How do y'all complete puzzles so fast?

[SheenaValentina]

I don't understand how people are completing the puzzles so fast. Like, on a challenging 4x7 puzzle how is 130 seconds even enough to read all the clues? It took me about 2000 secs to complete. Obv I'm not as advanced as some people (the average was 800) but I just don't see how it's possible. Do you have any insight and/or tips besides practice?

[boxeeboxeebox]

I occasionally get sub-200 times if I get lucky with a) an easy puzzle and b) it's solvable using the top row alone (almost every clue is "A is more/less than B").
By filling in all the "logical conclusions" in a grid, you end up spending a lot of time on housekeeping rather than solving. Of course, this practice can backfire in puzzles where those additional logical clues are necessary – you'd have to go back and re-evaluate each clue and fill in all the blank cells.
Oh, and using your browser's Find in Page function to highlight puzzle terms is a huge time-saver when you first start a puzzle. Particularly for puzzles where the numerical sequence has a unique term, like "votes".

[thecanuck]

I do my puzzles in groups so I don't keep going over things over and over again. First, the lists of multiple items so you can see them easily on the grid. Then I get all of the easy positives and negatives out of the way, then the weird ones with 3 or 4 comparing each other, so I'm left with all of the relationship ones by themselves. That way as much as possible is filled in before I tackle the remaining ones I've skipped that take the most time. Hope that helps.

BTW I usually get just within the green area using this method, and doing it fairly quickly.

[AuntieBeth]

I would say the same thing as Boxeeboxee, that when you just focus on the top row you can sometimes go lightning fast. Not every puzzle can be solved this way at least with my brain in its current state. Haha

[AnonCuttlefish]

Like everyone else said, top row only. It also takes a bit of practice. When I don't play for a while my scores tank and it takes a bit to get them back up.

[uigrad]

My opinions:

The vast majority of players aren't going to be able to get "fastest" scores. Don't focus too much on them.

Even the median scores are very hard to get if you are new-ish. The people that play lots and lots of puzzles tend to be the people that also are really good. If they play the puzzle 50 times and you play it once, then they contribute to the median 50 times more than you. So the "median player" is not going to be able to get a median score in most cases.

I first found the site one month after it moved to the new site. At the time, the medians were much higher than they are now, and I was able to do better than them, but not by a lot. But I stuck around. Eventually I got to the point that I could compete with the best on the site, but it took some time. Since then, the best has gotten better, and even though I've continued improving over the years, I'm barely keeping up.

Start with the small, easy puzzles. The people that compete for times on the "challenging" puzzles or on the 4x7 puzzles are even better than I am. Don't advance to the next size until you are happy with how you are doing on the smaller size. Personally, I'm happy with my performance on 4x4 and 4x5, but every time I go up to 4x6, it's obvious I've met my limit. If you aren't in the top 100 of all time, and are trying to keep up with 4x7 puzzlers, you are probably in the wrong place.

If you follow this advice, the site can be fun, no matter what your skill level is. Having fun is far more important than scores. If you have enough fun, you may stick around long enough to improve. Only after practicing enough that you see significant personal improvement can you start thinking about trying to get a record time.

Note to the admin: The circuits puzzles show the top 10 solvers for each puzzle. I feel like that improvement would be very welcome on the logic puzzle site also. It gives 9 more spots for people to compete for, and I think it would make it more fun for everyone.

[superdoggy]

I don't know how people complete so quick but I don't really do 4x7 I just do 3x3 but I've done a few 4x7s and they are challenging just to sometimes read all the clues. If somebody could just say in this post how they complete puzzles so fast I think we would all like it. But as long as I just do 3x3 it's pretty easy to do. And if everybody here does that I'm sure somebody in this post will have a top time

[fromalabama]

With apologies to the original poster, the only real trick to being fast is practice, and lots of it. Most people likely get good at the actual logic part relatively quickly, but the record times are probably more about familiarity with the puzzle clues than the reasoning. Once you've done enough you barely need to read the clues to extract the relevant data and apply it, and if you've got a good enough memory you may not need to spend time clearing out used clues or searching the list for a specific one or filling in the lower boxes, all of which can save a few seconds. Plus you get better at the sequencing, figuring out which order to do the clues in to minimize the effort. Some puzzles can be figured out without even needing every clue if you hit them in the right sequence, so that's more time saved. On the rare occasions they introduce a batch of new puzzle scenarios it takes a little while for the speed demons to acclimate to the new names and intervals and whatnot, and everyone is on more equal footing for a little while.
For example, I've been on this site for about 6 years, and have spent approximately 7,842,000 seconds playing 60,000 puzzles (10k of each category). That's some 2178 hours, more than 90 days. That's roughly the amount of time a person with a 40 hour per week job would devote to it in a year, and I can tell you there are some people who spend considerably more time here than I do. You don't need to spend that much time or effort to get good, of course, but if you joined a class or a sports team and everyone else had been learning or practicing for a year already, you probably wouldn't expect to come out on top right away, at least not regularly.

I would second uigrad's motion to expand the top times to a ten-person leaderboard.

[LGramlich]

First I do the "the seven people (or whatever) are..." hint. Then I move to the shortest hints, particularly the conclusive ones (for example, Hilda doesn't drink merlot, or Billy is the accountant.) While looking for those, I scan for any that start with "neither," and process them, as well. This lets me move a lot of clues out of the way, so I can focus on the ones that are not so obvious. Finally, I go through the clues that are left from the bottom up. I don't know why it should be, but this seems to be more effective than going from the top down.
Whatever I'm still left with, I start analyzing more deeply what they can or can't be, based on the open spaces in the grid. That usually gets me to the solution.
I'm not a person who can get it in 120 seconds, mind you, but I can get to the 400-600 range, typically (and sometimes the 300s, if I'm lucky.) I've also been playing for decades.
I can also be faster depending on what hardware you play on. I'm sure someone with a touchscreen tablet is much faster than I, dragging a mouse around for clicks with a desktop computer.
Either way, good luck and have fun! I don't really care much about times or scores, in general. I post times in the comments only to see if I beat my own time (or not,) the next time a grid comes up.

[agentkitty99]

I don't understand how people are completing the puzzles so fast. Like, on a challenging 4x7 puzzle how is 130 seconds even enough to read all the clues? It took me about 2000 secs to complete. Obv I'm not as advanced as some people (the average was 800) but I just don't see how it's possible. Do you have any insight and/or tips besides practice?

IQ is the way to go

[kgf]

First, there's usually one clue that either starts with "the seven people..." or ends with "..are all different people" (or giraffes, or stars or whatever). Do that clue first as it is much easier to do with a blank grid and yields a lot of info. Then I just start at the top and go down the list, filling in the grid using all the rows. once you've gone through the list, go through it again. look for one thing that is in multiple clues and see if you can cross reference them to eliminate possibilities. The clues that contain "doublets" are especially useful for this. These are the ones that say something like "Of Garay and the one that's 110 ft. tall, one is in Peru and the other was discovered in 1956". If you find another clue that mentions any of those things, that's a good cross reference. I also like to look for rows or columns where only two (or three) possibilities remain (i.e., the little boxes are not filled in), and then go across or down to the other big boxes to see if there are any clues I can eliminate because the row I'm checking has those same little boxes filled in.

[redlead]

I almost exclusively play 4x7s, so I'm used to the style. I only work the top row. I always use the relative measure as my focus at first, and ignore any clues that don't include that relative measure. For example, if the puzzle is about the relative prices of submarines, that is the relative measure for me (price), and the depth / make / manufacturer categories don't matter. My order for dealing with clues that include the relative measure is as follows:

1. Absolutes, either positive or negative, as those can be marked immediately and cleared (e.g., The $30,000 submarine is sold by Zephyr Waves.)

2. Relative clues that are specific (e.g., The submarine with a maximum depth of 5,500 feet costs 3,000 dollars more than the craft built by MarineCo.), focusing first on specific clues that reference large values. In the 4x7 puzzles, the maximum difference that will be referenced in a clue is 5 units (in my submarines example, that would be $5,000). Where this occurs, you know that the lesser referenced item can only be the smallest or second smallest value, and the larger item can only be the largest or second largest value.

3. Relative clues that are non-specific (e.g., The submersible built by MarineCo costs somewhat less than the craft built by Ecomarine Ltd..). I use these to narrow down margins around the items. Keep in mind, I am still only using the top row. It's more like I'm lining up values than using a grid.

After those clues have been resolved, I start moving through the other clues in an order that is mostly intuitive. I have some tricks for how I handle those clues as well, but it all comes down to an unrelenting focus on that relative measure, in this case, submarine price. I also have a bit of a running mental catalogue of which items I've identified the relative value for, and if there is a clue related to those items, I will mark them and move on.

These are my last 5 puzzle times:

Puzzle #73933649	6 minutes ago	131 sec.	Completed
Puzzle #73933532	10 minutes ago	199 sec.	Completed
Puzzle #73933461	13 minutes ago	164 sec.	Completed
Puzzle #73933375	16 minutes ago	222 sec.	Completed
Puzzle #73933256	20 minutes ago	199 sec.	Completed

The most recent one I solved was 73933649, and its clues were the following:

1. The Ukani Q2 doesn't have a maximum depth of 2,500 feet.
2. The submarine with a maximum depth of 5,500 feet costs 3,000 dollars more than the craft built by MarineCo.
3. The craft with a maximum depth of 2,500 feet isn't sold by MarineCo.
4. The $30,000 submarine is sold by Zephyr Waves.
5. The $31,000 submarine isn't sold by Abyss Toys.
6. The Marine V costs 5,000 dollars more than the Rubico One.
7. The submarine with a maximum depth of 4,000 feet costs somewhat more than the submersible with a maximum depth of 8,500 feet.
8. The submersible built by MarineCo costs somewhat less than the craft built by Ecomarine Ltd..
9. The Rubico One doesn't have a maximum depth of 7,000 feet.
10. The $25,000 craft is sold by Watertime LLC.
11. The Abby EX-4 doesn't have a maximum depth of 14,500 feet.
12. The Ompoc Eleven costs $30,000.
13. The Ukani Q2 isn't sold by MarineCo.
14. Of the submersible built by MarineCo and the submarine with a maximum depth of 8,500 feet, one costs $28,000 and the other is the Steado G5.
15. The Rubico One doesn't have a maximum depth of 14,500 feet.
16. The craft built by Abyss Toys is either the Steado G5 or the submarine with a maximum depth of 5,500 feet.
17. The craft with a maximum depth of 2,500 feet isn't sold by Pan Systems.
18. Of the $27,000 submersible and the $26,000 craft, one is sold by Deep Sea Inc. and the other is the Rubico One.
19. The Rubico One costs 1,000 dollars more than the Cafader Six.

I immediately marked #4, #5, #10 and #12 and clicked the clues to remove them from my screen so that they don't clutter up the space. I then zoomed in on #6, as it used the relative measure and referenced a large value. As I scanned the clues, I had also noted #18, as it referenced two relative values (narrowing down the value of the Deep Sea Inc. and Rubico One submarines to two possibilities. When #6 and #18 are read together, it is only possible for the Rubico One to cost $26,000, and that also tells us the cost of the Deap Sea Inc. and the Marine V. I marked all of those items and cleared clues #6 and #18. Given that I had now identified the prices of several submarines, I did a quick re-scan of the clues for clues that mentioned any of the submarines whose prices I had identified, marked and cleared them. This allowed me to clear clues #9, #15 and #19. At this point, I was left with about half of the clues I had initially started with, and a lot of the board filled in. These are the clues I had left:

1. The Ukani Q2 doesn't have a maximum depth of 2,500 feet.
2. The submarine with a maximum depth of 5,500 feet costs 3,000 dollars more than the craft built by MarineCo.
3. The craft with a maximum depth of 2,500 feet isn't sold by MarineCo.
7. The submarine with a maximum depth of 4,000 feet costs somewhat more than the submersible with a maximum depth of 8,500 feet.
8. The submersible built by MarineCo costs somewhat less than the craft built by Ecomarine Ltd.
11. The Abby EX-4 doesn't have a maximum depth of 14,500 feet.
13. The Ukani Q2 isn't sold by MarineCo.
14. Of the submersible built by MarineCo and the submarine with a maximum depth of 8,500 feet, one costs $28,000 and the other is the Steado G5.
16. The craft built by Abyss Toys is either the Steado G5 or the submarine with a maximum depth of 5,500 feet.
17. The craft with a maximum depth of 2,500 feet isn't sold by Pan Systems.

Scanning them again, you can immediately see that the MarineCo is mentioned in half of the remaining clues. That caught my attention as well, so I quickly worked to narrow the possibilities for the MarineCo. It's not easy to picture how my board looked without the benefit of seeing the top row of the grid. but I can tell you that at this point, it was impossible for the MarineCo to be the Steado G5 (clue #14). Based on how the board was already filled in, clue #5 narrowed the MarineCo's potential price to $28,000 or $26,000, and the Steado G5 could not be either of those prices, as we already know the Rubico One was $26,000, and clue #14 tells us that the Steado G5 can't cost $28,000.

That puts the MarineCo at $28,000, which lets us solve and put away clues #2, #3, #8 and #13. Even though we've "solved" #14 as well, I don't remove it until I've identified where on the top row the Steado G5 goes, and marked its depth alongside it. For the time being, I just remember that I know the Steado G5's depth.

At this point in the puzzle, (again it's difficult to see without a visual aid) we also know from the process of elimination that the MarineCo is the Abby EX-4. That allows us to clear clue #11 as well. The only two submarines whose price we don't know are the Steado G5 and the Ukani Q2, which are $27,000 or $29,000. We know the Steado G5's depth is 8,500 feet. Clue #1 tells us that the Ukani Q2's depth is not 2,500 feet. That means that we can mark the 2,500 foot depth craft as being not $27,000 or $29,000, and clear clue #1 as well.

We are left with just these clues:

7. The submarine with a maximum depth of 4,000 feet costs somewhat more than the submersible with a maximum depth of 8,500 feet.
14. Of the submersible built by MarineCo and the submarine with a maximum depth of 8,500 feet, one costs $28,000 and the other is the Steado G5.
16. The craft built by Abyss Toys is either the Steado G5 or the submarine with a maximum depth of 5,500 feet.
17. The craft with a maximum depth of 2,500 feet isn't sold by Pan Systems.

Looking at clue #16 and our board, the only possibility is that the Abyss Toys is the Steado G5, and the only open price square that those two share is $29,000. We mark those two together, along with the 8,500-foot depth, and that allows us to clear clues #14 and #16, and obviously clue #7 as well. By process of elimination, we have already solved the price of the Pan Systems submarine, so we can clear clue #17 as our final clue.



At the end of the day, doing all of this quickly requires luck. You have to spot the right clues together, and if you miss an important one when scanning, your time will slow down. But I scan and rescan the clues constantly, always looking for things that I can solve immediately, either on the top row or in my memory.

I hope this helps with speed. Also, I've played a sickening amount on this website, so as others have said before me, it's all about practice.

[contrary]

I mostly live in the 4x5 Challenging world because I'm doing them in my breaks when I need a mental reset, but I'm pretty good at that size. Mostly everything's been covered already, but my summary is:

- Practice. Sorry, but that's ultimately the best tip.

- I always stay on the top row. Usually it's faster. Sometimes it's not. It's good mental practice even on the puzzles that aren't fast. When I first started out, I stayed on the top row for as long as I could before I had to start using the lower boxes, and over time I got better and used them less and less. I don't use them at all now.

- If you know the puzzles, you don't have to take time to figure out which section each part of a clue belongs to. For example, in the shark research puzzle, the "Names" are the first section, the "Boats" are the second second section, the "Sharks" are the third section, and the "Months" are the ordered-component section. Again, this comes from practice.

- Scan the clues for the keywords rather than reading the whole clue: For example, in the clue "Of Nettie's produce and the vegetables that sell for 2.60 per pound, one type is from Susanville and the other type is the beets". When I scan that clue, I see "Of", "Nettie and 2.60", "Susanville and beets". "Of" tells me what type of clue it is: two separate items and two other items that are associated with those two items but we don't know which way yet, "Nettie and 2.60" tell me the first set of items, and "Susanville and beets" tells me the second set.

- When I get to the first clue that has an ordered relationship, I glance at that section to confirm what the steps are between those items. In some puzzles, "2 months before" may mean that there's one item in between the items in the clue, and in some puzzles it may mean that those two items are right next to each other. That's the first time I glance at the puzzle grid.

- I personally pick an order of the clues and go sequentially from there. I'm faster if I go in order because I'm basically building a mental index of clues and what puzzle items they refer to. For example, If I start at the top and there's a clue that says "Bonita is partnered with Luke", I'll remember that Bonita and Luke are related by a clue close to the top, although I may not always remember what that relationship was specifically. Then if I get halfway down and there's a clue that forces Luke into the first two items, I have a mental note that says there's a clue related to Luke at the top of the clue list, and I can go back and pick up the fact that Bonita also needs to be one of the first two items.

- As I go, I build mental ladders of the relationships. If the vegetables from Turner sell for 20 cents less than the vegetables from Frenchville according to one clue, and carrots sell for 10 cents more than Frenchville, then I can mentally ignore those clues if I build a mental mega-clue that says (Turner - ? - Frenchville - Carrots).

- Trust yourself. If you want to go fast, you assume that you've done everything right up to the point. Occasionally that means you go to hit "Submit" and it tells you you have a problem. But in most cases, it means you don't have to re-check the logic you've already done, which is a huge timewaster.

- I get rid of clues as soon as they're guaranteed to be true no matter what the solution is. For example, if Billy's cost 10 cents more than Jaime's, the puzzle step is 10 cents, and I know that either Billy must be 2.60$ or 2.70$, then I also know that Jaime must be 2.50$ or 2.60$, and 2.60$ must be either Billy or Jaime. At that point, if I mark Billy as 2.60$ then Jaime's only other option is 2.50$, and if I mark Billy as 2.70$ then the only option for 2.60$ is Jaime. Either way, the 10 cents relationship is maintained, so I no longer need that clue and can mark it used.




One of my favorite things is seeing the record progression, as we all get better at the puzzles. I would love to see a top-10 best times per puzzle.

[boxeeboxeebox]

Just to add, usually puzzles just need a single "breakthrough" moment, kind of like sudoku, where everything else falls into place after that. For the breakthrough itself, there are a few types on Puzzle Baron (and these aren't the technical terms, just how I categorize them in my head):

• Direct elimination, where just by filling in the grid, you're left with one possibility for a clue.
• Contradiction, where a clue that says "X is A or B", or "Of X and Y, one is A and the other is B" is determined by other clues.
• Indirect contradiction, same as above, but the determination is from the grid (this is by far the most difficult and slowest type of elimination).
• Relational contradiction, where you have two (or more) clues that say something like "A is 1 step above X" and "B is 1 step above Y", if A & B belong in the same group, then X & Y must be different.
• Locked candidates, where, for example, if the only scores for Floyd and Zachary are 20 and 24, then 20 and 24 can be eliminated for all other players (this can be expanded to 3 and 4 sets).
• Grid-based elimination, where you have to compare open squares between two candidates to determine if they can be eliminated as a match. For example, if John received either 20 or 24 points, while Mary received 28 or 32 points, then they cannot be partners on the same team (this isn't difficult, but can be annoying if it requires the second or third row to be filled in).

Of those, the two trickiest ones are the indirect contradiction and relational contradiction.

Relational contradictions can be tricky for players that play for speed, because the temptation there is to clear clues that you've already confirmed the positions of. So, using the same example as above, if you've already solved the position of A and X, you might clear that clue, but then you might not think to eliminate Y as a candidate for X.

Indirect contradictions require you to use the second or even third rows of the grid, and they can be quite gnarly. You can usually tell if this is the case if most of your clues don't have to do with the ordered/numerical component. The solution will hinge on a clue that says "Of A & B, one is X and the other Y", and you can only determine the elimination by filling in the grid based on another clue – which may not refer to any of the candidates in that first clue, but is in the same group, e.g. "W is one of C or D", where C & D are in the same group as A & B. In the latter case, A & B and X & Y have been eliminated for W, but it's a) not obvious and b) not easily tracked unless you're marking in the grid for all rows.

[tviolet]

One of my favorite things is seeing the record progression, as we all get better at the puzzles. I would love to see a top-10 best times per puzzle.

Oh my gosh, I would love to see a top ten added to the logic puzzles. I was doing a different Puzzle Baron puzzle and it did have a top list so it seems doable.

I'll add my two cents on speed. Like the poster above (Hi, contrary! It almost feels like I know you since I always see your name!) , I also do 4x5 exclusively, mostly just because the grid and the clues fit nicely on my laptop screen so I don't need to scroll.

When I first started, I was also in awe of the superfast times people were posting. But I just kept doing them and eventually I'm at a point where I'm reliably in the very fast range. I even have a couple top times which surprises me.

I used to try different tricks like doing certain types of clues first but I discovered it was faster if I just went through them in order. I concentrate on the top row of grids but I do use the other ones a little to eliminate relationships.

I mark clues as used to reduce clutter if I've gotten all the information I can from them. If it's a x=y, mark that and done. If it's a clue that just pertains to the top grids, mark those options and remove clue.

I think the thing that helped me the most was to keep the previous clues in mind and go back and use them as I do the following clues. If I do this as I go, I don't need to go over the clues again.

Also, how well I do can change a lot. Sometimes my brain is firing on all cylinders and I solve the puzzles really fast. Other times, if I'm really tired or have just eaten a big meal or had lots of sugar, I can feel my brain slowing down; the puzzles seem really hard and I get a lot slower. It happens, I don't stress tho', I'm just doing these for fun anyway.