genina.com
Follow GeninaDotCom on Twitter Follow Genina on Facebook Sign in Register
search | recent topics | hottest topics
Messages posted by lampshade
 
Forum Index » Profile for lampshade » Messages posted by lampshade
Author Message
hippa wrote:I use my left thumb to select the squares and my right thumb to select the numbers and pencil mode on/off.


That is a great approach. I feel like I waste a lot of time moving my finger down to change numbers and turn pencil mode on and off, then tank a good run when missing a button because I'm moving too fast to stay accurate. That said, you have a much different style than I do, and maybe the two handed approach wouldn't help me all that much.

My top hard and expert times are a mix between a careful study and place -- slower, and more cycling between numbers -- and a rapid, blind filling of pencil marks, followed by a fast paced picking off of easy targets. In my careful studies, I don't touch pencil marks until I get stuck, and if that happens, my time isn't going to be a top 100, so I slow down and enjoy a stress-free puzzle solve. Far more of my top daily 100 times are from using the heck out of pencil marks. Those times will always be among the slowest of the top times posted.

When it comes down to it, I'm not a great sudoku player, but I'm good enough to brute force myself into the top daily 100 often enough. With that, I was still able to get a 0:34 easy and 1:39 expert. Watching your expert video makes me realize how much faster I could be if I were better!

Thanks for posting those videos. I should try to get some of my own up on YouTube.
I've done a lot of competitions, and I've never come across anything that can't be solved, but yeah, it's irritating to see a time of 0:00. I have been accused of cheating for times I was proud to achieve legitimately, so it feels weird to criticize. I will jump on the band wagon and claim that any puzzle should take at least a few seconds to complete.
jo wrote:P.s. I don't use any kind of aid including pencil marks. I see no purpose in aids, unless a hostages life depends on my time.


But you're perfectly fine playing a puzzle game, on your phone, without aids? I'm pretty sure those hostages aren't going to care about whether you're using pencil marks while you're playing a computer game.

Pencil marks are an option to use in this application. I could play golf using only a putter, but unless it's mini golf, my bad score would only be my fault for not using the tools available.

I don't have a problem playing this game however you like. It's just a game! But this thread in entitled "timer exploits." There are quite a few people convinced that there is a great deal of "cheating" going on in this game, based on the observation that there are better times being posted then they have been able to achieve, and that they think the use of pausing enables offline solving. This upsets me because I have never worked a puzzle while paused, and I can consistently post scores to the top-100 lists. It's another matter to claim that legitimately using the features of the game the way they are intended to be used is cheating as well. The fact is that very few hard and extreme puzzle times in the top-100 list are from solving without pencil marks. If you want to be competitive with those times, you can choose to play however you like, but keep in mind that the decision to opt out of using pencil marks is purely a choice. It doesn't change the fact that others can use them, and it is not cheating.
Since I had nothing better to do this evening, I decided to find a free android screen video capture app and record myself playing a few extreme games. The app isn't that great, and crashed that phone a couple times, then on the first take, I recorded a 2:20 time (ties my 10th best)... at least I thought I had. (Like I said, the app isn't great.) On the second attempt, I captured a 2:35. I've never uploaded a video to YouTube, but I may get this one up. I'd like to get some audio commentary added before I post.

A couple things I note, based on watching the replay:
- I got a time that ranks in the top-100, but there are several excruciatingly long pauses between moves. Some of those were for yelling at my dog (for barking -- it was after 11pm, and I have neighbors), but many were scanning the board, looking for a next move that seems obvious as an observer.
- Tonight's extreme puzzles were pretty easy, compared to many. Play enough puzzles in this app, and you start to encounter vastly different difficulties writhing the same chosen level. I can't blame the app, or it's developers, for failing to provide Guinness book competition quality puzzles every one of the hundreds of times per day, every day. They cannot vet all of them. It was sheer luck that I started recording when I got a couple slow pitch lobs right over the plate.
- I play this game enough that I really should just buy the no-ad version. This is the best sudoku app I've tried.
- While not the form of cheating this thread seems to be focused on, I do use a number of "aides" provided by the app. These include digit-first, pencil marks, cell highlighting, etc. These are all considered cheating by some, and absolutely assist in keeping up speed and quickly finding solutions. Also, none are allowed when trying to record a time for Guinness. What is relevant is that everyone who plays sudoku with this app has access to all the same features. If you don't use then, you're making a choice to disadvantage yourself against the other competitors.
- When I went online to look at my time, only my 2:20 was there. Not a problem: only the best time from each player shows up in the top-100 lists. Also, the very top time was 1:35. This is fast, even for typical top-100 extreme times, but still not sub-minute. A sub-minute time does occasionally show up in extreme, but very rarely, and still isn't down at the times posted in the easy list. That's not statistically consistent with cheating.

I'll post the link to the video when I get around to posting. Keep in mind that I am lazy, and spend much more time playing sudoku that doing stuff I should be doing. I've got a sink full of dishes to do, for one.

Edit: spelling
Argle wrote:There is obvious cheating going on. The real question should be, what is Genina going to do about it? I have seen read some of their response, and it seems like their response is just deal with it. I would not care if people cheated, I just would like to know how I really compared to the other players. And yes, you can complete an "Extreme" puzzle under 4 minutes. Sorry if you haven't been able to yet.


So, what is your proof that there is cheating? I am, honestly, really interested to see the evidence.

As I have stated many times, I can reproduce times that are consistent with the top-100 times without cheating, so either cheaters are going to a great deal of trouble to mask the results with something statistically believable, or the cheaters aren't very good at cheating. In the TWO methods I am personally familiar with, BOTH are capable of producing times as low as one second, and NEITHER could realistically be prevented by Genina. To be clear, these cheating methods have nothing to do with taking an unsolved puzzle, pausing, solving elsewhere, then populating results within the Genina Sudoku app. I don't consider this to be a form of effective cheating, since IT IS MUCH FASTER FOR A GOOD PLAYER TO SOLVE THE PUZZLE THAN DOING THIS.

What are the methods I know about for cheating? I'm not going to disclose them in this forum, as I don't wish to educate people on these techniques. I will say that I have experience in software vulnerability assessments and reverse engineering. I am also intimately familiar with the methods used to discover, investigate, and prosecute matters like voting fraud, as well as how to mathematically demonstrate, to a level acceptable in court, that alleged voting fraud has not occurred. The same techniques can be used to investigate matters like cheating. There are many similarities, one of which is that, very often, people believe that some wrongdoing has taken place because they do not like, or are unfamiliar with patterns found in results. Also similar is the fact that when wrongdoing does occur, the effects are not only very obvious in the statistics, but also very different than what most people would expect.

When dealing with computer crime, I have also found that more than 99% of the time, when someone thinks their computer has been hacked, it has not. Alternately, in more than 99% of the cases in which someone has been hacked, they were completely unaware. Humans love patterns. We find patterns where there are none, yet many patterns that occur in nature do not align well with what we expect to see. This is most obvious when dealing with randomness. Ask a person to write out a 1000 count random sequence of 1's and 0's, then flip a coin 1000 times, recording heads as 1 and tails as 0. Two things always happen: most humans will think that the coin flip sequence is the human forgery, but there will be much lower entropy (the mathematical measurement of randomness) in the human sequence. That's just one of hundreds of tests that have been conducted over the last few hundred years that show how bad humans are at "sensing" the difference between a naturally occurring distribution, and one that is the result of human manipulation. It is also natural for people to jump to the conclusion that these mathematics are flawed after hearing the results of these studies. I am satisfied with the math, based on what I know about where it is used, and what we have been able to accomplish based on the use of this branch of mathematics. For one thing, much of the technology that is used to power the digital radios used by satellites and cell phones wouldn't work without it. So whether or not you believe that it works, you rely on the fact that it does (without knowing so) all the time. Luckily, it all works without you having to believe it does. The only technology that only works when you believe in it is Peter Pan's magic pixie dust. Ironically, the game of sudoku can continue to be played with this app with no cheating, no matter how much you believe that something else is happening.

And yes, you can complete an "Extreme" puzzle in under one minute. Sorry if you haven't been able to do that.
I just did the competition puzzle for this weekend. No pausing, no use of outside help, no clock/timer manipulation, etc. I did use pencil marks. I got a time of 2:51 on that Extreme puzzle.

I don't play sudoku every day, though I probably play on more days than I don't. When I play, I tend to do several puzzles in a row. This pattern of playing has resulted in playing several thousand puzzles at each difficulty level. Today's competition was, honestly, pretty easy within the extreme category. Some extremes can take me almost 15 minutes (without pausing). That said, my time for this puzzle didn't even break my top 10 times for extreme. None of those times involved cheating or pausing.

What I've said in the past is that cheating may be possible, but the fast times posted on the site are rarely the result of cheating. I should stop saying that cheating is possible - not because it's not true, but because that's what people get caught up on. Cheating is not a widespread problem. It's not much of a problem at all. It's definitely not a problem that merits the implementation of mitigations. What cheating really is: a concept that fuels a great deal of "sour grapes" attention on the forum without merit. I'm guilty of breathing life into this dead, dead horse. For that, including this post, I apologize. But it's time that it is said: There's no cheating problem. There are just good players getting low times.

I have no delusion that people who are convinced of cheating will read this and suddenly realize they have been mislead. But if you ever want to stop by and watch me play, the invitation is open. Also, as said before, I'm not that fast. I just happen to be faster than 99% of others posting times to this site on some days.
The topic of cheating -- specifically, cheating on times submitted to the Genina sudoku site -- is one that has been brought up several times on this forum. It is obvious that many players participating on this forum believe that there is cheating taking place. I have my thoughts on this issue, and although those thoughts can be found by reading my past posts, I'm keeping them out of this thread for now.

So, I have some questions I'd like for your input on:

Do you think there is cheating occurring on this site?
If so, do you think that the cheating taints the legitimate results?
How do you think cheating is likely being performed?

I'm very curious to see what people here think! Please feel free to add your own questions, and contribute any comments you believe to be pertinent to the topic.

Lampshade
psoul wrote:To developers:
Show two timers when the app is paused.
When you start the game, record system time and show a timer based on current system time - game start system time. As soon as the game is paused, start showing the "on-screen time" (which is the current way of showing the timer). When the game is completed, show the on screen time for personal tracking, but take (game end system time - game start system time) as the official time to be put up on the leaderboards.
Simple enough to patch this. No?


So, just in case it's not obvious, I'll state it clearly: if there are ever any times posted the leader boards that are a result cheating, those times are NEVER the result of using the in-game pause. The top times I usually see on easy are realistically achieved by straight gameplay by a very skilled player. Even those skilled players would be unable to post a faster time if they were to copy in values from a pre-worked solution. If anything, doing that would slow them down. When you see obvious cheats (times less than 10 seconds) it's the result of working around the game itself (and all of the rules placed within the game). Making the game more complicated to add more rules does absolutely nothing to stop cheating, while at the same time they could result in decreasing the reliability and usability of the app.

The top times on the statistics page as this post are:
Easy: 0:33
Medium: 0:49
Hard: 1:26
Expert: 1:41

Every single one of those times are possible to achieve by the most highly skilled players, and statistically consistent with a no-pause game. I have only very rarely seen a time posted to those leader boards that is likely the result of cheating. I can say this honestly to myself (and you can choose to believe or not), because of some reasoning based my own non-cheating play. I know what my own best times are. I see where I (somewhat consistently) place on the weekly challenges. I can take my own best times and extrapolate where the times the better players should fall relative that. What I see on the boards is almost perfectly aligned with my calculations. I post my personal best times in my signature. You can get the other times from what it available the genina site. If you want to do the calculations for yourself, feel free. You may find it interesting to compare your results with statistician (or other math professional -- I highly recommend talking to a cryptomathematician, as they have a very good sense of entropy and pattern consistency within a supposedly random series, and can give you great pointers on finding even good counterfeits).

Perhaps this is something that should be a new thread the forum, though I'll go ahead and start it here: What are you (that is, anyone complaining here) seeing on the top-100 and/or weekly competition time postings that is leading you to conclude that there is cheating being performed? As a follow-up: Assuming there is cheating, what leads you to believe that it is a result of pausing/solving/resuming?
In my experience, nearly all easy and medium puzzles can be completed with no pencil marks by using logical elimination. I cannot get below 1:20 with pencil marking, and my best times for easy and medium are both below that.

I can complete many hard puzzles with no pencil marks, but usually my times are slower that way. However, some of my best times for hard were achieved with no pencil marks. If I went without the pencil more often, I think I could eventually replace my entire top ten list with those times. I have been able to complete extreme puzzles without marking, but I never even try if I am aiming for a good time.

When it comes to the weekly completion puzzles, I will use pencil marks for anything but easy. For medium, I know the time I get won't be a personal best, but I can get a decent time consistently that way. For me, even if I know I'd have a 90% chance at a better time without marks, I still go for the consistency in the competition. Maybe that is why I'm rarely in the top handful of times, but nearly always in the top 5%.

In terms of advice, the best I can say is to put time towards practice, and practice without the pencil marks. You will start to learn techniques that work best for you, and see patterns emerge that help you perform logical elimination. Learning these through experience is infinitely more valuable than being told "how" to do them. Your times will naturally, though gradually, start to drop as you go, and that progress will help keep up your motivation. Remember to compare your progress to yourself, instead of to others. When you look at the times other people are posting, it can be distracting, and might motivate you complain, rather than to have fun and play more. Soon, you'll be posting the times that others will swear must have been the result of cheating!
I'm in the same boat as you for current speeds, so I don't have any advice to give for getting faster. I will point out one thing, though: you can't guarantee that you can get a great time on any given device. I have installed this app on several devices, and can attest to the fact that some android devices are just dog slow. There are some that can take more than a second to register a touch in a cell and respond by displaying a number. My Kindle Fire (first gen), for example, is so slow that I've never been able to get an easy puzzle solved in under a minute. As a comparison, I can generally crank out a sub-minute time for about one of every five easy puzzles on my Nexus 4. I don't blame this on the app. It's just a slow piece of hardware!
My best legit extreme time is 2:08. I have a top time of 1:30 on my high score table, but that was some glitch where an easy competition puzzle registered in the extreme category (weird, I know!).

My top time in Easy is 46 seconds right now, and I know I could do better, as i find myself hesitating for stupid reasons. After having played several thousand games using this app, I can completely believe that there are some players that can achieve sub-30 times in Easy, given the right puzzles. I come across at least one easy puzzle every day that requires no backtracking (i.e. using digit-first, only selecting each digit once).

When I was in junior high, the gym coach brought in a radar gun to see how fast we could throw a baseball. I threw as hard as I could, and could only get the thing to clock somewhere the 40 mph range. One kid managed mid-50s. At that same time Nolan Ryan was near retirement and still flinging balls in excess of 100 mph on a regular basis. If I used the logic of most of the posters on this board, my gym class experience would have led me to say that was impossible to pitch a baseball anywhere near fast as 100 mph without cheating, and call tv coverage of pro ball games "fakes." Seriously, folks, it reeks of sour grapes.
This is not the first time the subject of cheating has be brought up on this forum. Here's a few facts I'd like to point out:

1. It is possible to cheat on this game. It is still possible to cheat even if pausing were not allowed. (Please stop suggesting this.) I'm not going to post directions on how to cheat.

2. There are some really fast Sudoku players that play using the Genina app, and they post times to the boards here. A time of sub-30 is very possible to achieve if you have good Sudoku solving skill, great finger dexterity, and a phone/tablet/device with a good touch interface. My best times in Easy are just over 40 seconds, and I'm a bumbling oaf (who happens to spend a lot of time playing this game).

3. There are enough people playing and submitting scores on a daily basis to ensure that the top 100 times are all excellent times.

4. I checked the top-100 times for all difficulties this morning, and there are no times posted that would require cheating. (Note: I did not say that none of the times posted were the result of cheating, rather that the times are all in line with what can be achieved without cheating.)
lampshade wrote:My best time on easy is 1:50, and my best time on medium is 2:06. I suspect that my limitation is my fat fingers on the touchscreen. I would not be surprised if there were legitimate times for easy in the sub-1:00 space, but I doubt I'll ever be able to get there.


I switched to "digit first" a while back, and my times started dropping like a rock. The way I now play, there are many fewer touches required to complete any given puzzle. My best easy time is now just over 40 seconds.

I'd like to know the algorithm they use to determine the difficulty level. There are several Easy puzzles that I can tear through in less than a minute, but there are some other I encounter that take me over 4 minutes to solve. This can be a very frustrating experience, as my slowest top-10 time for Extreme on my phone is around 2:40 (and the best is under 2 minutes). Of course, to get the good Extreme times I am using the "notes" feature -- something I never use on Easy, and only on Mediums when there is a competition.

Back to the cheating thing -- I can now believe that there are people out there capable of solving an Easy puzzle in less than 30 seconds. I do not see how it would be possible to even touch the screen fast enough to solve any puzzle in 20 seconds or less. The single-digit times that you occasionally see are absolutely cheats.

I just looked at the top-100 times for today. In the Easy category, I see times from :34 to :57 (I'm at position #86 with a time of :53!). Based on my experience, I can believe all of these times. If there is anyone cheating in the Easy category, then their cheating system sucks, or they want the times to blend in. The other categories seem in order as well. In fact, my personal top-10 times for each category would all be top-100 daily times today (though I would only crack the top 10 of those in a couple of cases). It has been a while since I have seen a time that I would consider to be a cheat posted to the Genina boards.
sharkattack_nt@hotmail.com wrote:Reasons for Improvements:
1. I recently changed from cell input first to digit, and this has made a huge difference, particularly in easy.


I did this right before I got on the plane to fly home for thanksgiving, and since then I've completely replaced my top ten times for all difficulty levels. My top easy time dropped from around 1:40 to 1:08. I am pretty sure I can crack the one minute mark with a bit more practice and luck.
The name of this thread is "timer exploit," but the real theme here is cheating. I'm going to go out on a limb and say that the super-low times people are seeing are almost NEVER due to the use of the pause function. Take Easy mode as an example. A legit low time (no pause) can be as low as 40 seconds, as pointed out in a previous post. Lower legit times are possible, I'm sure. The times that we all assume are cheats include 11 seconds. I want everyone that thinks pause is a cheating method, for the sake of science, to try it! Seriously. Grab a camera, a pencil, some paper, and the device which you use to play Genina Sudoku, and try to match 11 seconds. Try to beat 40. Heck, try to get a time under 1 minute!

The use of pause might be a cheat for Extreme mode, but it certainty is not for Easy mode. If you look at the percentage of competition entries that you consider cheats for each skill level, I suspect you will find that they are all about the same. What this indicates to me is that people aren't using the pause mode for cheating on Hard and Extreme, unless they are using some other, more effective, method for cheating on Easy and Medium. My conclusion? Pause is almost never used to cheat! Due to the number of posts to this thread, it is obvious that cheating occurs often enough that it is seen as a problem.

I assume that the bulk of posters on this thread are NOT cheaters. What is happening here is that people are asking themselves the question "if I were to cheat, how could it be done?," and the conclusion often reached is "pause." Believe me when I tell you: the people actually cheating have thought about this question far longer than you, have much more knowledge of the internal workings of smart phones, and have come up with much more effective methods of cheating than using pause. Of the many things this means is that disabling pause will have absolutely no effect on cheating whatsoever. Please read that last sentence a few times and let it sink in.
 
Forum Index » Profile for lampshade » Messages posted by lampshade
Go to:   
Powered by JForum 2.1.8 © JForum Team
© 2009 genina.com Home Privacy Policy Contact Us