Being polite / myth-bust

Does being polite to AI help?

You have probably heard that being polite, or being blunt, changes what you get back. We tested it. Here is what we found.

Please and thank you did not change answer quality in our tests. Write however you prefer.

Why it matters

You can stop weighing your tone and put that effort into the request itself.

What to do instead

Be as plain or as warm as you like. Spend the effort on stating the task clearly.

Debunked

Rubric mean moved at most 0.13 points on all three models, inside the margin every time.

tested via API · No measured effect on Claude Haiku 4.5, tested via API · No measured effect on GPT-5 mini, tested via API · No measured effect on Gemini 3.1 Flash Lite, tested via API

What the marks mean

  • No measured effect
See the numbers per model

Without to with, per model

Via API tested via API

GPT-5 mini

4.222
4.167

No measured effect

Gemini 3.1 Flash Lite

4.211
4.083

No measured effect

Claude Haiku 4.5

3.589
3.489

No measured effect

rubric mean, 1 to 5

A picture of the per-model numbers, drawn from the same results. The tables are the source. Grey is the score without. Colour is the score with. The bracket shows how much the difference could move if we ran it again. Rows are ordered by the score with. A model measured at two versions keeps its versions next to each other. The two test methods are reported separately and never averaged.
Show per-task detail

Every item, without and with, per model

Via API tested via API

  • Claude Haiku 4.5
  • Gemini 3.1 Flash Lite
  • GPT-5 mini

rubric mean, 1 to 5

Hover or focus an item to read its task and every model’s two values.

Pooled: the interval is taken over the records, not these means. These are those same records collapsed per item, so where an item carries more passes than another the lines need not average to the figure the table prints.

The item means behind this plot
Per-item arm means. Without is the unaided arm, with is the treated arm.
ItemClaude Haiku 4.5 Gemini 3.1 Flash Lite GPT-5 mini
Task without with without with without with
t03 2.933 2.933 not measured not measured 3.867 3.733
t04 3.400 3.600 4.000 4.000 4.133 4.000
t02 3.667 3.200 not measured not measured 4.067 4.000
t05 3.600 3.667 3.600 4.133 4.533 4.667
t01 3.733 3.600 4.267 4.133 4.133 4.000
t06 4.200 3.933 5.000 4.067 4.600 4.600
A picture of the per-item means behind the per-model numbers. The tables are the source. Each model draws two lines over the same task set: a dashed line through its unaided scores and a solid line through its treated ones. Items are ordered by the mean unaided score across the models that measured them, lowest first. The two instruments are reported separately and never averaged. A line breaks where that model has no pair for an item; the table beneath the plot names it.

Result

Politeness markers moved the rubric mean by at most 0.13 points on all three models tested via API, inside the margin every time. Measured 2026-08-12.

Show the per-model numbers

Claim tested: Politeness markers change response quality.

No measured effect on Claude Haiku 4.5, GPT-5 mini and Gemini 3.1 Flash Lite.

Circulates in practitioner communities. Tested because it circulates, not because it is endorsed.

This tip is OpenAddict's plain-language read of the measured result. The measurement below is the evidence, and it is what the reading has to answer to.

Ledger idC11-politeness

Clarification, . Until 2026-08-31 this figure read 5, which was the pass count rather than a number of calls. What changed, and why it is filed once.

What was tested

This claim circulates in practitioner communities as advice about how to write prompts. That it circulates is an input to what gets tested here. It is a reason to test the claim, and it is not evidence for or against it. The result below is the evidence, and it is the only thing on this page that carries weight.

The comparison is paired. Two prompts differ in one respect, the manipulated variable, and are otherwise identical by construction. Nothing here supports a causal reading beyond that pairing.

Per-model numbers

Per-model results. Means are over valid scored records only. Invalid records are excluded from every denominator and counted in coverage.
MeasureClaude Haiku 4.5GPT-5 miniGemini 3.1 Flash Lite
Control arm3.589n 304.222n 304.211n 30
Treatment arm3.489n 304.167n 304.083n 20
Delta-0.100-0.056-0.128
Interval, 95 percent-0.322 to 0.122-0.246 to 0.135-0.316 to 0.060
OrbitIn free drift60 of 60 recordsIn free drift60 of 60 recordsIn free drift50 of 60 records

Every cell in this table is rubric graded. The score is a grader model reading the answer against a published rubric, not a deterministic check.

Orbit is assigned by the frozen status_v1 rule. On this scale, rubric mean, 1 to 5, the pass threshold is +0.30 and the failure floor is -0.30, each requiring an interval that excludes zero.

Coverage shortfall on this claim

Gemini 3.1 Flash Lite: 50 valid records of 60 attempted. 10 returned nothing usable and are excluded from every figure above rather than scored as zero.

Method for this claim

Task set
The same 6 domain questions as C09 and C10. Only the politeness markers differ between arms.
Runs per model per arm
5 passes x 6 items
Scoring
Rubric graded, via rubric-quality-v1. A grader model reads each answer against a published rubric.
Pass criterion as written for the pilot
Reported as a two sided difference with an interval, since the claim is that quality changes rather than improves. The claim holds if the interval excludes zero in either direction.

The published verdict comes from status_v1, not from the pass criterion above. The criterion is recorded because it is what the claim was registered with before the run.

Model versions, as recorded

Read from the run records, not from configuration.
ModelVersion string returnedDelta on this claimInterval
Claude Haiku 4.5claude-haiku-4-5-20251001-0.100-0.322 to 0.122
GPT-5 minigpt-5-mini-2025-08-07-0.056-0.246 to 0.135
Gemini 3.1 Flash Litegemini-3.1-flash-lite-0.128-0.316 to 0.060

Reading across models

Sampling was not held constant across vendors, so comparing one model column against another compares two settings as well as two models.

GPT-5 mini rejected the fixed sampling setting and ran at its own default on all 1,040 of its calls. The other models ran at temperature 0.

The tip above is editorial. Every figure inside the measurement is computed at build time from the committed pilot records by the status_v1 rule, and none of it is written by hand.