dzindzhi used to give ten minutes to one odds and a minor piece to people rated 2000 in the skittles room

he won about 90% of those games

he also used to playy at equal odds against 2000 rating strength and would tell player: you name the square and i will mate you on that square .... success rate over 50%

world opens in the 1980's at the adams mark hotel in philadelphia... many witnesses 

due to the older k factors, the difference between 2000 and 2200 (200 points) is not the same as the differesnce between 2400 and 2600

as we go up the scale, rating points are factored differently and a 100 point gap at the top (2700) is significantly greater than it is at the expert level 

Even a toddler knows that.

oh the reason Kasparov was down a piece is because it is all in his calculations!lol

No - this is wrong. 

K factor means that you need more games, if you are worth 2600, to get there from 2400. 

However, the percentages of success are calculated the same way, 2000 to 2200, or 2400 to 2600. This means that winning or losing, will be rewarded or punished at the same rate. 

If I'm 2400, but worth 2600, it will take me LONGER to get there, but it's just as certain that I do, than if I'm 2000 and worth 2200. The difference in CLASS, or winning percentage, is exactly exactly exactly the same. 

About the other question, when I was playing at 1850-1900 FIDE strength, I had a few friends who were 2500+ GMs, and were pretty regular sparring partners. 

One of them gave me 5:1 odds, and beat me 14:1 (!) I beat him the first game, strategically and tactically - and thereafter he adjusted his play, and it was all hopeless, game after game after game after game. 

The other, with the same odds, only managed to beat me about 6:4

A third one (who was a candidate to the world championships in the eighties), giving me the same time odds as the other two, and no material advantage, lost 3 times more games than he won against me. Then he started giving me 3:1 minute time odds, and then the trend reversed. 

So this is highly individual also among 2500+ GMs. Maybe Dzindzi has a special gift for speed playing. Don't underestimate the 2000 player! :-)

solskytz, he probably meant that it's harder to improve your game the more you go higher up the scale, e.g. every rating point will cost you more and more time and effort.

It's just the law of diminishing returns. A beginner can easily improve hundreds of points in a matter of months, but then it gets progressively harder.

For example I think getting from 1800 to 2000 is harder than getting from wherever you start to 1800.

You can prove it by yourself.
Set up a position without a Knight (for example) against Houdini and if you can beat it, maybe you can win against Kaspárov too.

I´ve just watched your game.

Nice! you played very solid.

I'm just posting the game from post #164:

I actually played out that final position in the game above me with 1 minute on the clock and its not exactly easy to play, it actually makes good endgame pratice against houdini because i did win one but also messed it up as well.

 Its just enough of an advantage but white has enough counter play to force you to be very precise.