Authoritative Minesweeper

On 19 August 2016 someone claiming to be Harryck Repse wrote in the Guestbook saying his scores and many others were fake.

He said he was many players and their scores were fake:

1) Harryck Repse - Best Latvian - On ranking with 1-12-43.
2) Vazgen Geghamyan - Best Armenian - Not on ranking. (Was on ranking with 1-11-55 but then submitted obvious fake 10-37).
3) Matthew Rendo - Youngest to join World Ranking - On ranking with 2-19-65

He also accused three other people of cheating:

4) Shen-Jia Zhang - Former best Chinese and webmaster of the China Ranking - On ranking with 1-11-38 (Clone videos).
5) Jake Warner - Former best American - Not on ranking. (Was on ranking with 1-10-36 but removed in 2007 after refusing to send more evidence).
6) Stevan Gzovdenovic - Best Montenegrin - On ranking with 2-12-41.

All 6 players have been accused at least 1 time about cheating. For example, Shen-Jia submitted fake Camtasia videos in 2003 (using Windows XP Minesweeper) to the first China ranking made by Ruan Yi, but later made the famous China Ranking website and introduced thousands of players to the ranking.

At first I thought this was someone trying to make the community more suspicious about these players. 'Harryck Repse' then gave more evidence to prove he was telling the truth, and then I found a report written by Jonian Grazhdhani in 2007 (emailed only to a few selected people). This is a 48 page investigation into players who used a cheat feature in Arbiter 0.43 to load Clone boards into Arbiter and play with UPK in normal mode to make records. The report talks in detail and is suspicious about Harryck, Vazgen, Jake, Stevan, Luke (a confirmed cheater) and other players.

I think this Harryck is telling the truth, and I will try to find more evidence to make the World Ranking clean. The report is attached:

Edit: Best way to verify this is me is geolocate the IP address of this post versus my old posts. It'll show the same city as I still live in the same house. I have no beef with any players here, so my accusations are only based on my statistical analyses over the years, and not intended to be malicious towards anyone. My posts are pretty vague because it's been nearly 5 years since I've been interested in what's going on here, so I'm doing my best to speak from memory. I bricked my old computer in 2012, which was loaded with videos from all players and a ton of info that I had collected - unfortunately this is all gone, and would have been a great help as I would have had much more to offer back then, definitely more so than the vague info that's been pushed to the back of my memory.

I know you asked for a personal email, Damien, but I'll post most of my comments here. I will email you from harryckrepse@mail.com, let me know if you receive it, though I unfortunately don't have much more to offer.

I have no way to prove I was Matthew Rendo, but the scores I submitted for him were actually real and are indicative of my true minesweeper skill at the time, so I don't know if you wish to keep "Matthew" on the high scores. The birth year 1995 is true as well as the scores, I only falsified my name and birthdate. So, in truth, I was the youngest player to ever make the world rankings... though there is no pride in this as I was simultaneously cheating on different accounts.

I was a few other players too, none of them particularly good, and none of them cheaters (just fake names/countries). I can't even recall the names or the countries I used.

My 56.97 done by Harryck Repse on 2011-02-24 is a legitimate score, and I now do remember actually hitting a sub60 randomly when not trying to cheat. I don't think I actually got any other legitimate sub60s. I could certainly not sub50 at any point. All this is besides the point; I should be removed entirely from the rankings.

I recall reading Jonian's paper when it was published. Regarding all players mentioned in the paper:

Vazgen Geghamyan: This was me. I wanted to test the waters to see how much you could get away with. I used one of my earlier Harryck Repse boards and slightly tweaked it to make it EXTREMELY easy to solve. Compare Vazgen's expert score to Harryck's 77.21 or 58.98. I don't recall which board I used.

Jake Warner: Cheat cheat cheat, as I've said before look at the fourth column from the right on every one of his submitted Arbiter 0.43 expert videos. There will be no mines in this column. Statistically it's extremely unlikely for this to happen to someone on multiple boards. Look at all my Arbiter 0.43 videos as well - same issue with the fourth column from the right.

Thomas Kolar: No he's not a cheater lol. I think everyone should know this

Harryck Repse:

James Custer: Legit

Luke Chiang: Already known to cheat, nothing to discuss

Stevan Gvozdenovic: This guy's a mixed bag - he's a talented player but his expert score is very obviously cheated. You don't even need to analyze the average length of clicks to come to this conclusion, the cheating is so poorly done it's hard to believe this 11-year-old score is still not rescinded. I believe that his intermediate score may be legitimate, and that some of the other videos of slower expert scores are legitimate - perhaps his expert score should be altered to a previous, legit score. Then again it isn't possible to tell if ANY of his scores are legitimate, because he could have used speed-up software on all his games, just upped it high enough to make it obviously fake on his expert score.

Elmar Zimmermann/Tam Minh Bui/Ian Fraser/Dennis Lütken/Jonian Grazhdani/Robert Farnik/Alex Poehner: Yeah all these guys are clean as could be.

---

Shen-Jia Zhang used an altered version of clone 0.97 to procure fake scores, as did several other Chinese players of his time - unfortunately I do not recall the names of any of these players as it has been many years. To my knowledge, funny mode was distributed without the "funny mode" thing appearing in the video files. That's not to say Shen-Jia Zhang isn't a talented player, just that the legitimacy of his scores may be up to debate; obviously the Chinese community knows much more about this than I. I believe the current Chinese community is incredibly devoted to minesweeper and features many extremely talented players, which is exciting to see because I believe minesweeper is something that has much to be discovered.

I wish I had all those videos and statistics I collected between 2005 and 2012. I would have been a much greater help than I am now... all I can do is apologize for my cheating and tell you what little information I remember.

I tried hard to give you more information about Shen-Jia Zhang. Years ago I received a .zip file which contained most of Shen-Jia's sub50, so the date of files inside was just the accurate time the replay was saved.

I joined saolei.wang around 2008-12-21, the first thing I did was to download all Shen-Jia's records and studied how to play. In the picture below, I select blue rectangle identifying all replays which were not contained in the .zip file(so they were all downloaded after the real solved date), and use analyzing tool to give you the accurate solved date of them. Only three sub50 games of Shen-Jia was solved after 2008-01-26, the date of his sub40.

In the first column of the picture, there're 17 sub50 games with 140+ 3bv, and 35 sub50 games total. Almost half of them are sup140.
In the second column of the picture, there're 28 sub50 games, but how many 140+ 3bv are there? None!
In the third column of the picture, there're the only three sub50 games with no games with 140+ 3bv.

So the conclusion is, after the day 2008-01-19, no 139+ board is solved with sub50 score.

In another picture, I put a red line on the key date 2008-01-19 so you can easily find it out.

I knew the story about Zhang Shen Jia years ago.I just want to know how did Harryck found out this.
Harryck said he spent 100s of hours ananysing videos.I did the same thing and got terible conclusion about 20+ our players cheating.

This time your anouncement give us minesweeper players a good chance to do sth. to clean up the dirts.

btw the Gvozdenovic's case Post by me on Mon Jan 05, 2015 10:49 am time to remove him.

Harryck_Repse wrote: Thomas Kolar: No he's not a cheater lol. I think everyone should know this

Well, that's a huge relief In case anyone is interested years later, by the way - I was using both the computers at my school and my laptop at the time XD And may have installed the new Arbiter version next to, instead of over, the old one? I don't remember clearly. But yes, Joni's analysis is very accurate XD

Anyway - @Harryck? Matthew?: As far as I can see, you've come clean, and I personally don't see why your legit scores shouldn't stay on the ranking. Would be nice to have your name there, though!

Remove Gvozdenovic's 40.72 and Zhang Shen Jia's 37.82 as soon as possible
Remove Harryck Repse now and many thanks to him
Big doubt of Kyle's 45 and 10.94 int game after ananysing his stats.dat
A very very little bit doubt of Gottlieb's 3 new records for it seems too easy to get USA.1 back

Hehe, I admit my 39 was very lucky, but I have been getting a lot of sub50s as well as two more 42s recently :p If anyone wants to see my arbiter stats file or all my recent games I'm happy to send a zip their way. Really, I've just been playing a lot lately, and trying to complete boards more often.

It's very unexpected to me that Harryck was a cheater this whole time, since usually cheating comes from people who pop in to a community and post a few great scores before either leaving or getting into a huge legitimacy debate. There was one big exception in cubing (around 2008, a guy who was very dominant at the time and had won a lot of money at world championships turned out to have been cheating at blindfold solving for who knows how long) but it's still not something you think will happen, and just participating in the community for a long time usually means someone is for real. It makes me wonder about the density results and the many nonosweeper results too (are none of them real?).

Personally I'd rather be a little more strict about possible cheating. If the community thinks someone is very likely to be cheating, remove them. Then only let them back in if they have shown they are capable of the results they claim - and even then, only accept their scores that are not suspicious (too fast movement or clicking compared to their other games, bad clone version, etc.).

qqwref wrote:It makes me wonder about the density results and the many nonosweeper results too (are none of them real?).

http://www.stephan-bechtel.de/nono-worldch.htm

I have seen Stephan and Christoph play and they really are incredibly fast. I'm pretty sure that most nonosweeper results are legit.

Oh, I didn't mean to imply that any nonosweeper results *but Harryck's* were cheated. He did get a lot of results though, if you consider all those tournaments.

Zhang Shen Jia_Exp_50.61(3bv193).mvf 2009-3-29

alix wrote:Zhang Shen Jia_Exp_50.61(3bv193).mvf 2009-3-29

2007-02-20 04:29:41

download here