TCEC Season 9

[Hai]

Problem 1: TCEC Season 8 Superfinal 99,9% draw score already.
Problem 2: Only 378 viewer at the superfinal.

Season 9:
Solution 1: To decrase the draw score in the superfinal the engines will play the 20 possible opening moves (draw score ~50% to 75% depends on hardware and time control). + We will see new openings + openings created by engines.
And without tb (to decrease the draw score again and because the difference of the endgame knowledge between these engines is very big/ elo.)
Time control will be changed from 180' + 30" to 120' + 60" or 90' + 60".
-Don't forget next time Stockfish and Komodo will be like every time +~100 elo stronger.

Solution 2: To have more viewer, the next superfinal will be played with 4 engines and everyone vs everyone will play 40 games.
Stage 3 will have 8 engines and not 6. + more games.
Stage 2 will have 16 engines and not 12.

[Carlos777]

I'd like something like this:

Stage 1: 2 groups of 12 engines, double round robin. 10 ply book. 90' + 30". First 4 of each group qualify to the next stage.

Stage 2: 1 group of 8 engines hexa round robin. Same book and time control. First 2 qualify to the superfinal.

Stage 3: Superfinal, 100 games, 120' + 30" (IMO this is enough), 16 ply book of unbalanced openings, engines play the same openings with reversed colors.

[Graham Banks]

In my opinion it would be more popular if it wasn't overhyped so much.
In addition, I suspect that actively seeking money from viewers puts many off.
However, I guess this sort of thing comes with any commercial venture.

The chatroom could be tidied up a bit too - too much crap goes on in there.

Great chess regardless.

[Modern Times]

The superfinal should be chess960 if both engines support it

[jdart]

I don't think you can conclude from the number of live viewers that people are not interested. Some of us will look at the games later or offline.

The openings were already selected to be somewhat unbalanced. I think what you are seeing re draws is that the contestants are very evenly matched. And with depths of 35 ply or so they are not going to miss much.

--Jon

[Leto]

Right now there's over 900 people watching the TCEC 8 superfinal. I've seen it go over 1000 plenty of times. I guess it depends on the time of the day, as viewers tune in from many different countries.

[JJJ]

Maybe TC is too long ?

[fenchel]

Maybe TC is too long ?

The problem is 4 straight hours of moves eval'd to 0.0 by both engines, which seems to occur often.

[S.Taylor]

Time control will be changed from 180' + 30" to 120' + 60" or 90' + 60".

Why not go all the way, and play all games at blitz TC's?

Or even better, just use the engines of 5 years ago which were weaker than the new ones, and disqualify any engines from after 2010.
In fact, then, MORE engines can join in and have good chances!

[fenchel]

Time control will be changed from 180' + 30" to 120' + 60" or 90' + 60".

Why not go all the way, and play all games at blitz TC's?

Or even better, just use the engines of 5 years ago which were weaker than the new ones, and disqualify any engines from after 2010.
In fact, then, MORE engines can join in and have good chances!

no need to be so critical; it's a fact that there are often fewer viewers this round than during the last, and relatedly this will affect martin and everyone's else's motivation, which no one wants.

IMO there are many ways to adjust things to go around this. one is adjusting the adjudication rules. Another is that rather than spending a lot on top-of-the-line servers, instead have two nearly-of-the-line, and play the same opening with player colors reverse simultaneously. etc., the list goes on.

there's lots of experimentation going on these days with how chess is presented... I'm sure we all want more people to enjoy chess..