Pistol reliability: attributing your malfunctions instead of enduring them
Published on · 8 min read · Anthony Baillon
A stoppage does not say where it came from. The pistol, the ammunition lot and the magazine produce the same symptoms, and none of them stands out on an isolated malfunction. What separates them fits in three columns logged as the malfunction happens — and in one division.
Why attribution is impossible from memory
The three main causes of a malfunction produce the same symptoms. A feeding failure can come from a magazine's lips, from underpowered ammunition or from a fouled ramp: what you see on the firing line is identical in all three cases. Our guide to pistol malfunctions sets out that reading, symptom by symptom.
Hence the real difficulty: the diagnosis is not made on the malfunction, it is made on the series. And a series cannot be reconstructed from memory at the end of a season. Nobody remembers which magazine was in the gun on the third Saturday in March.
That is the whole problem three columns solve.
The three variables, and the only one that takes effort
| Variable | How to log it | Cost |
|---|---|---|
| The pistol | model, and the round count if you keep one | none |
| The ammunition lot | brand, bullet weight, and the lot number on the box | none |
| The magazine | its number — assuming it has one | one punch mark |
The first two are free. The third is the highest-return step in the whole method, and it is the one almost nobody takes.
An unnumbered magazine is invisible. It blends into the pile, its malfunctions get charged to the pistol, and it keeps circulating for years. Numbered, it gives itself away in two or three sessions. A punch mark on the base plate is enough, provided it survives cleaning and reads without pulling the magazine out of your pocket. Our magazine checks then cover what to look at on the one that stands out.
The rate per thousand, and why the denominator decides
A raw malfunction count does not compare from one session to the next. A rate does.
The calculation fits on one line: malfunctions ÷ rounds fired × 1,000. Three malfunctions over a thousand rounds make three per thousand.
The point is not the precision of the figure, it is comparability. The same number of malfunctions describes radically different situations depending on the volume fired.
Three malfunctions, four verdicts: why the denominator decides
Rate per thousand rounds obtained with the same number of malfunctions — three — depending on volume fired; an arithmetic demonstration, not a field record
This chart is not a record but an arithmetic demonstration: the number of malfunctions is held constant and only the volume fired changes. It illustrates why a rate quoted without its denominator says nothing, and why comparing two magazines means having put comparable volumes through them.
Voir les données
| Volume fired | Malfunctions | Rate per thousand |
|---|---|---|
| 200 rounds | 3 | 15 |
| 500 rounds | 3 | 6 |
| 1,000 rounds | 3 | 3 |
| 3,000 rounds | 3 | 1 |
Three malfunctions is a rate of fifteen per thousand over two hundred rounds, and one per thousand over three thousand. The first pistol deserves a closer look, the second is ordinary. The malfunction count is identical.
A practical consequence for comparing magazines: you need to have put comparable volumes through them. A magazine you only bring out at the end of a session, when the pistol is hot and dirty, will accumulate malfunctions for a reason that is not its own. Rotate magazines in the same order, or log the order.
The threshold is not a number
This is the point where many pages invent a figure. We will not give one, because none would be honest.
What signals a problem is not a rate crossed, it is an attributable repetition: the same variable involved across several separate sessions. An isolated malfunction proves nothing, even on a new pistol. Three malfunctions on magazine no. 4, spread over three sessions, while magazines 1 to 3 produce none, prove a great deal — whatever the pistol's overall rate.
The process of elimination follows from that order:
| What you observe across several sessions | Reasonable conclusion |
|---|---|
| The malfunctions follow a numbered magazine | the magazine |
| They appear with one lot, vanish with the next | the ammunition lot |
| They follow the pistol whatever the magazine and lot | the pistol |
| They space out without your having changed anything | a break-in under way |
That last line is the one most often misread.
Break-in: what manufacturers actually document
Malfunctions on a new pistol are not all defects, and two manufacturers put it in writing.
CZ is the most explicit. Its manual warns that on the Shadow 2 Target, "the difficult cocking of the slide during the first approximately 2,000 rounds" can come with failures to lock and failures to eject, and that the remedy is thorough lubrication of the slide guide1. It adds a warning worth reading: "In no case the recoil spring shall be replaced with a weaker one"1.
Glock, for its part, notes that the firing pin on a brand new pistol "may not move freely until several hundred rounds have been fired"2.
The trap with a break-in is fixing it. A shooter who fits a weaker spring to stop break-in malfunctions removes the symptom and installs a lasting problem — precisely what CZ forbids in its manual. Before modifying anything on a new pistol, you need to know whether the rate is falling on its own. That means having measured it.
What separates a break-in from a defect is therefore the slope, not the level. A break-in sees its rate fall session after session. A defect sees it hold steady. Both can start from the same figure at the first reading.
When the conclusion is the pistol
It comes last, after ruling out the magazine, the lot and the break-in. And it splits again in two.
If malfunctions appear on a pistol that produced none, look first at what changed: a lot, a new magazine, a skipped cleaning, a spring at the end of its life. Our wearing parts are judged on that kind of sign rather than on a counter, precisely because almost no manufacturer publishes an interval in rounds.
If the malfunctions have been there from the start and the rate does not fall after the break-in the manufacturer announces, the subject belongs to a gunsmith — and you will arrive at the bench with dates, lots and magazine numbers rather than an impression. That is the difference between "it jams a lot" and "four per thousand over twelve hundred rounds, across all magazines and two lots".
What to log, at minimum
The useful record is short. One line per session is enough, provided it carries the three variables and the volume.
| Column | Example |
|---|---|
| Date and pistol | 12/03, pistol A |
| Rounds fired | 150 |
| Ammunition lot | 124 gr, lot 7 |
| Magazines used | 1, 2, 4 |
| Malfunctions, with the magazine in the gun | 1 failure to eject, magazine 4 |
That is exactly what a shooting log records with no extra effort, since the session is already written down — part of why we built one. If you also chart what each ammunition weight changes, our guide to 9mm bullet weights covers what to expect before blaming the pistol.
The essentials
Three columns beat any opinion
- No cause stands out on an isolated malfunction: the series is what decides.
- Numbering your magazines is the highest-return step, and the cheapest.
- A rate without its denominator says nothing: express it per thousand rounds.
- A break-in is recognized by its slope, not its level — and is not to be corrected.
Sources
- CZ — Instruction manual CZ Shadow (PDF, 02/2025 edition), malfunctions chapter, consulted 7 August 2026: "Due to the special features of CZ Shadow 2 Target, the difficult cocking of the slide during the first approximately 2,000 rounds can be accompanied by occurrence of the 'Slide failure to lock' and 'Failure to eject' malfunction, which should be prevented by thorough lubricating of the slide guide. In no case the recoil spring shall be replaced with a weaker one!" ⚠️ This passage concerns one specific model, the Shadow 2 Target; it is cited as a documented example of break-in, not as a general rule.
- Glock — Preventive Maintenance of the GLOCK Semi-Automatic Safe Action Pistol (PDF, 7 pages), firing pin channel inspection step, consulted 7 August 2026: "Firing pins on brand new pistols may not move freely until several hundred rounds have been fired." ⚠️ This booklet is made of page images: its text is not machine-extractable and was read page by page.
- Oplo — method note. The rate per thousand rounds is a plain division, malfunctions ÷ rounds × 1,000, and the chart in this article is an arithmetic demonstration at a constant malfunction count, not a field record. No numerical threshold is offered: we know of none that can be sourced, and attributable repetition serves far better than an invented number.
This method attributes, it does not repair. It tells you which part to suspect before acting, which saves you changing a spring when a magazine was at fault. It replaces neither your pistol's manual nor a gunsmith's judgement. If a pistol behaves unusually — noise, recoil, a suspected obstruction — stop shooting immediately and have it examined, without waiting to build a series.
Frequently asked questions
How do I tell whether a stoppage comes from the pistol, the ammunition or the magazine?
By logging all three at each malfunction, then comparing. None of those three causes separates from the others on an isolated malfunction: they produce the same symptoms. What separates them is repetition. If the malfunctions follow a numbered magazine while the others produce none, the cause is the magazine. If they appear with one ammunition lot and vanish with the next, it is the lot. If they persist whatever the magazine and the lot, only then is the pistol at fault.
How is a malfunction rate calculated?
Number of malfunctions divided by rounds fired, multiplied by a thousand. Three malfunctions over a thousand rounds make a rate of three per thousand. The point of expressing it per thousand rounds is to make two sessions comparable, which raw numbers cannot: three malfunctions over two hundred rounds and three over three thousand describe very different pistols, even though the count is identical.
How many malfunctions should start worrying me?
The question has no honest numerical answer, and you should be wary when you read one. What signals a problem is not a threshold but an attributable repetition: the same magazine, or the same lot, involved across several separate sessions. An isolated malfunction, even on a new pistol, proves nothing. Three malfunctions on the same numbered magazine, across three sessions, prove a great deal.
Are malfunctions on a new pistol normal?
Some are, and manufacturers document them. CZ writes that on the Shadow 2 Target, stiff cocking of the slide during the first two thousand rounds can come with failures to lock and failures to eject, and that the remedy is thorough lubrication of the slide rail — emphatically not a weaker recoil spring. Glock notes for its part that the firing pin on a new pistol may not move freely before several hundred rounds. What separates a break-in from a defect is the slope: a break-in sees its rate fall, a defect sees it hold steady.
Should I number my magazines?
It is the highest-return step in the whole method, and the cheapest. Without a number, a faulty magazine is invisible: it blends into the pile and its malfunctions get charged to the pistol. With a number, it gives itself away in two or three sessions. A punch mark on the base plate, or an engraving, is enough — what matters is that it survives cleaning and reads without pulling the magazine out of your pocket.