What do these percentages/numbers behind the moves mean?

Google is your friend:

Tom Martell (semisober) on TwitterName Tom Martell; Location New York, NY; Web http://www.tommar... Bio Ex-code monkey turned ex-lawyer turned financier aspiring to poker pro with a side of ...
twitter.com/semisober - Cached - Similar


Cliffs: ´our´ Tom Martell is semisober on FICS, has a 1591 USCF rating and claims whitewashing hackdeath and montrealcorp 8-0 was legit, despite strong likeliness with engine play.

I would like to point out that Tom's 1591 USCF rating is based on exactly one tournament that has been played in the last 9 years. He appears to have been an active collegiate player, but since 2001 has played only one rated tournament, in 2007. He played three games in that tourney, losing to a 1992 and an 1884, and beating a 1627.

The point being merely that his published USCF rating may be a rather poor indicator of his current playing strength.

Now, I'm certainly not convinced his games were legit, far from it I feel the evidence strongly suggests they weren't. The USCF rating isn't the strongest argument in this case though.

Time spent on each move, that's how it copied from http://www.ficsgames.com/.

I've never claimed to be a 2300. I don't play competitively much any more; the last event I played in was an A section that I won 3.5-.5. I am currently studying / working with a pair of 2200s and I think I am a little worse but in their league. With respect to blitz, I routinely play chess with IMs and GMs who I work with or who I know from scholastic chess from years ago. I generally lose, but I've improved a ton from it.

@Nezh: you seem like a really angry person. I looked at some of your other posts in other threads and they often drip with hostility. I don't really see any reason to engage you on this, so I'm not going to respond to you again.

I came to play a few games for fun. I have plenty of other time sinks in my life; I don't need ones that cause a headache. I certainly don't want to waste any more of my time defending myself to a group of people I don't know and don't plan on interacting with much more. I know that I'm not a cheater but I also know that it is impossible for me to prove this. Feel free to delete me from your ratings list; I don't know why you would add someone with only 8 games played to begin with.

Best,
Tom

Someone who is a little worse than 2200 doesnt play blitz like that. someone who is a GM does. You lie about your rating in your notes, almost all your moves match those of an engine. That martyr gambit just doesnt deserve to work.

Hmmm...that does not reflect my self image very well, but maybe I´m deluded....Could you point out these ,,postst dripping with hostility" so I can engage in some self-reflection? Dunno, with regard to your ´case´, when it looks like a duck, walks like a duck and sounds like a duck, it usually is a duck. So for the moment, to me you are a duck.

For someone not interested, you show a lot of interest lol:


Tom Martell
adept

Send Message User Lists Last Activity: Today 11:02 PM
Current Activity: Viewing Thread 2p2 chess players RATING LIST

He replied by pm to my how did you crush me "i dont really play much online. I'm ~2000 USCF, maybe a little stronger now as I've been working with some IMs recently." Don't know the significance of this, but I thought I'd post it.

I reviewed some of our games. This game's opening is highly suspicious. http://www.ficsgames.com/cgi-bin/sho...62;action=show
He tanks for 15 seconds and then plays 4...dxe4, an unknown move after playing 3...Nc6. What kind of "strong" player plays the uncommon 3...Nc6 variation of the French and then doesn't know how to respond to the most popular move by white.

I don't know. I seem to recall a story where it ends up being a swan.

Well, I have went over his games:

http://www.ficsgames.com/cgi-bin/sho...04;action=show

Black played really bad in this game, even if TM's moves 100% corresponded to the ones of the engine it would mean nothing as everything in this game was simple.

http://www.ficsgames.com/cgi-bin/sho...01;action=show

again, White plays bad out of the opening, loses a pawn, TM goes on to win. The only move that is interesting is b4, apart from that everything is straightforward.

http://www.ficsgames.com/cgi-bin/sho...99;action=show

This game is more interesting and in several spots TM founds quite strong moves. However, I think that most people above 2000 would be able to find them, so in my mind it is not conclusive proof at all.

http://www.ficsgames.com/cgi-bin/sho...22;action=show

this game feels a little weird, I don't think a lot of 2000 players would be able to play the d5 - d4 - Qxd4 sequence but again, nothing conclusive as the rest of the game is pretty straighforward IMO.

I haven't went over the ones with Hackdeath yet, but for now it seems to me that he is just a quite good player.

Of course I might be wrong

I the first game 18.Rhf1 is NOT simple or obvious at all. It is however Rybka´s first choice, by a tiny margin. Rxd4 is the standard speedchess move, Rf1 required calculating some complicated lines and it seems unnecesarily complex compared to the simple Rxd4, which is a perfectly fine, natural move.

I do not believe someone who gets crushed like an ant by a much weaker player in this fashion a few months ago (http://www.ficsgames.com/cgi-bin/sho...50;action=show , I know sample size, but still) is suddenly capable of complex GM-like sequences like shown in the eight games he played in the 2+2 Tournament. Also replay the FICS-games with the ´realtime´ option and you´ll se even many natural (re)captures take TM several seconds. Notice how much faster he plays in the game he gets crushed.

This thread makes me sad. This is the reason I haven't played in any of these tournaments even though I'd love to. You guys are way too quick to call someone a cheater.

In blitz games, I have over 8 wins against GMs and 8 losses to 1400 players. A sample size of 8 is pretty small. I analyze some of my better games and sometimes 95% match Rybka from early middlegame on. Rare, but has happened with a lot of forced lines. Means I have a high chess understanding, though I blunder a lot.

Now I'm not attacking or defending anyone, just realize please that some players play close to there rating all the time, others might not. On a good day, I doubt there is anyone on this forum I can't beat 4-0 given enough tries. Likewise, I'd go 0-4 against some of the lower guys from time to time too.

Now that poker is a good part of my living, I can't afford to be part of a witch hunt. I just ask that you give this some thought before making accusations that, for people with integrity, is very hurtful.

I haven't looked at any of the games in question. I just know two 4-0 matches means little. I personally don't know one person on 2+2 IRL, so I'm not biased here, just trying to shed some light from my perspective.

My two cents,

Karl

This is the first time this has happened in the time I´m around at least. And where you are right one has to be very careful, when there is plenty of call it ,,likely evidence", there is also a good reason to state one´s thoughts.
Now what is Tom´s real rating or playing strenght? We don´t know. We do do know there is a lot of contradicting information about it, both from Tom, the USCF rating list and other sources, like this one:


http://www.chess.com/livechess/profile/tmartell


No 1850 player (not even taking into account for the habit chessplayers have to adjust their stated rating to what they think it should be) plays speedchess on the level Tom Martell did in his eight games in the 2+2 tournie. And this is the last I will say about this as this isnt some personal crusade against someone I don´t know. But for a lack of an arbiter, the players do have have to police the games themselves.

Nezh,

This isn't personal against you. Just pointing out that I happen to be one of those guys you say don't exist. My rating peaked near master level around college. I'm 40 by the way. For whatever reason (inactivity for years I suppose, plus I moved a lot for my job), USCF gave me a new rating. I have very little play on my new rating, I think it is around 1900. I'm way stronger than that as I've improved since college days considerably. If I play in the World Open again (I hope to), I'd play in the U2100 or U2200 section as it would only be fair. Sometimes there are reasons for anomalies. I have blitz wins on Playchess over GMs. Again, I know nothing of Tom's games. I just know your premise is off a little. I promise you that if I played a four game match with every 2+2er, the results would have you puzzled. In fact, on Playchess, I won 72 straight games at one point, all blitz. I also lost 26 straight games. I assure you I am not two different people.

It's very important to emphasize that I, at least, have leveled no accusations of any kind. I have suspicions, because no matter HOW good a player is, it seems extremely unlikely for them to play the computer preferred move almost every time. Especially over multiple games, where the "forced lines" argument is a lot less likely to come into play.

In six of his eight games, Tom Martell made a TOTAL of 5 "errors" according to Firebird. That seems like excessively accurate play for ANYONE under 5'2" time controls, I would find it suspicious even if I knew for sure he was a GM. The investigations into his actual rating make sense, because the higher his rating, the more likely is that he could play that sharply. Of course cld343 makes a good point though that it is only a measure of likelihoods. A lower rating in no way proves conclusively that he did cheat, it merely makes it less likely (not impossible) that he would have found moves the computer considers perfect with that frequency.

From what I've seen, when looking into his actual rating, it looks like it's not a very good measure of his actual strength. I'm pretty sure that his claim to be actively playing with strong players, and to have improved greatly because of it, is accurate. In fact I'm pretty sure I know who one of those players is, and could guess at least one other, thanks to a conversation I overheard on chess.com during today's Anand-Topalov match. So as far as I know, he has told nothing but the truth in this thread.

There are also the two games (one in each match) where he very clearly wasn't using an engine. I'm not sure if this can be used to extrapolate that he didn't use one in the other six games either, or if it just means he played those two legit. Obviously he is a good player, who is capable of beating either of his opponents legitimately.

I still find it really hard to ignore the EXTREME accuracy of the six suspicious games though. And when I look at the amount of time used on each move (particularly seemingly obvious recaptures that weren't played instantly) it adds to my suspicion. As the person who is taking the time to maintain this rating list, I care about its integrity and so I'm not going to just ignore something that seems suspicious without investigating further.

@Tom, I do hope you can understand why we have our suspicions, and not hold it against us that we do. I want to emphasize again that I'm not accusing you of anything, I just want to make sure your games were kosher. As for why I'm including someone with 8 games in my ratings list, it's because that's what the computer spits out. I just keep a database of all the games, and have the computer generate a ratings list from it.

@Nezh, although I appreciate your zeal, you have approached this from a more accusatory perspective than I would prefer to see on this forum. I'm not surprised that Tom came away offended. I strongly disagree that you come of as an angry person in your posts in other threads, but ITT... well... you perhaps could have handled yourself better.

@cld, well said. I should mention that the only reason I initially started looking at the games was because smilingbill, who is both a VERY good chess player, and an impartial observer (he didn't participate in the tourney in question, and isn't on the ratings list yet), mentioned they looked suspicious.

Overall, I do fully intend to delete from the database (and thus the ratings list) any game where cheating is conclusively determined to have occurred, now or in the future. At this point in time I do NOT consider Tom Martell's games to be in this category. Once again, they are suspicious, but there is nothing conclusive that I've seen yet. I would appreciate if YouKnowWho would be willing to take a look at the games vs. Hackdeath and weigh in. I welcome additional debate over the point. I ask though, that we please try to keep this an investigation and discussion, as opposed to being a witch hunt. I don't want to see people driven away from the forum because we're "too quick to call someone a cheater", but I also don't want to see anyone driven away from our tournaments because they suspect foul play. Hopefully we can strike a reasonable balance.

Thank you.

Cld343, so you basically say you havent even reviewed the supposed evidence but are still symphatizing. "Sometimes matches 95%" is a whole lot different to "always matches nearly 100%".
I have played on internet chess servers for 15 years and tens of thousands of games, and have come across only about a dozen players who I thought to be cheating (and have usually been right with (C) appearing next to the player's name later on). I am not randomly calling names here

Hi smilingbill,

I appreciate your efforts. I will never condone cheating for any reason, ever. It was the approach overall I didn't like, not by you, but in this thread. I'm all for investigating said possibilities and getting to the bottom of it first. The truth is what is important here. I can run the games through Playchess' engine assistance detector to help out. Maybe after SCOOP is finished. Roundtower already dealt with the placing of labels before all investigation is complete. That was my only disappointment.

BobJoeJim

Appreciate your efforts as well.

I love this forum and I like almost every poster, with the exception of a banned poster, but that worked itself out. I too am done on this topic, until after SCOOP.

I have no experience in trying to evaluate whether or not a game was played with engine assistance, so if someone who does have that experience would be willing to help out, that would be great. cld343, if you were willing to run them through the Playchess engine assistance detector I would really appreciate it.

For the time being though, while I don't necessarily consider the issue closed, I'm considering the games legit. I played through all six suspicious games again, with my engine running, and kept a close eye on the evaluations at the moments when Tom Martell played his moves. My previous run-through at depth 14 was much deeper than an engine would have been able to get in the time he had available. His moves were actually significantly LESS accurate when measured at the depth an engine would have had time to reach, and he made a lot of moves in each game that would have made absolutely no sense at all if he did have access to the evaluations I was looking at. Obviously proving innocence is extremely difficult, and who knows what things would have looked like if I did the analysis with Rybka or Fritz, but I'm pretty convinced that, at the very least, Tom Martell did not use Firebird to assist him in playing his games.

BobJoeJim -

I did spend some time and reviewed the games in question and it is my opinion that an engine definitely wasn't used throughout the entirety of games. In a game against Hackdeath 37 ...Bf2+ would be every computer's first choice for instance. Tom didn't play that.

Pointing to Tom's loss against a 1700 player, that player played well in my opinion.

Anyway, I ran one game through Chessbase's engine assistance detection and it found nothing. This was the first game against MontrealCorp with 18.Rhf1. I can do the rest later.

Of course, someone could use an engine every third move or something, seems silly though. It really looks at this point like Tom is a strong player. If confirmed, I doubt he'll be joining us again.

Okay, I ran all 8 games of TM's of 2+2 through Chessbase and no indication of engine assistance was detected.

To address taking time for routine captures, some players get in a rhythm and spend a few seconds on every move just to make sure they aren't missing the obvious. When I play tournaments on Playchess, I almost never move in under 3 seconds.

None of what I posted proves someone didn't cheat, but it does cast sufficient doubt that said player did cheat. Until I see evidence to the contrary, I personally feel apologies are warranted. I hope if this subject comes up in the future with any other players, it will be dealt with delicately and respectfully.

If someone calls you a cheater on a chess site, it's easy enough to laugh off I suppose. But when they attach that label to your name on a poker forum that may be your living, it's a little more serious. I'll get off my soapbox now, just disappointed.

Here are updated rankings. These include all games from the Candidate's Tournament, as well as all the Tourney IV games that have been completed so far. If the unaligned columns bother you, then copy/paste the data into Excel, because I don't care enough to fix it

Notes: One new player earned his way onto the list, smilingbill, and he's sitting quite pretty with his 6-0 record. Looking forward to seeing how he does against Montrealcorp or Hackdeath (and if he wins, against Nezh or nikitheone). 6-0 is nice, but the opponents are ranked 18th and 27th, so we have to wait until he plays some other top players to see how that rating holds up. As for the 2010 2+2 Chess Championship that is about to begin, the players are ranked 3, 5, 6, and 7, but are the top 4 rated players with 10+ games so the field seems about as right as it could be.

Here's the latest update to the rankings. These are the rankings being used to seed the 2p2 Open V. They include all games from Tourney IV and from the Championship that have been played as of the date of this post.

This is frustrating