I beat Telepath

[Dirt]

I was actually taking that into account.

Ok, I wasn't sure of what the correct numbers were.

[Uri Blass]

I had Telepath play without pondering and I always played the move from its PV.

I think you were quite lucky. Weren't you playing at a disadvantage of one ply?

yes, but that is less than 100 Elo. so you should actually win about 1.5 games out of every 4...

scoring 1.5 out of 4 is clearly less than winning 1.5 games out of 4 because there are draws.

Uri

[bob]

I had Telepath play without pondering and I always played the move from its PV.

I think you were quite lucky. Weren't you playing at a disadvantage of one ply?

yes, but that is less than 100 Elo. so you should actually win about 1.5 games out of every 4...

scoring 1.5 out of 4 is clearly less than winning 1.5 games out of 4 because there are draws.

Uri

What are you talking about?

1.5 out of 4 could be 3 draws or one win and one draw, it makes absolutely no difference in Elo calculations. If you are +200 better than your opponent, you should only lose about one of every 4 games, but you could draw 2 out of four instead...

[Uri Blass]

I had Telepath play without pondering and I always played the move from its PV.

I think you were quite lucky. Weren't you playing at a disadvantage of one ply?

yes, but that is less than 100 Elo. so you should actually win about 1.5 games out of every 4...

scoring 1.5 out of 4 is clearly less than winning 1.5 games out of 4 because there are draws.

Uri

What are you talking about?

1.5 out of 4 could be 3 draws or one win and one draw, it makes absolutely no difference in Elo calculations. If you are +200 better than your opponent, you should only lose about one of every 4 games, but you could draw 2 out of four instead...

I say that if you can expect 1.5 out of 4 then you can expect to have less wins than 1.5 wins out of 4.

It can be 1 win out of 4 and 1 draw out of 4 and it can be 3 draws out of 4
and 0 wins out of 4.

In other words I disagree with the following sentence of you

"so you should actually win about 1.5 games out of every 4"

You can actually expect 1.5 points out of 4 points but it is different than winning 1.5 games out of 4 or having win probability of 0.375

Uri

[bob]

I had Telepath play without pondering and I always played the move from its PV.

I think you were quite lucky. Weren't you playing at a disadvantage of one ply?

yes, but that is less than 100 Elo. so you should actually win about 1.5 games out of every 4...

scoring 1.5 out of 4 is clearly less than winning 1.5 games out of 4 because there are draws.

Uri

What are you talking about?

1.5 out of 4 could be 3 draws or one win and one draw, it makes absolutely no difference in Elo calculations. If you are +200 better than your opponent, you should only lose about one of every 4 games, but you could draw 2 out of four instead...

I say that if you can expect 1.5 out of 4 then you can expect to have less wins than 1.5 wins out of 4.

It can be 1 win out of 4 and 1 draw out of 4 and it can be 3 draws out of 4
and 0 wins out of 4.

In other words I disagree with the following sentence of you

"so you should actually win about 1.5 games out of every 4"

You can actually expect 1.5 points out of 4 points but it is different than winning 1.5 games out of 4 or having win probability of 0.375

Uri

This is only different for someone too anal to read. First, how can you win .5 games? I think the meaning was perfectly clear..

[CRoberson]

I tried the experiment again (on my laptop - slow). This time
I played Rybka 2.3.2a and let it ponder. I used the Chessbase
interface that came with Deep Fritz 8.

Played only one game and it ended in a draw. This fits the stats.

However, I can honestly say I ran the same experiments 20 years
ago using dedicated units and never achieved the results we have
been predicting here.

Why?
1) A ply isn't worth as much today as it was back then. (1ply = 200 Elo 20 yrs ago and 100 today.)
2) 20 years ago, the depths were much shallower thus the games were more impacted by the horizon effect.