Just what I have been looking for to help me understand how to go about effectively solving puzzles. I had been an all options proponent but this convinces me to switch.

I found this because I wanted to learn and use the Snyder notation. You don’t need to convince me of the other puzzle solving!

I've been an "all candidates" solver for many years, but have always wanted to develop quicker solving techniques. Your process intrigues me, but I can only use it to limited success, often reaching a point where I can't go any further. The all candidates method, while certainly much slower and often cluttered, seems to offer more avenues toward a solution. I don't question your obvious speed with this technique, I'd just appreciate further explanation regarding where to go when the two candidate progression stalls. I find myself staring at the grid longer than if I'd just completed filling in all the candidates and resorting to old practices.

Great compare/contrast of those two methods, thank you! I was going to say that the 'all the candidates' method might be easier for beginners, but watching this I definitely would go with Mr. Snyder's notation.

Very helpful comparison!

Most helpfull vid so far. - Many thanks.

This was very helpful. Thank you.

I have developed a technique called "shooting holes". For instance: From the initial position, you can place 3 in the middle box. This is called a "Naked 3", right? (Tryin' to pick up the lingo!) Then you can place a 1 in the top left box, another "Naked 1" which leads to a 1 in the top middle box. Now, looking at the 6th column with 1 and 3 filled in, my eyes sort of gets attracted to the fact that this column only lacks 3 numbers: 2,5 and 7. I instantly spot that 2 and 5 "shoots into the hole" at top right corner of the top middle box, which leads to a 7 as well as a pair, 2 and 5, at the two "holes" at the bottom middle box. Is this a "technique" and do you pro's have a name for it? I have only recently stumbled into this channel which REALLY has opened my eyes to improve my Sudoku-solving: Thanx a lot for the "X-wing" presented in various videos! Looking forward to learn how to apply it! (Edit: 12:50 you spot the same thing! Looking for rows, columns and boxes with a lot of numbers already filled in, deduct which is missing etc. I call it "shooting holes"! ;O)

I have been struggling as to which penciling method is best Thanks for simplifying things for me Excellent explanations / comparisons. Thanks again!.

Ipersonally use Snyder notation as far as I can and after it has driven it's course I revert to populate all possibilities. But I don't do it like a computer would, I'd still take note of my snyder notations as they give definite answer where numbers can be. It isn't unusaul that while populating all possibilities I stuble on a naked singular. And if that happens I go back to Snyder and see what I will get with the new number. And I try to do it in numerical order but sometimes some things just looks right to do. For exmaple in this sudoku I first put down all the 1's noticed that there nothing to do with 2's and concentrated on 3s. The minute I've put 3 in r4c6 I noticed that this allows me to pencil mark the 2's in the two middle bottom squares and I also noticed that 5 cannot go to r1c6 and thus there's 2-5 pair down. I stopped what I was doing and scanned what was the missing number and noted the 7 in r1c6. Then I continued scanning what I could do with numbers 4, 5, 6, 7, 8 and 9.

You have persuaded me to change my notation system but it's tough: today I made an error in solving for the first time in quite a while. But I'm confident that the change will be worth it in the end. Thanks!

I was rather farther along, but stuck, when you said, “and that cracks it”. WHAT?! Oh, well, actually it does! Learned something today.

Simon, you just miss to spot the singleton 4 in row 4 column 1 you already find before you find the x-wing pattern. If you do that, you crack the puzzle very fast indeed.

Is Snyder notation helpful ? Initially I thought not. The added clutter seemed to inhibit the thought process. But avoiding notation is a BIG mistake. Too much time is wasted by overlooking simple structures. Holding too much in head is fatiguing. One sits much of the time staring at the board, waiting for inspiration to hit. Even worse, the solver might resort prematurely to unnecessarily complicated techniques, increasing the likelihood of errors. This video is a treasure. Mr Anthony exhibits fluency and ease in demonstrating the Snyder notation. Solving puzzles this way is more enjoyable and, at times, even artful. And one moves more quickly and reliably to the real point of each puzzle: namely, identifying the main principles and structures which crack it. Using Snyder notation is a skill. I went slowly and easily at first until it became a fluid, unconscious habit. Even though I am primarily interested in extremely difficult puzzles that require techniques from mathematical graph theory, I now appreciate that a mastery of Snyder notation is a skill basic and useful to the art of Sudoku. Many thanks to Simon and Mark for these brilliant tutorials.

I tend to look at the whole puzzle and see if there are any obvious answers. Here, the first two ones jumped out at me and then the remaining ones.The two sevens in the middle were next and then I would fill in all the rest of the pairs. I think triplets etc are best left out till at the end as it all looks a mess...i.e. only what a computer would need for a brute force attack in hacking parlance. So, I agree with you, Simon, not to build triplet walls when there are singletons and pairs for the taking.

Thanks again Simon, I really appreciate the explanation of reasoning for adopting this technique, or in my case something very similar. I have only been doing Sudoku for about 2 years on and off and independently came upon the idea of using pairs as an opening strategy about 6 months ago after struggling with the visual clutter of the "all candidates" approach. However the strategy I have developed includes pairs that can exist in only 2 cells in columns and rows as well as blocks. Often they will coincide but just occasionally I find the extra pairs of assistance, Of course it does mean there will be ragged singles in blocks resulting from row or column pairs which might be mistaken for naked singles but usually I find no confusion: perhaps because I tend to think and focus on columns and rows when scanning rather than blocks. I'm painfully slow but the real problem arises when there seem to be no further pairs to be found.

So what if you have a box where you have 2 possibilities for each number but there’s an overlap where one of the squares has 3 possible solutions? Throughout this video each square only had 2 possibilities.

very helpful Simon. how do you solve it in under 5 min or 3 min like Tom Collyer?

What a difference in presentation style compared to the CTC videos nowadays...

I'm of the belief that Snyder notation is superior to that of filling in all candidates. But, when you said you couldn't see that 2s were limited in box 4 (paraphrase) at 20 minutes, I was looking at it and was surprised you couldn't see how they were limited. There was clearly a 24 pair in that box. Thus forcing 2s in to C2 in box 7.

I'm at 12:11 and very confused why this video is titled "What is the Snyder Notation", since you don't actually seem to be explaining nor really using it. In the beginnging you briefly mention the fact that you can pencilmark pairs inside a box, but you never explain why one would do that or how this is useful. The following part where you used the all candidtes method was confusing since this has nothing to do with Snyder. I was drawn to this video because I wanted to learn the rules of the Snyder Notation and I'm leaving without having learned much sadly.

What do you think if you use software that shows all candidates, but can highlight each number of a candidate, like all threes for example?!? Then it is obvious again, that there is a X-Wing in C7 and C9, R5 R9. Would this be considered cheating?

Ha! I came up with this method on my own. I've never even looked at a video about the game...

Took me 3:44 after importing and generating the candidate list. No advanced techniques required...pretty simple for a “super fiendish” puzzle.

Solved on 12/10/22 in 17:26

The notation system is very good, but with this puzzle, I would start by easily filling out the remaining 1s and some 9s.

I have an 8yr old …I had her play the 4x4 and by her 6th game she was solving them in Under 1 minute … is this normal ? … should I immediately move her to the classic soduku ?

I ruined myself through years of using an app that had auto-fill pencil marks and highlighting candidate numbers. It became a game of pattern recognition, barely even looking at the numbers themselves, where I'd finish a puzzle rated 'annoying' in around a minute. Doing it this long way now takes forever, and the way the puzzles are generated on the app, means that it's very hard to solve anything above 'difficult' with just snyder notation, because they deliberately require certain logic steps that don't lend themselves well to the technique. Makes if feel like this kind of notation is a massive step backwards, but actually I've just been cheating the whole time...

You would spot the 3 and the 5, but not because of the "all candidates" notation. You spot them despite the "all candidates" notation, The "all candidates" notation is noise in these cases.You spot the naked singles because of information elsewhere.

The "Snyder" notation has been around for a while. If you look at youtube sudoku amateurs you see it a lot. What Snyder doesn't do is combine it with noting the spaces where only two numbers can go. You need an extra symbol, like a "/" or a way to use the geography of the space, so you won't confuse the "twoshare" notations with the "pairs in squares" notations, but it can make things a lot faster.

My time was 20:36

Fiendish? Solved in 4 minutes. No big tricks necessary.

I don't make notations. It's messy and never helps much.