Discussion of anything and everything relating to chess playing software and machines.
Moderators: bob, hgm, Harvey Williamson
Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.

fastgm
 Posts: 409
 Joined: Mon Aug 19, 2013 4:57 pm

Contact:
Post
by fastgm » Mon Oct 21, 2019 6:47 pm
Brainfish after
45 hours and
depth 75 from the start position:
Brainfish 191015 64 POPCNT:
Code: Select all
67/170 8:10:57 427.471.609.979 14.511.274 +2,53 1.Qf6+ Bxf6 2.gxf6+ Kg8 3.Bd2 Qa5 4.Bh6 Qxc3 5.bxc3 b2 6.Bxf8 Rxf8 7.Rb1 Rb8 8.Ng5 Nxa3 9.Rxh7 Be8 10.Rxb2 Rxb2 11.Bg4 Nc2 12.Rg7+ Kh8 13.Bxe6 fxe6 14.Ra7 Rb8 15.Nxe6 Ne1 16.Ng5 Rb1 17.h4 Nf3+ 18.Kg2 Nxg5 19.hxg5 a3 20.Rxa3 Bf7 21.Ra8+ Kh7 22.Rf8 Be6 23.Re8 Bf7 24.Re7 Kg8 25.e6 Bxe6 26.Rxe6 Rc1 27.Re7 Kf8 28.Rg7 Rc2+ 29.Kg3 Rxc3+ 30.Kf4 Rc1 31.Ke5 c3 32.Rc7 Re1+ 33.Kxd5 Rc1 34.Kd6 Rd1 35.d5 Rc1 36.Ke6 Re1+ 37.Kd7 Rc1 38.d6 c2 39.Rc3
...
75/180 44:42:22 2.452.656.797.931 15.239.290 +6,05 1.Qf6+ Bxf6 2.gxf6+ Kg8 3.Bd2 Qa5 4.Bh6 Qxc3 5.bxc3 b2 6.Bxf8 b1Q 7.Rxb1 Rxb1+ 8.Kg2 h5 9.Bc5 Ne3+ 10.Kf2 Nd1+ 11.Bxd1 Rxd1 12.Rg3 Be8 13.Bb4 Bc6 14.Ke2 Ra1 15.h3 Rb1 16.Ke3 Rd1 17.Rg2 Rb1 18.Ng5 Be8 19.Rc2 Re1+ 20.Kf3 Rg1 21.Kf4 Rb1 22.Rf2 Kh8 23.Kg3 Kg8 24.Kh4 Bd7 25.Be7 Rb3 26.Rc2 Be8 27.Nf3 Rb7 28.Bb4 Kh7 29.Ng5+ Kh8 30.Rf2 Ra7 31.Bd6 Rb7 32.Be7 Kg8 33.Bb4 Rb6 34.Rc2 Rb7 35.Rc1 Bc6 36.Rf1 Rd7 37.Bd6 Rb7 38.Be7 Be8 39.Rc1 Rb2 40.Bb4 Bd7 41.Rf1 Be8 42.Kg3 Bd7 43.h4 Bc8 44.Kf3 Rc2 45.Rb1

Master Om
 Posts: 330
 Joined: Wed Nov 24, 2010 9:57 am
 Location: INDIA
Post
by Master Om » Mon Oct 21, 2019 7:14 pm
fastgm wrote: ↑Mon Oct 21, 2019 6:47 pm
Brainfish after
45 hours and
depth 75 from the start position:
Brainfish 191015 64 POPCNT:
Code: Select all
67/170 8:10:57 427.471.609.979 14.511.274 +2,53 1.Qf6+ Bxf6 2.gxf6+ Kg8 3.Bd2 Qa5 4.Bh6 Qxc3 5.bxc3 b2 6.Bxf8 Rxf8 7.Rb1 Rb8 8.Ng5 Nxa3 9.Rxh7 Be8 10.Rxb2 Rxb2 11.Bg4 Nc2 12.Rg7+ Kh8 13.Bxe6 fxe6 14.Ra7 Rb8 15.Nxe6 Ne1 16.Ng5 Rb1 17.h4 Nf3+ 18.Kg2 Nxg5 19.hxg5 a3 20.Rxa3 Bf7 21.Ra8+ Kh7 22.Rf8 Be6 23.Re8 Bf7 24.Re7 Kg8 25.e6 Bxe6 26.Rxe6 Rc1 27.Re7 Kf8 28.Rg7 Rc2+ 29.Kg3 Rxc3+ 30.Kf4 Rc1 31.Ke5 c3 32.Rc7 Re1+ 33.Kxd5 Rc1 34.Kd6 Rd1 35.d5 Rc1 36.Ke6 Re1+ 37.Kd7 Rc1 38.d6 c2 39.Rc3
...
75/180 44:42:22 2.452.656.797.931 15.239.290 +6,05 1.Qf6+ Bxf6 2.gxf6+ Kg8 3.Bd2 Qa5 4.Bh6 Qxc3 5.bxc3 b2 6.Bxf8 b1Q 7.Rxb1 Rxb1+ 8.Kg2 h5 9.Bc5 Ne3+ 10.Kf2 Nd1+ 11.Bxd1 Rxd1 12.Rg3 Be8 13.Bb4 Bc6 14.Ke2 Ra1 15.h3 Rb1 16.Ke3 Rd1 17.Rg2 Rb1 18.Ng5 Be8 19.Rc2 Re1+ 20.Kf3 Rg1 21.Kf4 Rb1 22.Rf2 Kh8 23.Kg3 Kg8 24.Kh4 Bd7 25.Be7 Rb3 26.Rc2 Be8 27.Nf3 Rb7 28.Bb4 Kh7 29.Ng5+ Kh8 30.Rf2 Ra7 31.Bd6 Rb7 32.Be7 Kg8 33.Bb4 Rb6 34.Rc2 Rb7 35.Rc1 Bc6 36.Rf1 Rd7 37.Bd6 Rb7 38.Be7 Be8 39.Rc1 Rb2 40.Bb4 Bd7 41.Rf1 Be8 42.Kg3 Bd7 43.h4 Bc8 44.Kf3 Rc2 45.Rb1
This is nice. It takes time for SF to know Ng5 is draw and Bd2 wins instantly. There is No answer to Bh6!!.
Always Expect the Unexpected

Spliffjiffer
 Posts: 210
 Joined: Thu Aug 02, 2012 5:48 pm
 Location: Germany
Post
by Spliffjiffer » Mon Oct 21, 2019 7:16 pm
fastgm wrote: ↑Mon Oct 21, 2019 6:47 pm
Brainfish after
45 hours and
depth 75 from the start position:
Brainfish 191015 64 POPCNT:
Code: Select all
67/170 8:10:57 427.471.609.979 14.511.274 +2,53 1.Qf6+ Bxf6 2.gxf6+ Kg8 3.Bd2 Qa5 4.Bh6 Qxc3 5.bxc3 b2 6.Bxf8 Rxf8 7.Rb1 Rb8 8.Ng5 Nxa3 9.Rxh7 Be8 10.Rxb2 Rxb2 11.Bg4 Nc2 12.Rg7+ Kh8 13.Bxe6 fxe6 14.Ra7 Rb8 15.Nxe6 Ne1 16.Ng5 Rb1 17.h4 Nf3+ 18.Kg2 Nxg5 19.hxg5 a3 20.Rxa3 Bf7 21.Ra8+ Kh7 22.Rf8 Be6 23.Re8 Bf7 24.Re7 Kg8 25.e6 Bxe6 26.Rxe6 Rc1 27.Re7 Kf8 28.Rg7 Rc2+ 29.Kg3 Rxc3+ 30.Kf4 Rc1 31.Ke5 c3 32.Rc7 Re1+ 33.Kxd5 Rc1 34.Kd6 Rd1 35.d5 Rc1 36.Ke6 Re1+ 37.Kd7 Rc1 38.d6 c2 39.Rc3
...
75/180 44:42:22 2.452.656.797.931 15.239.290 +6,05 1.Qf6+ Bxf6 2.gxf6+ Kg8 3.Bd2 Qa5 4.Bh6 Qxc3 5.bxc3 b2 6.Bxf8 b1Q 7.Rxb1 Rxb1+ 8.Kg2 h5 9.Bc5 Ne3+ 10.Kf2 Nd1+ 11.Bxd1 Rxd1 12.Rg3 Be8 13.Bb4 Bc6 14.Ke2 Ra1 15.h3 Rb1 16.Ke3 Rd1 17.Rg2 Rb1 18.Ng5 Be8 19.Rc2 Re1+ 20.Kf3 Rg1 21.Kf4 Rb1 22.Rf2 Kh8 23.Kg3 Kg8 24.Kh4 Bd7 25.Be7 Rb3 26.Rc2 Be8 27.Nf3 Rb7 28.Bb4 Kh7 29.Ng5+ Kh8 30.Rf2 Ra7 31.Bd6 Rb7 32.Be7 Kg8 33.Bb4 Rb6 34.Rc2 Rb7 35.Rc1 Bc6 36.Rf1 Rd7 37.Bd6 Rb7 38.Be7 Be8 39.Rc1 Rb2 40.Bb4 Bd7 41.Rf1 Be8 42.Kg3 Bd7 43.h4 Bc8 44.Kf3 Rc2 45.Rb1
thank you fastgm !
this score seems devestating and gives a new "best line" as well...impressive analysis, thank you
...this score is much higher than that i was ever analysing...from my pov this should most probably gives us the conclusion that Lyudmil Tsvetkov's assumption that 1. Qf6 (sac) was correct and HIS move wins !!?...we can call now this position as a "testposition of first class" imo...thx again awesome analysis !
Wahrheiten sind Illusionen von denen wir aber vergessen haben dass sie welche sind.

Master Om
 Posts: 330
 Joined: Wed Nov 24, 2010 9:57 am
 Location: INDIA
Post
by Master Om » Mon Oct 21, 2019 7:46 pm
Spliffjiffer wrote: ↑Mon Oct 21, 2019 7:16 pm
fastgm wrote: ↑Mon Oct 21, 2019 6:47 pm
Brainfish after
45 hours and
depth 75 from the start position:
Brainfish 191015 64 POPCNT:
Code: Select all
67/170 8:10:57 427.471.609.979 14.511.274 +2,53 1.Qf6+ Bxf6 2.gxf6+ Kg8 3.Bd2 Qa5 4.Bh6 Qxc3 5.bxc3 b2 6.Bxf8 Rxf8 7.Rb1 Rb8 8.Ng5 Nxa3 9.Rxh7 Be8 10.Rxb2 Rxb2 11.Bg4 Nc2 12.Rg7+ Kh8 13.Bxe6 fxe6 14.Ra7 Rb8 15.Nxe6 Ne1 16.Ng5 Rb1 17.h4 Nf3+ 18.Kg2 Nxg5 19.hxg5 a3 20.Rxa3 Bf7 21.Ra8+ Kh7 22.Rf8 Be6 23.Re8 Bf7 24.Re7 Kg8 25.e6 Bxe6 26.Rxe6 Rc1 27.Re7 Kf8 28.Rg7 Rc2+ 29.Kg3 Rxc3+ 30.Kf4 Rc1 31.Ke5 c3 32.Rc7 Re1+ 33.Kxd5 Rc1 34.Kd6 Rd1 35.d5 Rc1 36.Ke6 Re1+ 37.Kd7 Rc1 38.d6 c2 39.Rc3
...
75/180 44:42:22 2.452.656.797.931 15.239.290 +6,05 1.Qf6+ Bxf6 2.gxf6+ Kg8 3.Bd2 Qa5 4.Bh6 Qxc3 5.bxc3 b2 6.Bxf8 b1Q 7.Rxb1 Rxb1+ 8.Kg2 h5 9.Bc5 Ne3+ 10.Kf2 Nd1+ 11.Bxd1 Rxd1 12.Rg3 Be8 13.Bb4 Bc6 14.Ke2 Ra1 15.h3 Rb1 16.Ke3 Rd1 17.Rg2 Rb1 18.Ng5 Be8 19.Rc2 Re1+ 20.Kf3 Rg1 21.Kf4 Rb1 22.Rf2 Kh8 23.Kg3 Kg8 24.Kh4 Bd7 25.Be7 Rb3 26.Rc2 Be8 27.Nf3 Rb7 28.Bb4 Kh7 29.Ng5+ Kh8 30.Rf2 Ra7 31.Bd6 Rb7 32.Be7 Kg8 33.Bb4 Rb6 34.Rc2 Rb7 35.Rc1 Bc6 36.Rf1 Rd7 37.Bd6 Rb7 38.Be7 Be8 39.Rc1 Rb2 40.Bb4 Bd7 41.Rf1 Be8 42.Kg3 Bd7 43.h4 Bc8 44.Kf3 Rc2 45.Rb1
thank you fastgm !
this score seems devestating and gives a new "best line" as well...impressive analysis, thank you
...this score is much higher than that i was ever analysing...from my pov this should most probably gives us the conclusion that Lyudmil Tsvetkov's assumption that 1. Qf6 (sac) was correct and HIS move wins !!?...we can call now this position as a "testposition of first class" imo...thx again awesome analysis !
I have completely done the analysis . Will post it tomorrow.
Always Expect the Unexpected

Spliffjiffer
 Posts: 210
 Joined: Thu Aug 02, 2012 5:48 pm
 Location: Germany
Post
by Spliffjiffer » Mon Oct 21, 2019 7:53 pm
Master Om wrote: ↑Mon Oct 21, 2019 7:46 pm
Spliffjiffer wrote: ↑Mon Oct 21, 2019 7:16 pm
fastgm wrote: ↑Mon Oct 21, 2019 6:47 pm
Brainfish after
45 hours and
depth 75 from the start position:
Brainfish 191015 64 POPCNT:
Code: Select all
67/170 8:10:57 427.471.609.979 14.511.274 +2,53 1.Qf6+ Bxf6 2.gxf6+ Kg8 3.Bd2 Qa5 4.Bh6 Qxc3 5.bxc3 b2 6.Bxf8 Rxf8 7.Rb1 Rb8 8.Ng5 Nxa3 9.Rxh7 Be8 10.Rxb2 Rxb2 11.Bg4 Nc2 12.Rg7+ Kh8 13.Bxe6 fxe6 14.Ra7 Rb8 15.Nxe6 Ne1 16.Ng5 Rb1 17.h4 Nf3+ 18.Kg2 Nxg5 19.hxg5 a3 20.Rxa3 Bf7 21.Ra8+ Kh7 22.Rf8 Be6 23.Re8 Bf7 24.Re7 Kg8 25.e6 Bxe6 26.Rxe6 Rc1 27.Re7 Kf8 28.Rg7 Rc2+ 29.Kg3 Rxc3+ 30.Kf4 Rc1 31.Ke5 c3 32.Rc7 Re1+ 33.Kxd5 Rc1 34.Kd6 Rd1 35.d5 Rc1 36.Ke6 Re1+ 37.Kd7 Rc1 38.d6 c2 39.Rc3
...
75/180 44:42:22 2.452.656.797.931 15.239.290 +6,05 1.Qf6+ Bxf6 2.gxf6+ Kg8 3.Bd2 Qa5 4.Bh6 Qxc3 5.bxc3 b2 6.Bxf8 b1Q 7.Rxb1 Rxb1+ 8.Kg2 h5 9.Bc5 Ne3+ 10.Kf2 Nd1+ 11.Bxd1 Rxd1 12.Rg3 Be8 13.Bb4 Bc6 14.Ke2 Ra1 15.h3 Rb1 16.Ke3 Rd1 17.Rg2 Rb1 18.Ng5 Be8 19.Rc2 Re1+ 20.Kf3 Rg1 21.Kf4 Rb1 22.Rf2 Kh8 23.Kg3 Kg8 24.Kh4 Bd7 25.Be7 Rb3 26.Rc2 Be8 27.Nf3 Rb7 28.Bb4 Kh7 29.Ng5+ Kh8 30.Rf2 Ra7 31.Bd6 Rb7 32.Be7 Kg8 33.Bb4 Rb6 34.Rc2 Rb7 35.Rc1 Bc6 36.Rf1 Rd7 37.Bd6 Rb7 38.Be7 Be8 39.Rc1 Rb2 40.Bb4 Bd7 41.Rf1 Be8 42.Kg3 Bd7 43.h4 Bc8 44.Kf3 Rc2 45.Rb1
thank you fastgm !
this score seems devestating and gives a new "best line" as well...impressive analysis, thank you
...this score is much higher than that i was ever analysing...from my pov this should most probably gives us the conclusion that Lyudmil Tsvetkov's assumption that 1. Qf6 (sac) was correct and HIS move wins !!?...we can call now this position as a "testposition of first class" imo...thx again awesome analysis !
I have completely done the analysis . Will post it tomorrow.
you guys are awesome, thank you Louis, Peter, MasterOm, fastgm and all others...finally we got a conclusion...cant wait for your analysis MasterOm...we had a long way to go
...great stuff !
Wahrheiten sind Illusionen von denen wir aber vergessen haben dass sie welche sind.

jp
 Posts: 815
 Joined: Mon Apr 23, 2018 5:54 am
Post
by jp » Tue Oct 22, 2019 3:55 am
Spliffjiffer wrote: ↑Mon Oct 21, 2019 7:16 pm
thank you fastgm !
this score seems devestating and gives a new "best line" as well...impressive analysis, thank you
...this score is much higher than that i was ever analysing...from my pov this should most probably gives us the conclusion that Lyudmil Tsvetkov's assumption that 1. Qf6 (sac) was correct and HIS move wins !!?...we can call now this position as a "testposition of first class" imo...thx again awesome analysis !
But, Spliffjiffer, it looks like the hard part is further down the line, i.e. not 1. Qf6+ ("easy"
) but 3. Bd2 (and maybe later moves too). I think most reasonably good chessplayers would play 3. Ng5.

Spliffjiffer
 Posts: 210
 Joined: Thu Aug 02, 2012 5:48 pm
 Location: Germany
Post
by Spliffjiffer » Tue Oct 22, 2019 6:06 am
well its probably a matter of taste wether one prefers Ng5 or Bd2 instead...as im not a good player my interpretation is not worth much but for my taste Bd2 seems natural because of 3 reasons: 1./2. it avoids simplifications by securing the bishop from trading and opens the way for the rook to the kingside, 3. it is positioning the bishop to exactely there where it belongs to imo, namely the structural problem of the f7/g6/h7 pawn formation which weakens the black squares with the aim to get to g7 and mating ideas...on the other hand Ng5 seems most aggressive and is seeking the immediate knockout which seems to work nearly...so from my pov Bd2 seems to be the positional approach and Ng5 the tactical one filled with pure aggression...which approach works best though is just a matter of precise calculation...my own impression was and is that Bd2 seems more natural (to me personally)
btw: that Lyudmil is the aggressive type of player he is showing not least in an other brilliant assessment in his attempt to claim a win in this position (no diagram, offtopic):
rnb2rk1/1pq1bppp/p3p3/2npP1P1/3N1P2/2NBB3/PPP4P/R2QK2R w KQ  0 1
1.Bxh7 !!?, Kxh7 2. Qh5+, Kg8 3. Rg1?!...thats the pure aggression i mean, chesswise, and ironically the 3rd move again
...but his initial assumption is correct again according to my analysis !?
Wahrheiten sind Illusionen von denen wir aber vergessen haben dass sie welche sind.

jp
 Posts: 815
 Joined: Mon Apr 23, 2018 5:54 am
Post
by jp » Tue Oct 22, 2019 9:05 am
It's possible that my guess about humans' preference was biased by knowing that Lc0 & SF choose Ng5 first and stick with it for a long time.

Vinvin
 Posts: 4373
 Joined: Thu Mar 09, 2006 8:40 am
 Full name: Vincent Lejeune
Post
by Vinvin » Tue Oct 22, 2019 11:51 am
Master Om wrote: ↑Mon Oct 21, 2019 7:14 pm
fastgm wrote: ↑Mon Oct 21, 2019 6:47 pm
Brainfish after
45 hours and
depth 75 from the start position:
1r3r2/4bpkp/1qb1p1p1/3pP1P1/p1pP1Q2/PpP2N1R/1Pn1B2P/3RB2K w   0 1
Brainfish 191015 64 POPCNT:
Code: Select all
67/170 8:10:57 427.471.609.979 14.511.274 +2,53 1.Qf6+ Bxf6 2.gxf6+ Kg8 3.Bd2 Qa5 4.Bh6 Qxc3 5.bxc3 b2 6.Bxf8 Rxf8 7.Rb1 Rb8 8.Ng5 Nxa3 9.Rxh7 Be8 10.Rxb2 Rxb2 11.Bg4 Nc2 12.Rg7+ Kh8 13.Bxe6 fxe6 14.Ra7 Rb8 15.Nxe6 Ne1 16.Ng5 Rb1 17.h4 Nf3+ 18.Kg2 Nxg5 19.hxg5 a3 20.Rxa3 Bf7 21.Ra8+ Kh7 22.Rf8 Be6 23.Re8 Bf7 24.Re7 Kg8 25.e6 Bxe6 26.Rxe6 Rc1 27.Re7 Kf8 28.Rg7 Rc2+ 29.Kg3 Rxc3+ 30.Kf4 Rc1 31.Ke5 c3 32.Rc7 Re1+ 33.Kxd5 Rc1 34.Kd6 Rd1 35.d5 Rc1 36.Ke6 Re1+ 37.Kd7 Rc1 38.d6 c2 39.Rc3
...
75/180 44:42:22 2.452.656.797.931 15.239.290 +6,05 1.Qf6+ Bxf6 2.gxf6+ Kg8 3.Bd2 Qa5 4.Bh6 Qxc3 5.bxc3 b2 6.Bxf8 b1Q 7.Rxb1 Rxb1+ 8.Kg2 h5 9.Bc5 Ne3+ 10.Kf2 Nd1+ 11.Bxd1 Rxd1 12.Rg3 Be8 13.Bb4 Bc6 14.Ke2 Ra1 15.h3 Rb1 16.Ke3 Rd1 17.Rg2 Rb1 18.Ng5 Be8 19.Rc2 Re1+ 20.Kf3 Rg1 21.Kf4 Rb1 22.Rf2 Kh8 23.Kg3 Kg8 24.Kh4 Bd7 25.Be7 Rb3 26.Rc2 Be8 27.Nf3 Rb7 28.Bb4 Kh7 29.Ng5+ Kh8 30.Rf2 Ra7 31.Bd6 Rb7 32.Be7 Kg8 33.Bb4 Rb6 34.Rc2 Rb7 35.Rc1 Bc6 36.Rf1 Rd7 37.Bd6 Rb7 38.Be7 Be8 39.Rc1 Rb2 40.Bb4 Bd7 41.Rf1 Be8 42.Kg3 Bd7 43.h4 Bc8 44.Kf3 Rc2 45.Rb1
This is nice. It takes time for SF to know Ng5 is draw and Bd2 wins instantly. There is No answer to Bh6!!.
Maybe the position before 3.Bd2 deserve a test position for itself.
That's recurrent problem with Stockfish : when it has a bad position it often satisfies with a drawing move even when there's a winning move. Another version of this problem : when SF finds a move with eval "+3" it has trouble to find a "+10" move.
Engines can find 1.Qf6+ only to make a draw. That's not bad already.
I'll test deeper this evening : 1r3rk1/5p1p/1qb1pPp1/3pP3/p1pP4/PpP2N1R/1Pn1B2P/3RB2K w   1 3.
Is Bd2 the clear best move here ?

Master Om
 Posts: 330
 Joined: Wed Nov 24, 2010 9:57 am
 Location: INDIA
Post
by Master Om » Tue Oct 22, 2019 2:57 pm
Vinvin wrote: ↑Tue Oct 22, 2019 11:51 am
Master Om wrote: ↑Mon Oct 21, 2019 7:14 pm
fastgm wrote: ↑Mon Oct 21, 2019 6:47 pm
Brainfish after
45 hours and
depth 75 from the start position:
1r3r2/4bpkp/1qb1p1p1/3pP1P1/p1pP1Q2/PpP2N1R/1Pn1B2P/3RB2K w   0 1
Brainfish 191015 64 POPCNT:
Code: Select all
67/170 8:10:57 427.471.609.979 14.511.274 +2,53 1.Qf6+ Bxf6 2.gxf6+ Kg8 3.Bd2 Qa5 4.Bh6 Qxc3 5.bxc3 b2 6.Bxf8 Rxf8 7.Rb1 Rb8 8.Ng5 Nxa3 9.Rxh7 Be8 10.Rxb2 Rxb2 11.Bg4 Nc2 12.Rg7+ Kh8 13.Bxe6 fxe6 14.Ra7 Rb8 15.Nxe6 Ne1 16.Ng5 Rb1 17.h4 Nf3+ 18.Kg2 Nxg5 19.hxg5 a3 20.Rxa3 Bf7 21.Ra8+ Kh7 22.Rf8 Be6 23.Re8 Bf7 24.Re7 Kg8 25.e6 Bxe6 26.Rxe6 Rc1 27.Re7 Kf8 28.Rg7 Rc2+ 29.Kg3 Rxc3+ 30.Kf4 Rc1 31.Ke5 c3 32.Rc7 Re1+ 33.Kxd5 Rc1 34.Kd6 Rd1 35.d5 Rc1 36.Ke6 Re1+ 37.Kd7 Rc1 38.d6 c2 39.Rc3
...
75/180 44:42:22 2.452.656.797.931 15.239.290 +6,05 1.Qf6+ Bxf6 2.gxf6+ Kg8 3.Bd2 Qa5 4.Bh6 Qxc3 5.bxc3 b2 6.Bxf8 b1Q 7.Rxb1 Rxb1+ 8.Kg2 h5 9.Bc5 Ne3+ 10.Kf2 Nd1+ 11.Bxd1 Rxd1 12.Rg3 Be8 13.Bb4 Bc6 14.Ke2 Ra1 15.h3 Rb1 16.Ke3 Rd1 17.Rg2 Rb1 18.Ng5 Be8 19.Rc2 Re1+ 20.Kf3 Rg1 21.Kf4 Rb1 22.Rf2 Kh8 23.Kg3 Kg8 24.Kh4 Bd7 25.Be7 Rb3 26.Rc2 Be8 27.Nf3 Rb7 28.Bb4 Kh7 29.Ng5+ Kh8 30.Rf2 Ra7 31.Bd6 Rb7 32.Be7 Kg8 33.Bb4 Rb6 34.Rc2 Rb7 35.Rc1 Bc6 36.Rf1 Rd7 37.Bd6 Rb7 38.Be7 Be8 39.Rc1 Rb2 40.Bb4 Bd7 41.Rf1 Be8 42.Kg3 Bd7 43.h4 Bc8 44.Kf3 Rc2 45.Rb1
This is nice. It takes time for SF to know Ng5 is draw and Bd2 wins instantly. There is No answer to Bh6!!.
Maybe the position before 3.Bd2 deserve a test position for itself.
That's recurrent problem with Stockfish : when it has a bad position it often satisfies with a drawing move even when there's a winning move. Another version of this problem : when SF finds a move with eval "+3" it has trouble to find a "+10" move.
Engines can find 1.Qf6+ only to make a draw. That's not bad already.
I'll test deeper this evening : 1r3rk1/5p1p/1qb1pPp1/3pP3/p1pP4/PpP2N1R/1Pn1B2P/3RB2K w   1 3.
Is Bd2 the clear best move here ?
Yes it is. All moves loses to Bh6!!
Always Expect the Unexpected