[pinkythepig]

He was prone to 'rattling the bars' so to speak and out of boredom, learned how to break the 'secure' file cabinets/safes that all the top secret stuff was kept in.

Basically, all of the combination locks had 2 digits of leeway. So if your first code was 20, you could set the lock to anything from 18 - 22 and it would still open.

This meant that instead of the locks having 100 x 100 x 100 = 1,000,000 combinations, they instead had something closer to 20 x 20 x 20 = 8000 combinations. He would then brute force the 8000 combinations through a combination of trying the more likely ones first and if that failed, just brute forcing it over a few hours.

The locks were made easier to brute force in that unlike modern locks, the 3 combinations would be input on 3 separate dials so you can set the first two to the 'right' value, then just spin the third while attempting the handle at the same time. This meant you could try literally every possible combination in ~10 hours of work.

As for an example of easy ones would be a birthday or anniversary.

- code 1 = day 1-30(6 combos)

- code 2 = month 1-12(3 combos)

- code 3 = year 1-99 (20 combos) (This would likely be 9-10 combos if he accounted for the distribution of likely dates)

So to try every date combo, you would only need to try 360 attempts in the worst case.

[VLM]

Over the last 20-30 years there have been innumerable microcontroller projects involving a stepper motor and a servo and very recently some 3d printed parts to open combo locks. Its an interesting real world project.

At my high school some decades ago we had the 3-turn 0-39 master locks but mfgr sloppiness meant there were really only ten or so possibilities and if you knew the last digit you only had perhaps 100 or so combos to try which doesn't take long.

A frenemy of mine got into a practical joke war and my friends collected tens of thousands of magazine subscription cards from the school library over the course of weeks and filled his locker with them when I finally brute forced the lock. He responded by filling my locker with many thousands of 4-40 sided machine tool nuts and some washers fed in thru the top ventilation slots. Well, it all seemed like a good idea at the time.

Small time barely noteworthy events happened all the time like remove the lock and attach it to the locker upside down, or replace it with a different lock, or swap it with a neighbors lock, or remove and/or change the numbers on the locker door. Oh another move was breaking into a locker, and respectfully not touching any personal property but disassembling the interior of the locker such that the victim no longer had a coat hanging hook.

[digler999]

I would imagine if you kept notice of what number series was left on the locked safe, it would leak a little bit of info. for instance, if the last number of the combo is 42, but when its locked you consistently see the number 10, its likely the person spins the dial a predictable number of times each time he locks it. all you'd need is to observe him -from a distance- locking it and count the number of spins and you'd have a good guess what the last number is.

then there's the paddle technique, not sure what its called, but after the first number is set, the 2nd number has to be at least past the point where you feel the "paddle" contact the 2nd wheel in the series. Higher-end safes counter this somehow.

[btilly]

If I remember correctly, when the lock was open it wasn't hard for him to spin the dial and look for a click. That told him the last number. So he just collected that idly for every lock he could and now his safe cracking was under an hour of work.