How over-rated are tactics trainer ratings?

Over the board play is much different then solving chess tactics problems.

One difference between tactics problems and real over the board problems and measuring performance is that the player gets 2 or more minutes to evaluate the position before we make our move to counter it. And this is just one of the many differences I can think of.

I really don't think one should expect tactics ratings to correspond to over the board ratings. What the rating does show me is if I am improving or not.

As a point of reference my chess.com tatics rating is 1705 and my USCF rating was, at one time, 1985. I have not played in years and my tactics rating is slowly going up.

The best part of tactics problems is that they are fun! And I am having fun while improving my game while providing a more enjoyable experiance for my opponent. :)

Ratings aren't meant to be compared across playing pools, let alone across playing pools *and* games.  

I agree the tactics ratings are only so you can chart your personal progress and its meaningless to compare them to anything else.  Positional players have to have a strong tactical eye.  A positional player needs to see everything.  That is one of the things that made Petrosian great he saw an attack 10 moves before his opponent thought of it.  That is why all beginners should start with open positions and tactics.

Ratings are over-rated.

Just wanted to say my tactics rating is now over 2000. And I am not 1985 OTB strenth any longer :)

I consider myself to be a positional player, but have a 2309 tactics rating while my USCF is 1766.

The TT ratings were overrated when this thread began, but now they have reversed themselves. I find it very difficult to get my TT rating up to the level of my USCF rating.

it does seem that way doesn't it?I used to be able to get to 2100-2200 in TT....now I consider myself lucky to be able to maintain a 2000-ish rating

I'll let you know as soon as I've stopped beating my wife.

If that is the case then I am a 300 - 400 rated player.   Probably not too far off the mark since I am just really a beginner with 90% of my time dedicated to study and 10% to actual play (5% HIARCS/Sigma and 5% OTB). Since I have not yet played a game here on chess.com, compared to my horrible record against a 1400 rated HIARCS engine (that rates me 1148) I like to think that TT is a "tool" to help my overall game get better.  

I have a love-hate relationship with the timer. Whereas at chess tempo I broke the 1900 barrier more than once, I never got my TT rating much over 1700 here - and when I can't resist to have a go at it when just coming back from work, it has a nasty habit of plummeting several hundred points ... 

Everyone's game has weaknesses. Below IM level, those weaknesses can be significant, be it a lack of opening knowledge, or a lack of tactical ability or whatever. It is not hard to find 2 1600 rated players, one of whom has the tactical abilities of an average class-A or Expert player and another who has the tactical abilities of an average 1200 player. 

Presuming that tactical problem "ratings" have any relationship to chess rating, particularly long time control ratings, is folly. The two have next to nothing to do with each other. The only purpose of ratings for the tactical problem sets is to help users find problems relatively close to their tactical ability. And even then it's really something of a crap shoot. 

By way of an example, on another site that has very detailed statistics, I score close to 90% for skewer type problems but only about 50% for removal of the guard type problems.  

lol

The problem I have is that I spend a while looking for something deep, then finally shrug my shoulders and say, "Well, all I see is RxQ"...and then that turns out to be the solution. 

I'm not saying the weaknesses are fatal such that 1600's will be taking down 2400s. I'm saying that they are exploitable by people who happen to match up well with that specific player and are otherwise competative with them. A 2000 can take down a 2200 if they are better at the chess skills that are critical for a particular situation.  

Is it likely or frequent? Of course not -- guys who get ratings in the titled player range are there for a reason after all. But I've also seen an FM get tactically blown out by a 2100 before too. 

It's always true that anybody matched up well with someone else can beat them...whatever the rating (or title).

Ratings are accurate within their own pool of users. You can not relate the rating in tactics trainer to say 5 min games unless the pool of 5 min players all use the tactics trainer. 

Yes Chesstempo is a much better proposition. I use it, no time limit as such, and also a chance to move again if you pick an equally good move but not one the computer has designated - very good feature.

Chess.com take heed the time pressure is too much on the tactics trainer - not ideal for the slow but deep accurate thinker.