My thinking on moderating is to treat everyone as adults. I don't like to "lord it over" people, and instead try to give everyone as much leeway as possible.

I agree that uhoh's posts have been marked by aggressiveness and sarcasm that grate on other people's nerves. 

I also agree that Markovich's posts have served to ratchet up the animosity. I don't know whether uhoh is rdecredico (did I spell it right), but I do know it doesn't matter. 

I've given the respondents a chance to show they can salvage this discussion and make something useful or interesting out of it. 

After viewing the comments since I split the thread, I've decided to lock this thread.

  Stigma wrote on 03/22/11 at 14:15:05:


Yes, that's what selection bias is (not what Markovich is saying here). 

Stigma wrote on 03/22/11 at 14:15:05:


Uhhmm.... no. The first example is an example of selection bias. The second is not, it would appear to be one you just tried to sneak in along with a one you found elsewhere on the web that happens to conform with Markovich's claim. The only thing the second has in common is that they "select" the openings based on how they're going to play. 

Selection bias isn't anything to do with what the sample itself does. It is what the researchers select. I don't know if you get the humour of it of trying to nonchalently sneak in this obviously different example, but I find it kind of amusing (I swear I'm not trying to sound condescending here, just expressing how I find it).   

Stigma wrote on 03/22/11 at 14:15:05:


No, that is the exact definition of selection bias. You just stated it yourself above, when they take any type of non-random selection. 

Selection bias isn't limited to the social sciences, any type of experiment you do is susceptible to selection bias. If you do an experiment on mice, then you may have selection bias because of the mice you use in your sample. If some of the mice choose to eat a certain type of food and you know the reason why, then that would not constitute selection bias. That would just be a normal part of the experiment, nothing to do with selection bias. If it affected or prohibited your conclusions, that would be called "confounding factors", not selection bias.

I don't want to rekindle the flames of this thread again, but there are two confusing issues I don't want to leave behind just yet:

- You are really talking about two different things when you say "selection bias". Markovich's concept is the well-known problem from the social sciences when you use any kind of convenience sample instead of a truly random one: People may have selected themselves to the groups (or indeed to the entire study) for non-random reasons that could distort the results. Examples: Surveys on the internet (avid internet users are not a random subset of the population), and people with peaceful intentions playing certain openings because of their drawish reputations (their intentions rather than the opening's real merit may account for the larger drawing percentage).

Uhoh on the other hand is talking about researches taking the wrong games/data, which is also a possible definition of "selection bias" I suppose, but not the standard one. Just thought I'd try to clear that up.

- Contra Markovich I think there are clearly drawish openings even below 2250, but these may be different openings from the ones that are considered drawish on top level. The way to go about drawing with Black is to find openings where the most frequently played lines lead to calm, positional play. The same openings may well contain wild, tactical lines in theory, but if those are rarely played on the level in question it doesn't matter. Usually that means finding openings that are not the most common, so opponents won't be well-prepared.

Exhibit A: The Black Knights' Tango. White has both critical and sharp lines, but most common between 2000 and 2250 are either 3.Nf3 e6 4.Nc3 Bb4 5.Qc2 reaching the Zürich variation, or 3.Nf3 e6 4.a3. Both lead to calm, positional play, especially if White doesn't know any concrete, critical lines.

Exhibit B: The English Defence. The most common line on 2000-2250 level is 1.d4 e6 2.c4 b6 3.a3 f5 followed by calm development with Nf3 and g3 or e3. Theory says White should play an early d5 to get any kind of advantage, but people in this rating group often fear some preparation and avoid that move. The result: An equal and quiet game.

Exhibit C: The Bogo-Indian. Has a solid reputation even on GM level, but the concrete strategies White uses to fight for an edge there, especially nursing "the 
advantage of the bishop pair" throughout the game, are often beyond players below 2250, myself included.

TN wrote on 03/22/11 at 04:17:34:


Personally I would ban him, either by user name or IP address.  But that's just me.

This Topic was moved here from General Chess [move by] Smyslov_Fan.

TN wrote on 03/22/11 at 03:42:25:


You're really not a very intelligent man are you? 

trw wrote on 03/22/11 at 03:00:47:


They are not wrong.

If it doesn't come up as an image please click on the dots. 



Search criteria is: C90-C99. Rating: 1900-2100 (both). Years: 2000-2010.  

What *I* would like to know is how *YOU* are getting your statistics. How are you getting them? I am just interested. Are you really incompetent or is it a different database or what is it? 

I would NEVER say a person was straight out lying about what they said like that, especially not something as straightforward as this. I would ask "ok, how are you getting your figures" or I would say how about we find out what the discrepency is, and if someone has made a mistake fair enough as we all make mistakes sometimes. 

MAYBE I would say "I doubt you're doing it right, I think you did it wrong", and try to sort out the discrepency but I would NEVER repeatedly state that someone was reading something wrong even after they have repeatedly checked it. It really offended and upset me physically. 

I would therefore like an apology for this. 

Markovich certainly wasn't very polite to me! What Markovich has done before shouldn't be an issue, I came here in peace to talk about chess. 

I thought the discussion was going well until he started disrupting it, trolling it, and trying to get me to go into an argument, as he has before. Maybe he feels I am an easy target, I don't know. 

FWIW: Selection bias is a completely different thing. Selection bias is when the researchers themselves are taking the wrong games... the selection of games under study is not random. I thought about pedantically saying that, but I said no, I would hold off and not be so pedantic. It can't make much if any difference here because the number of games amateurs are looking for draws in are relatively small. Even if Black was desperately looking for a draw he still wouldn't receive it as much times with the Petroff. It could be proven with numbers.      
  
Using his logic, we couldn't say any openings are drawish from statistics because even Super GMs may be under some sort of mass delusion that the Petroff is a drawish opening but in fact it only draws a lot because other GMs used it as a drawing weapon when they really needed a draw and the self-perpetuating circle continues!!! 

Anyone who has studied and or is in science or is interested in it knows that, it has little to do with being an economist (wow, an economist).

I have briefly read your latest post Markovich and am largely unimpressed, as I have generally been with your posts so far.

How exactly is YOU starting to butt into everything I'M saying that has nothing to do with you, trying to start a fight, me being hostile and me picking on you? Telling me I'm "unstable". I think you have some issues bro'. 

I ignored your accusations of me having "open hostility" and being "smug" and everything else you threw at me for a reason. It's you who are being provocative. I just pointed out the fact that some openings are indeed drawish or at least more drawish in the best and most appropriate way I could. 

I wasn't rude on these forums, is that what you're accusing me of? If you don't like how I talk you're entitled to your opinion, but don't be surprised if everyone just ignores you for it. I obviously wasn't trying to offend anyone, just pointing out something after I did my own research. 
  
I thought you were accusing me of being Richard as a joke. 

At this stage I don't care if you doubt my figures that I got from Mega Database, after I quadruply checked them specially for you. But I would just like to assure other people that the figures I posted are correct, and Markovich is clearly an unreliable and unstable person himself. And it would indeed NOT be "rather strange" for there to be a higher drawing percentage than winning percentage for White (as there is). Are you still taking that medication Richard?

And we are supposedly not allowed to make a questioning comment anymore, we're all meant to go along with flow so as not to upset the sensibilities of the forum is it? I think maybe you need to settle down a bit as the only one trying to get into a big argument here is YOU. Leave me alone, and don't accuse me of getting figures wrong that OBVIOUSLY aren't.