Toool competition – we have a winner!

July 30th, 2025

The Toool lockpicking competition is held yearly from 2004 onwards. About 25 locks are selected and it is up to the Toool members to try and pick them at Toool gatherings. They can spend as much time as they like, and can retry to open a lock as often as they like. Scores are given out depending on opening speed, 10 points for the fastest person up to 1 point for the 10th time on the lock. Since some locks can be opened with just a single rake, times under 5 seconds all count as the same time.

Every year, locks are selected on what we think are good locks, meaning: they cannot be opened in a single rake, but can be opened within the year. This year, there were 26 locks and the mix looked like this:

The locks from the 20th Toool NL competition. More details and pictures are found on: https://toool.nl/competitie2025/

Having won 18 of the 20 competitions so far, I was keen to win again, but everybody is getting better! I am a very fast picker, but it is not just speed, it is also making sure you open all the locks. In the end, there were a few locks that were only opened by 5 or less people.

The Yardeni and Sargent looked simple, but turned out to be hard. The Abus XP1 was left unpicked for a long time, mainly because it has a reputation of being very hard to pick. When it turned out to be (quickly!) pickable, it got 5 openings.

Then to the hardest locks. The EVVA 3KS was my personal nemesis. There is only so much time you can practice on these locks (only during gatherings). For the first time, I resorted to practicing at home. I took an EVVA 3KS I had in my collection, created specialised picking tools and practiced until I could open it. However, the 3KS in the competition would not budge. The main problem being that, when moving some sliders, the lock would reset (which my own 3KS would not do). I’ve spent quite a few hours on this 3KS to no avail. It was opened by Henri who became third, and by Tom who won this year’s competition by 1.5 point. Congratulations Tom! The three of us opened the same locks (except the 3KS for me) but I scored higher than Henri because of my fast openings on the locks I did open.

Since a 3KS is easy to disassemble without having the key, I took a look at the insides, to see if there was anything that can explain the resetting of the lock, but everything in there seemed very normal.

These are the two locks that no-one opened. The TrioVing does give feedback while picking, but nobody succeeded in opening it. The DOM was picked into a false set but also never opened. The pins have a shape that makes picking quite hard.

I am looking forward to the next competition. This one was exciting right until the end. The last gathering I decided to not work on the 3KS but try and gather a few points by improving times on simpler locks. At the start of the gathering, I was only half a point in the lead, and was able to gain another point by improving my time on the Benco lock. The Benco I don’t like, as I don’t really get a good feel of how to pick it. Tom was trying to open a new lock (such as the DOM above) for maximum points, but then changed strategies and improved his time on the Desmo, giving him a 1.5 point lead. I only realise now, writing this, that improving my time on the Benco with another 2 seconds would again have been enough for the win. That is how close this years’ competition was!

If you have locks you are willing to part with, please consider donating them to Toool for the next Toool NL competition. We are particularly looking for the difficult locks. Equivalent to Desmo, 3KS, Trioving, and IX6. As we are building two cases, one for us and one to share with other Toool organizations, two locks would be ideal.

Walter.

https://toool.nl/competitie2025/

Bookleg: Locksport

July 19th, 2025

Dear hacker friends, it has been brought to our attention that physical bootleg copies of Locksport are in circulation. While imitation is the sincerest form of flattery, bootleg copies are a near 1:1 copy without any additions or changes. This is uninspiring freeloading.

At the moment of writing, two official versions of Locksport by No Starch Press are in circulation, the Defcon version which doubles as the uncorrected proof with 367 pages, and the release version with 383 pages. Several pre-print versions of selected chapters were shared as PDF for preordered books directly from No Starch Press. Similarly, a DRM-free copy is included with the purchase of physical copies.

After thorough investigation — by placing the copies next to the original work and comparing the differences — we have determined several key differences. Please take your copy from the shelf and compare it to the pictures below.

The book on the left of the image is the version released by No Starch Press. This book measures 177 mm in width and 235 mm in height. 7″ by 9.25″. The bootleg copy measures a centimeter (50/127″) smaller in the width. This resulted in the image depicting the authors with their favorite tools seemingly squished, and two more black lines on the bottom are added, The No Starch Press logo is slightly more oval as well. You may also note the slightly more orange tint of yellow used in the bootleg copy.

The book on the bottom of the image is the version released by No Starch Press. The differences are quite clear, where the original has the full name written out, the bootleg has the title and subtitle on the edge. The names on the No Starch Press version are the last names of the authors, and the bootleg version has the full names as shown on the cover.

The last difference we want to share is the glaring lack of color in the pages of the bootleg copy. The authors of Locksport and No Starch Press take pride in the full color images in the original book. This makes this book a bit more expensive, but the results are totally worth it.

Thanks, Jos, for lending me the bootleg copy of the book and the pictures for this blog. I’m glad you can laugh about it.

All No Starch Press copies have several key indicators to make counterfeit books practically impossible. Before accepting a Locksport book, please check if the book is legitimate. Furthermore, if you ask the authors to sign your counterfeit book, it’ll be ceremoniously thrown through the room, and you are expected to be uninvited to all Locksport authors birthday parties.

We will keep monitoring the situation and keep the hacker community up to date when new discoveries are made. Thanks for your continued support!

CCBY4.0 Jan-Willem Markus. Toool Blackbag

Miniature Roman golden lock measures just 11 by 12 mm

February 26th, 2025

On November 14th 2024, the Landschaftsverband Westfalen-Lippe (LWL) at Münster presented a tiny golden lock from around the year 400 (1600 years ago). This news was published on the 28th of January by the LWL, and widely shared. Most articles written on the topic used the smallest version of the photos they could find, and glanced over the details. We chose to write this article after re-discovering the original article and studying the inner workings of the lock.
Feel free to check out the original article, first. The article has loads more details on the history, as well as videos of the CT scan and the working principle of the 4x model: https://www.lwl.org/pressemitteilungen/nr_mitteilung.php?urlID=60628

This is the lock itself on a black background. The Lock is 11 mm by 12mm in size. DKZ 3619,0094:109 – Fundnr. 15. Photo copyright: LWL /S. Brentführer

While it doesn’t look too special at first glance, this lock is s quite an engineering marvel. At just 11 mm high with a diameter of 12 mm, it’s tiny. The engineering is quite similar to engineering in horology and takes incredible skill to make something like this today, let alone with the tools from 1600 years ago.

The lock looks great from the outside, but to really, appreciate the complexity. We will have to look at the Neutron tomography CT-scan the researchers created. The images may not look like much to the untrained eye, but the researchers published the post-processed images to show the specific parts.

CT-scan (Neutron tomography) of the lock revealed the inner workings of the lock.
Photo copyright: Paul-Scherrer-Institut/Villigen [CH]/ David Mannes

The next image is the processed CT-scan where specific parts are colored. The article gives the following explanation: frame with spring (red), bolt (blue), broken bolt (?) (yellow), piece of the key (green), baseplate (purple) and inserted chain link (orange). The original article provides a video of the CT-scan as well. The video is in the attachments to the original article, with the name: “Anlage 1: Dosenschloss_Neutronen_CT_Animation.avi”.

This image shows four views of the processed CT-scan. Three side views and one top view.
Photo copyright: Paul-Scherrer-Institut/Villigen (Schweiz)/David Mannes; Montage: LWL/Corinna Hildebrand

With the information from the CT-scan, Stefan Brentführer set out to make a functioning replica. He chose to make the replica larger, likely to better show how the lock works, as well as it is easier to work on larger parts. This exploded view is quite well done, but it doesn’t show clearly how the parts fit together. If you are up to the task, take a moment to puzzle to see how these pieces fit together.

Replica of the lock at a 4:1 scale. DKZ 3619,0094:109 – Fundnr. 15.
Photo copyright: LWL /Stefan Brentführer

While locks which require a rotational motion to open can only be opened in either clockwise or counter-clockwise rotation, this lock operates clockwise, as do most locks in use today. Does someone know examples of counter-clockwise operating locks from 1600 years ago?

I’ve puzzled that the steel piece in the middle fits upright and horizontal on the round backplate. It has a matching piece which fits the rectangle hole. This leaves a square opening for the square linkage. The second diagonal rectangle hole is for the round post. (Yes, this tiny lock has a pipe key!) The cutout on the key corresponds to the bridge on the middle plate. From there it got confusing, and where I got some help. The bolt fits under the bridge into a barely visible rectangle hole to the top left next to the spring.

Single frame from the video demonstrating the workings of the lock.
Photo copyright: LWL /Stefan Brentführer

The lock’s operating is rather clever as it reuses the spring to keep the bolt in place in either positions. With the bolt closed and the key rotating in the clockwise direction, the fat bit of the key pushing down the spring that traps the bolt in the locked position. The triangle tip of the key perfectly fits the triangle cutout on the bolt. The bolt is moved on top of the spring, while the key holds the spring down. The closing operation is identical, but in reverse. Stefan Brentführer made a video of the operation and added it to the original article under “Anlage 2: Dosenschloss_Rekonstruktion_Schliessvorgang.MP4”.

Toool didn’t have a role in the research, but we believe these moments should be chearished and the news shared widely. My contribution, beside this blog, was the work to demonstrate how small the lock really is. I’ve made a mock-up of the lock in OpenSCAD and printed a copy with an FFF 3D printer. The files are attached, in case you are curious or want to print your own.

3D printed mock-up of the lock in purple filament on top of a cabinet lock with very much the same mechanism but designed as a mass-produced low security product. CCBY4.0 Jan-Willem Markus. Toool Blackbag.

Nigel from Toool UK went a different route, and chose to model the mechanism in a 3D CAD program, with great result!

A 3D render of Nigels model. Photo copyright Nigel K. Tolley. Toool UK.

The photos in this post are mostly from LWL /S. Brentführer. They kindly give permission to share the pictures in the context of the finding. The article is the creative work of Jan-Willem Markus. This text is copyright CCBY4.0 Toool Blackbag.

January Challenge

February 2nd, 2025

Jan-Willem wrote about the UKLS January Challenge last year. This year, Thice decided to participate, as a way to learn how to pick more types of locks. I myself started a few days too late to make it a real January Challenge. Nkt sometimes skipped a day, but then made up for it the next. But that is not a problem, the great thing about doing such a challenge is that you get to work on different types of locks, even locks you haven’t picked before. It is a nice boost when at the end of the challenge, you’ve learned some new skills!

These were the ones I picked:

Thice picked these:

Nkt not just picked pin tumbler locks, but quite a few lever locks as well.

Nkt kept the best for last, however!

Join next year if you can, or just start now! Just pick a month, or a number of days, and see where it will get you.

Walter.

Impressioning a key for an Ingersoll 6 padlock

January 10th, 2025

This keyless padlock was sent to me by a friend as an impressioning challenge a while ago. At LockCon 2024 I was able to get some fitting blanks from a friend, so I could finally get to work. Below you can read about my experience with impressioning a key for this lock.

First two attempts

Starting out I approached the impressioning like I would with a pin tumbler cylinder, by preparing the blank. Since the blanks were steel, I didn’t need to remove any plating, but the surface was very rough so I made it smooth (a mistake, but more about that later).

Tried to figure out the spacing, but it was difficult to see so I tried to get some marks. The first mark was not difficult to find, because the levers bite in the edge of the blank. Marks stayed on the edge, but switched sides from time to time. As cuts got deeper, marks started to get wider and cover more of the surface of the edge.

From time to time marks reappeared on levers I worked on earlier, I think this might be caused by the shape of the levers. I remember from picking other locks with a similar mechanism (Miwa U9, PR) that before getting them open I spent some time tapping multiple levers to set them again. So I took the reappearing marks as a clue that I was getting closer to opening.

I decided to stop my first attempt after working on several levers like this for quite some time without feeling I was getting closer to opening the lock, so I decided I must have filed too deep somewhere earlier. When I tried to figure out the spacing earlier, I removed some material on the first cut on the left and I suspected that might have been too much.

I started the second attempt by preparing the blank again, but didn’t try to find the spacing. Instead I use the key from my first attempt as a reference to make some guiding marks on the side with a marker. Most of that second attempt went like the first, and when I decided to take a break I was looking at two near identical attempts.

Time to think

With only one blank left I decided to take some time to work on other projects and think about possible mistakes I had made in the first two tries.

A conversation with someone about the blanks used in the SSdeV impressioning competition at the Wendt Hausmesse earlier this year finally put me in the right direction. During that competition, in the final round we had to try to make keys for several different cylinder models. Obtaining blanks for some of those had turned to be difficult so on those we had to use universal blanks. One of the downsides we encountered was that sometimes preparing the blank like we were used to, left us with a blank already below the zero cut height. Since I already suspected there could be zero cuts on Ingersoll keys, could it be that I had been setting myself up for failure by preparing the blanks?

Third attempt

Feeling motivated again, I decided to clamp up that final blank. Not having prepared the blank, I took my time finding the first mark. It was a pretty clear “bite”, but before I started filing I wanted to make sure I didn’t miss any other obvious marks. After this the impressioning was very similar to the previous two attempts. I just had to remind myself to check if new levers started to make marks from time to time. And of course the biggest difference was the key finally turning. I did encounter an unexpected surprise though: the key turned freely but I was unable to remove it.

Open! But the key is stuck…

Troubleshooting

After some online research and asking around, I managed to remove the shackle from the padlock, so I was able to remove the core and find out what prevented the key from being removed. My initial thought was canyoning, even though I kept an eye out for that during the impressioning. The real cause was totally different, and not really something I could have prevented. The key wasn’t perfect, but had allowed the gates to align just enough for the sidebar to drop in under tension. After that though, It was stuck so tight that the sidebar springs were too weak to lift the sidebar out again. 

Obvious damage to the sidebar where it was forced in a gate. Also note the shallower traces caused by pushing the bar against the levers during the impressioning process

Conclusion

I think the only thing that might have prevented this from happening is having a list of all possible cut depths and making sure to file exactly to those depths on the exact right location of the blank. Should I ever try this again, obtaining a list of the cut depths and making some sort of jig or guide to help with the spacing would help out a lot.

Henri/Greenish

Dutch open lockpicking 2024 at LockCon

December 4th, 2024

The Dutch open lockpicking is one of the largest lockpicking competitions of its kind. In several rounds with ever more complicated locks the participants attempt to open the locks as quickly as possible to secure their place in the finals.

The qualifier consisted of six tables with ten participants at each table. In ten rounds of five minutes each, the participants decided who continued into the quarter-finals. The best three of each table, eighteen in total, continued. From there, more difficult locks were introduced. The competition went from eighteen to nine to three participants. The complete rule set can be viewed at https://blackbag.toool.nl/?p=4652

For the five minute locks we had a wide selection consisting of, Eras, Destil, Kibb, Yale, Mastermate, Dom, Basi, Nemef, and several others. For the fifteen minute rounds we introduced more difficult locks, including Winkhaus, CES 1SB, Gege ANS2, DOM Plura, DOM Sigma, Oxloc, and more.

While we cannot make guarantees, we strive for consistency.
You’ll very likely encounter the same or similar locks in the next competitions at LockCon.

The final score of the Dutch open lockpicking at LockCon 2024.
Dutch open lockpicking 2024 at LockCon. Henri (3rd), Nitiflor (1st), Walter (2nd)

Nitiflor won the competition by opening the most locks in the finals. He won the Multipick Elite Locknoob set with 39-pieces. Walter placed second and picked the Multipick Quick easy pro key duplication kit as his prize. Henri received a Sparrows Kirigami lockpick set as the third place winner. These prices were kindly sponsored by Multipick and Sparrows.

We are hosting a lockpicking competition like this one almost every year, and we hope to see you at the next one.

Scores of the half finals for inspiration

Dutch open impressioning 2024 at LockCon

December 4th, 2024

The Dutch open impressioning is traditionally the first of the large competitions at LockCon. The challenge in impressioning is to make a working key by reading the feedback from the lock and adjusting the key accordingly. Eventually the lock opens, and the opening time is registered. The most consistent and fastest impressioners get to attempt another six locks in the finals. This year, 40 people joined the competition. The complete rule set can be viewed at https://blackbag.toool.nl/?p=4652

For this years competition, two locks were to be opened in the qualifier. This was a little confusing for some, but the effect was clear. By opening more locks, we have effectively reduced the luck factor and rate the participants on their skill. All twelve finalists opened both the qualifier locks.

Dutch open impressioning 2024 at LockCon. A Finale: Torsten (3rd), Jos (1st), Lasse (2nd)

Jos won the competition by opening the most locks in the fastest time. He picked the Multipick Kronos electropick. Lasse won second place and received the Multipick Quick easy pro key duplication kit. Torsten placed third and chose a Mulipick cutaway training lock as his price. These prices were kindly sponsored by Multipick.

Dutch open impressioning 2024 at LockCon. B Finale: Rubberbanned (3rd), Cocolitos (2nd), Oli (1st)

The B finals in Impressioning was win by Oli, who received the new Multipick advanced Sandman lockpick set. The second prize in this final went to Cocolitos, who picked a Multipick triple gauge lockpick set. Rubberbanned received third place and won a Multipick cutaway training lock. These prices were kindly sponsored by Multipick.

We are looking forward to organizing the next impressioning competition at the next LockCon. In the meantime, SSDeV will hopefully organize an event as well.

Toool NL competition

December 4th, 2024

In the Toool NL competition, the Toool members compete by picking a selection of locks during Toool meetings. This year was the 19th Toool competition, with 45 meetups and at least a hundred hours to attempt these locks. However, only the quickest opening time for each competitor counted.

The points are assigned by opening times, where the quickest opening gets ten points, the second-quickest gets nine points, and so on. Sometimes points are shared when several people have the same opening time, or opened the lock within five seconds. (We count these as the same opening time, as we have a lockpicking competition and not a stopwatch operating competition.) The full breakdown of the competitions, locks, statistics, and the rules in Dutch can be found on. https://toool.nl/competitie/

The competition went strong, from the first meetup after LockCon already many competitive times were set. More locks were opened every meetup, and the opening times were significantly reduced. The locks which were predicted to be difficult were not, while difficult locks were underestimated. The competition remained close until the end. The analysis on lock difficulty is found at https://blackbag.toool.nl/?p=4684.

Screenshot of the competition score. https://toool.nl/competitie2024/

Walter won the competition and received a Multipick Ares Disc detainer lockpick. Henri, with second place, chose the new Multipick Elite dimple picks community edition. Tom, who place third, won the Sparrows Vorax set. These prices were kindly sponsored by Multipick and Sparrows.

The next competition has started. Do you want to join this one as well? Become a member, and join us at the Toool meetups in Amsterdam and Eindhoven. May the best picker win. https://toool.nl/competitie2025/

Toool NL competition 2024 – analysis

November 20th, 2024

The 19th Toool NL competition is concluded (https://toool.nl/competitie/). The Toool members worked on lockpicking a set of 27 locks over a period of one year, and registered the quickest opening times. The full competition details can be found on https://toool.nl/competitie2024/.

Before we started picking, Walter asked the Toool members to make a guess on the best opening times for each lock. With this data he expected to see which locks are perceived to be difficult, but were not, or vice versa. Four lockpickers made their guess, and we analyzed the data.

Please note, the final opening times are heavily skewed in favor of short opening times, as the locks are picked several times at the Toool meetups and only the lowest time counts. A lock may be opened in 15 minutes during the first attempt, but during several meetups the final opening time can be reduced to five seconds. Many of the locks in the competitions are perfectly suitable locks, and us opening a similar lock in five seconds, shouldn’t be the reason to replace yours.

We found this data isn’t easy to visualize in a single graph, as each lock is unique. We chose to break down the analysis results and report on the five most underestimated locks and the five most overestimated locks.

CylinderBrandFastestLockpicker 1Lockpicker 2Lockpicker 3Lockpicker 4
T1906Abus E6066.5074.00630.005.0020.00
T1907GTV36.435.00940.0030.0015.00
T1910EVVA39.00420.00304.0060.0015.00
T1915Mul-T-Lock Classic216.04360.00312.00120.0045.00
T1924Sargent27.64150.00925.005.005.00
Underestimated locks from the Toool NL competition 2024
CylinderBrandFastestLockpicker 1Lockpicker 2Lockpicker 3Lockpicker 4
T1913DOM iX 5KG47.41240.00985.00120.00200.00
T1916ISEO8.31180.00781.0060.0025.00
T1918ASSA 5005.00120.00610.00180.0045.00
T1919Best SFIC5.00240.00558.00180.0020.00
T1926Medeco Biaxial5.00132.00121.00600.00180.00
Overestimated locks from the Toool NL competition 2024

It is quite interesting that only a few locks were underestimated, while most locks were overestimated. Several of the underestimated locks have a tight keyway, but this in itself is not a guarantee for the lock to be difficult. For example, it’s common for the old Evva to be full of standard pins. The biggest surprice is the Sargent lock, which is much better quality than they are usually given credit for.

On the overestimated side, we have locks which are usually quite secure. The Medeco has good tolerances, and the SFIC rely heavily on their double shearline for security. While the Assa 500 and DOM IX are well-made locks from Europe. However, with the right tools and techniques, they just opened without too much difficulty. (The Medeco was raked several times.)

This analysis and report may not have solved the mystery of why some locks are easy, and other locks are difficult, but it was interesting to analyze these nonetheless. We will not take guesses for the 20th Toool competition, but do take a look. https://toool.nl/competitie2025/ We chose to save all the keys of this new competition, which enables us to see if the bitting is what makes the difference after the competition concludes.

The LockCon LockQuiz

November 1st, 2024

For this years’ LockCon, I decided to not do a lecture about a lock-related subject, but to organise a lock-themed pubquiz, the “LockQuiz”. I like the old-skool quizzen where you work together in teams and write down your answers on paper. Modern app-based quizzen have their own charm, but that was not what I was going for. When I play a quiz, I like questions where teamwork leads to the correct answer.

So I went about to see if I could come up with enough locksport-related questions and categories. I ended up with 68 questions in 6 rounds. In round 1, I showed lock cylinders and padlocks, with the brand name removed and the teams had to tell the brand. In round 2, lock-related books were shown, with one word (partially) blurred, the question was what the word was. Round 3 was another picture round with pictures of tools, where the question was which toolmaker the tool was from. After a small break, we continued with round 4, in which I asked for names of movies in which something lock(picking) related was shown. Round 5 showed keyways, asking for the brand and model, and round 6 was a music round, showing clips that had something to do with locks as well. Requested was the name of the artist.

I got very good feedback and everybody seemed to really enjoy the quiz. I have made generic pubquizzes before and know that the biggest pitfall is to make them too hard, which is no fun for the people doing the quiz. That is why I included answer sheets for the (rather difficult) movie and music rounds, that showed the number of letters in the answer and also had some letters filled out in advance.

The quiz had a clear winner, which was good, as I had not thought of a tie-breaker question. In the end, 77% of the questions were answered correctly, which is wonderful, as that was what I sort of aimed for, but it was hard for me to know how difficult it would be in advance. The lock brands round turned out to be the easiest, the music round the hardest, but even that round had 74% correct answers, thanks to the answer sheets.

If anybody wants me to do the quiz elsewhere, contact me. Obviously, the people at LockCon already know the answers. Or, if you have suggestions for other topics for rounds, or questions, I’d be interested in those as well. And finally, if you’ve made pictures of me presenting, I’d love to have a copy as well.

And again, congratulations to Team Baguette, who won!

Walter.