certain GPU configurations affected number selection during language processing. OpenAI says it rolled out a fix and confirmed that incident was resolved. The write-up concerns that specific technical incident; it does not explain why a casual prompt might produce a repeated or familiar number. Read OpenAI’s incident report.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What should you use when randomness matters?
For a casual, low-stakes choice, a physical die is a simple option you can observe directly. For work that needs statistical guarantees, use an external tool designed for the required sampling task rather than treating a chat response as a random-number generator. The cited evaluation supports that distinction, but does not test particular dice or recommend a specific generator or vendor.
Rank #4
- Roll A Random Number 1 to 10,000,000!
- 7 Dice Set
- Great for Random Numbers & Loot in RPGs
- The Dungeon Master's Friend
The ACL paper’s authors conclude: “These findings indicate that current LLMs lack a functional internal sampler, necessitating external tools for applications requiring statistical guarantees.” That is the authors’ conclusion from their evaluation, not an official claim about every ChatGPT version.
Quick Recap
Best Value
- There are approximately 100 factory second dice in this bag.(a little more or less than 100pcs, all in random) These dice are mixed in a variety of sides, shapes, designs,styles, colors. Every bag does not necessarily have every style & color. Sides may include 4, 6, 8, 10, 12, 20, 50,60,100 sided dice (and possibly other sizes) in colors that include (but are not limited to) marble, glitter, opaque and translucent.
- They are all better rejects dice.
- Can be used for replacement dice. Customize your games with your very own personal set of dice. Dice come in all sorts of types, styles and colors.
- Each set in plain polybag packing.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
