Stay With Your Draw
Asystole

I put the easiest cases on my team.

That's a selection strategy.
Try a trial.

Random allocation puts chance in charge;
Known and unknown causes can balance at large.
Balance in expectation; samples still can lean;
A random list won't make all baseline columns clean.
Conceal the next assignment while recruiting to the test;
No peeking at tomorrow's group and choosing who looks best.
After allocation, blinding plays its part:
Who knows which treatment? Say that from the start.
Patients, staff, assessors: name who couldn't tell;
"Double blind" alone won't tell the whole tale well.
Expectations can change behaviour and the scores;
A masked assessor helps protect those measurement doors.
Some treatments make blinding staff or patients tough;
Still plan fair assessment. The label isn't enough.
You called it random because your plan was improvised;
That word has filed a complaint about how you advertised.

Chance makes the groups; protect the draw.
Measure what you planned; show what you saw.
Intention-to-treat: keep assigned groups in view.
Your favourite finishers don't stand for the crew.
Chance makes the groups; protect the draw.
Measure what you planned; show what you saw.
Intention-to-treat: keep assigned groups in view.
Your favourite finishers don't stand for the crew.

A patient changes treatment; another stops the pills;
Keep collecting outcomes with the safeguards and the skills.
For the assignment effect, analyse as assigned;
Intention-to-treat keeps the original groups aligned.
You moved every poor responder to the other side;
Your miracle needed a taxi and a doctored ride.
An adherence effect asks a different thing;
Per-protocol questions need assumptions that you bring.
Simply dropping those who stopped can bias what you see;
Reasons for stopping may predict how things will be.
Missing outcomes still require an honest plan;
The letters I-T-T won't find a vanished scan.
State the target effect; explain treatment changes on the way;
Probe missing-data assumptions before you claim the day.
Predefine the outcome, the timing and the score;
You changed the finish line after seeing who reached the door.

Chance makes the groups; protect the draw.
Measure what you planned; show what you saw.
Intention-to-treat: keep assigned groups in view.
Your favourite finishers don't stand for the crew.
Chance makes the groups; protect the draw.
Measure what you planned; show what you saw.
Intention-to-treat: keep assigned groups in view.
Your favourite finishers don't stand for the crew.

Allocation concealment: keep the next assignment unknown.
Blinding: keep treatment knowledge masked, and state for whom.
Random assignment and random sampling differ;
One supports the comparison; one concerns who's in the picture.
A trial's internal strength still needs a careful read;
Who was recruited? Does this setting match the need?
Benefits and harms both belong upon the sheet;
A beautiful design still has to earn its clinical seat.

Chance makes the groups; protect the draw.
Measure what you planned; show what you saw.
Intention-to-treat: keep assigned groups in view.
Your favourite finishers don't stand for the crew.
Chance makes the groups; protect the draw.
Measure what you planned; show what you saw.
Intention-to-treat: keep assigned groups in view.
Your favourite finishers don't stand for the crew.

And my ten dropouts are probably all fine.

You'd better find out.

Fair. Follow-up first. Then the victory line.