1. Mark what?

• Marking a bullet itself is pointless. When a bullet hits a target, the soft lead quickly deforms significantly enough to obliterate any markings on it.
• People who advocate this marking probably are referring to the shell of the round. (I'll mention only here that people who don't know what they're talking about (i.e. bullet vs. round vs. shell) probably shouldn't be talking about it.) Let's correct the issue and discuss marking the shell.

• Shells are freely ejected by semiautomatic pistols onto a nearby surface (usually the ground).
• Criminals usually don't clean up spent shells, leaving them for police to find.
• Not all shootings are done with pistols. Revolvers are still very popular, and a revolver retains the shells when fired.
• Assume roughly 1/3 of all handguns are revolvers (no statistics available on which are used in criminal shootings), so shells are not left behind in 1/3 of cases.

• Tens of billions of rounds of ammo are produced each year.
• 10,000,000,000 is an 11-digit number.
• Ammo lasts a long time, so let's add a couple digits to account for >10-year storage.
• That's 13 digits that the round must be marked with.

• Printing a 13-digit number in ink on a shell, esp. a .22, is hard. (Remember we're trying to print on ~10 billion shells a year.)
• If ink is used, it is easily rubbed/sanded off by a criminal.
• Printing 13 digits legibly on a small _round_ shell is hard.
• Stamping the number would be unremoveable, but much harder.
• Stamping the number also is a severe saftey hazard, as it unreliably weakens the casing which is designed to contain a near-explosion.

• A lot of bullets are currently made from one lead ingot.
• Trying to evenly diffuse taggants thru an ammo-intended ingot adds excessive cost and complexity.
• Tagging an ingot (similar to tagging a batch of poweder) gives only a marginal aid in tracing, as it gets distributed among hundreds (thousands?) of stores.
• Tagging individual bullets requires proper mixing of tags into lead in a rather small object.
• You'd still have to get 32-36 bits (11-13 digits) of data onto a single tag for a single bullet.
• Many bullet types are precision-built objects, requiring multiple layers of precisely-formed metals. Taggants would complicate the precision build.

• Assume 10,000,000,000 rounds made per year.
• Assume 10,000 shootings (killings) where tracking the round is worthwhile.
• Assume 1 in 10 (possibly 1 in 100) shootings would benifit from round tracking. Standard police work usually does just fine without it.
• Remember, roughly 1/3 of handguns are revolvers which do not leave shells behind.
• That means for each bullet tracked, roughly FIFTEEN MILLION (start counting!) rounds had to be recorded.

• Consider: the majority of violent felons are repeat offenders. The charges are often plea-bargained down, sentences are often less than maximum possible, and convicts are typically released early serving notably less than 80% of their time. Those released usually go back to what they were doing - assaulting civilians.
• Catch a criminal, throw him in jail for a long time. Does it work? sure - he can't kill a civilian if he's behind bars.