<Replace the word novelties with endgames and I would agree.

If each of us could research one winning endgame that had the same pawn structure and a few of the same pieces, then we could try to steer Rybka into one of those won endgames.

I think that would take a subset of the current voters who are going to vote for the best score as calculated by either Rybka or Fritz.>

"Let me emphasize again, we need to find a Novelty or in other words improve the best theoretical line in order to win."

We would need to convince influential members like Capafan,Kutztown for work assignments and most important we need to get our top analysts to see the benefits of Goal setting which is derived from backsliding potential winning endgame positions.

Rybka's strength could then be harvested for such a task. Its tactical ability will provide us the ways from arbitrary selected endgame position realistically reachable via currently used opening lines. This method applied by our members would quickly weed out irrelevant lines or rewrite known theory as to its true evaluations.

If we get acceptance then the work load needs to be distributed to avoid duplication where ever possible.

The findings of lines leading to proven winnable positions would then become our goal to aim for and branch out to search for deviations to our main line.

Yes "centi pawn" evaluations are important during the tactical prove reading of our lines. But moves should be fed to Rybka first to influence its evaluation procedure. Its learning function will then try to avoid long term loosing moves and adopt moves even when considered to be a short term reliability.

Such a method would elevate our playing strength to new heights and worthy to be part of it.

In summary: We must first start conceptualizing what we want to build as humans are capable of and then start looking for ways to get there with the use of Rybka and Co.

Comments?>

Yes, I think there is the danger that someone would suggest an analysis forum, you create it, and then it sits there and dies (i.e. nobody posts there or reads there).

So I agree that you don't want to make it so easy to start a new analysis forum that a mere suggestion, without commitment from the sugggester nor agreement from others, would suffice.

I think we are on the same page here. If a team member says "There should be a forum for line x", that's not good enough. The statement needs to be "There should be a forum for line x, and I am prepared to lead the analysis and report conclusions back to the team on the main page." I think that level of commitment is needed in order for the idea to succeed.

<Tabanus>

Changes to our forum system has been discussed many times before. But any changes just for the sake of it makes no sence at all.

I understand your desire to improve our playing strength via the avoidance of analysts duplication work and to have a condensed recommendation ready for our best move candidate at voting time.

But what we need is an attitude change of our perception that we can not win against best play and therefore we accept Rybka's lines produced via centi pawn evaluations.

We have come a long way from a three or four move search to the acceptance of 30 ply or 15 moves long line posting. Now we need more work to search for potential improvements to existing theoretical opening lines and transition to winnable endgame positions.

I am convinced that our analysts could find novelties in such lines by spending more time looking at replaying games and not just accepting an opening explorer verdict without prove reading.

For example if we play e4 next,we have to prepare also for the likely hood to face the Robatsch.

We need data of the most successful lines against it. Analyse those games with winning endings by back sliding to identify possible improvement for both sides. Also check games which have a negative score to see why and could we improve the lines.

Once we have data showing the winning percentage based on played games which must be verified first before adopting it as our main line. All recommended positive evaluated moves via backsliding done by Rybka and CO should give us our tree data or goals to aim for.

Such an undertaking is an immense task and thus require managing to avoid duplication but more important identifying and hopefully assigning the tasks for the various expected openings from our next challenger.

In this game we were caught unprepared by GMMU's moves several times. In the sense that they were analysed only after or a few days before they were actually played. BECAUSE we had a "main line" which caught too much attention from the analysts - relative to other lines. Thus after GMMU's off-beat move was played, main page popularity coincided with Rybka's top infinite analysis move.>

27.Nxd5 Nxd5 28.Ng5 Bxg2 29.Qc2 Be4 30.Nxe4 Qe5 31.h4 f6 32.Nc3 Nb4 33.Qb3+ Kh7 34.Qf7 Rc8 35.Qc4

If 35.Qc4** instead of 35.Qd7** there is a chance for black win..

35...b6 36.Be3 Qf5 37.Na2 Qb1+ 38.Kg2 Qxb2 39.Nxb4 Qxa3 40.Qe4+ Kh8 41.Qf5 Rf8 42.Nxc6 Qd6 43.Qxh5+ Kg8 44.Nd4 f5 45.Qf3 Qd7 46.Nb5 Bd8 47.Bd4 Qe6 48.Qb7 Rf7 49.Qb8 Qe8

[49...Qc6+ 50.Kg1 Rd7 51.e3 Qc1+ 52.Kg2 Qc4 53.Kg1 Qc6 54.Qe5 Qe4 55.Qb8 Kh7 56.Qc8 Qd5 57.Qb8 Qe6 58.Qc8 Be7 59.Kg2 Bc5 60.Qb8 Qg8 61.Qf4 Qd5+ 62.Kg1 Qe4 63.Qb8 Qb1+ 64.Kg2 Qb3 65.Qe8 Qf7 66.Qb8 Qd5+ 67.Kg1 Re7 68.Qf4 Qe4 69.Qxe4 Rxe4 70.f3 Re8 71.Kf2 Bxd4 72.exd4 Kg6 73.d5 Kf6 74.d6 Rc8 75.Ke3 g6 76.Kd4 Ke6 77.Na3 Kxd6 78.Nc4+ Ke6 79.Nxb6 Rc1 - ]

50.Kh2 Rf8 51.Qb7 Qf7 52.Qc6 f4 53.e4 fxg3+ 54.fxg3 Bxh4 55.Qd5 Qxd5 56.exd5 Bd8 57.Kg2 Rf7 58.d6 Rb7 59.Kf3 Kf7 60.Ke4 Ke6 61.Be5 Bf6 62.Bf4 Rb8 63.g4 Rh8 64.g5 Bb2 65.Nc7+ Kd7 66.Nd5 Re8+ 67.Kf5 b5 68.axb5 Rb8 69.b6 a4 70.Be3 a3 71.Nb4 Ra8 72.Na2 Kxd6 73.Ke4 Ra4+ 74.Kd3 Rg4 75.Ke2 Bd4 76.Kf3 Rh4 77.Bc1 Bc5 78.Bf4+ Kc6 79.Be5 Bd4 -

click for larger view

Analysis by Fritz 11: after <28.Ng5>

28...Bxg2 29.Qc2 Be4 30.Nxe4 Qe5 31.h4 f6 32.Nc3 Nb4 33.Qb3+ Kh7 34.Bxb4 axb4 35.Qxb4 Bd6 36.Qxb7 Rb8 37.Qxc6 Bxa3 38.bxa3 Rb2 = (-0.05) Depth: 29/51 11:11:22 49284mN

In order for us to achieve a very complex middlegame position in which theoretical novelties are possible, novelties that leave our opponent with a slightly inferior position, we need to have greater control over which opening is selected, as a result of considerable brainstorming that would actually precede the game.

The best way to achieve this control over, let's say the first 12 moves or so, is to have a steering committee that has veto power over some of the less than effective opening recommendations that have occurred as a result of the party atmosphere and circus mentality that always seem to accompany the start of our contests.

A case in point was the time 500 pages of kibitzing occurred as we argued the merits of <1.e4> over <1.d4>, and vice versa!

We would play at a much higher level if this type of nonsense could be avoided in future encounters. A steering committee could help ensure that, don't you think? Unfortunately, this idea of a steering committee was rejected out of hand as being undemocratic.

1) obviously it cannot be enforced. In fact there may be rebellion; members deliberately voting against the steering committee choice.

2)That the game may turn, in effect, into our best player (and computer) versus the GM (and computer) with the likely result.