hms123: <ronpaz1> Here's a quick example of how I would go about analyzing the line you asked about. I have truncated all of the lines to save space. <34.Qxc4 exf4 35.Bxf4 Rxd5 36.Rxd5 Rxd5 37.Be3 g6 38.Bd4 Bg7 39.Rf1 Bxd4+ 40.cxd4 Qe6 41.Rd1 Qe4 42.Qc8+ Kg7 43.Qc4 Rh5 44.Qc1 Qh4 45.Qc7 Rf5 46.Qg3 Qxg3 47.hxg3> After <40.cxd4>
(notice that <41….Qe4> is no longer in the top four moves)
 click for larger viewAnalysis by Fritz 10: 20 ply, deep position analysis
1. = (-0.02): 40...Kg7 41.Rd1 Qd6
2. = (-0.02): 40...Qd6 41.Qb3 Kg7
3. = (-0.02): 40...Kh7 41.Rd1 Qd6
4. = (0.00): 40...Rxd4 41.Rxf7 Rd1+
After <40…Qe6>
 click for larger viewAnalysis by Fritz 10: 21 ply, dpa
1. = (-0.01): 41.Qd3 Kg7 42.Rd1 Qd6
2. = (-0.01): 41.Rd1 Qe3+ 42.Kh1 Qe5
After <41.Rd1>, Fritz gives <41…Qe4 or Qd6 or Qe3+ or Kh8> all with <-0.01> evaluations at 20 ply. After <41…Qe4> Fritz gives <42.Qd3 or Qc8+ or Qc3 or h3> all with <-0.01> After <44…Qh4>, <45.Qc7 and h3> are the only choices Suddenly, after <45….Rf5>, Fritz finds <46.Qc2> more attractive than <46.Qg3>
 click for larger viewAnalysis by Fritz 10: 20 ply
1. = (-0.01): 46.Qc2 Qf6 47.h3 Qd6
2. = (-0.17): 46.Qc4 Qf2+ 47.Kh1 Rg5
3. (-0.34): 46.g3 Qh5 47.Re1 Qf3
4. (-0.54): 46.Qg3
Sliding forward on <46.Qc2 Qf6>, I find this:  click for larger viewAnalysis by Fritz 10: 19 ply (I would ordinarily let this run longer) 1. = (-0.01): 47.Qa2 Qb6 48.h3 Rh5
2. = (-0.01): 47.Qe4 Qb6 48.Rf1 Qe6
3. = (-0.01): 47.Qc4 Qb6 48.h3 Qb5
4. = (-0.01): 47.h3 Qd6 48.Qe4 Rd5
At this point, I would slide back down the line to see if there were other changes in the line. The main points are that there are lots of possible paths through the lines that are given by the engines, that the evaluations change as the hash tables fill up with positions, and that it is important to look at the board at each point to decide which of the branches hold the most promise. |