We double in this patch the weight of the capture history table in the
local scoring of captures for move ordering.
The capture history table is indexed by the triplet (capturing piece,
capture square, captured piece) and gets information like "it seems to
have been historically good in that part of the search tree to capture
a pawn with a rook on g3, even if it seems to lose material", and affect
the normaly pure « Most Valuable Victim » ordering of captures.
Finished yellow at STC after 228842 games (posting a +1.36 Elo gain):
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 228842 W: 50894 L: 50152 D: 127796
http://tests.stockfishchess.org/tests/view/5b714bb00ebc5902bdba332d
Alain SAVARD [Tue, 14 Aug 2018 06:31:56 +0000 (08:31 +0200)]
Remove pawncount array in imbalance
This is a natural follow up to last commit where values on the
QuadraticOurs diagonal and some piece value deltas were changed.
@Stefano80 tried to simplify the newly introduced pawncount array
using QuadraticOurs[1][1] =52 and a -30 adjustment on pawn values
So I started a 80000 30+0.3 SPSA on the QuadraticOurs diagonal and
on the piece values using @Stefano80 start values.
SPSA gave the new values QuadraticOurs[1][1] =38 and a -33 on pawn
values (the other changes on QuadraticOurs were kept, but were not
ignificant according to this test
http://tests.stockfishchess.org/tests/view/5b710ccb0ebc5902bdba2f27)
GuardianRM [Sun, 12 Aug 2018 16:40:03 +0000 (18:40 +0200)]
Non-linear bonus for pawn count
This patch introduces a non-linear bonus for pawns, along with some
(linear) corrections for the other pieces types.
The original values were obtained by a massive non-linear tuning of both
pawns and other pieces by GuardianRM, while Alain Savard and Chris Cain
later simplified the patch by observing that, apart from the pawn case, the
tuned corrections were in fact almost affine and could be incorporated in
our current code base via the piece values in types.h (offset) and the diagonal
of the quadratic matrix (slope). See discussion in PR#1725 :
https://github.com/official-stockfish/Stockfish/pull/1725
How to continue from there?
- Maybe the non-linearity for the pawn value could be somewhat tempered
again and a simpler linear correction for pawns would work?
Original patches:
- [Piece values](http://tests.stockfishchess.org/tests/view/5b6d2cc00ebc5902bdba02d5) by Stefano Cardanobile
- [Stat bonus](http://tests.stockfishchess.org/tests/view/5b6adbc90ebc5902bdb9da73) by Stefan Geschwentner
- [Rook on pawn](http://tests.stockfishchess.org/tests/view/5b62a95b0ebc5902bdb961c0) by Mark Tenzer
- [Hanging bonus](http://tests.stockfishchess.org/tests/view/5b5d2fa00ebc5902bdb90855) by Ivan Ilvec
- [ss tweak](http://tests.stockfishchess.org/tests/view/5b58b7240ebc5902bdb89025) by miguel-l
Simple razoring: depth 1 only, no distinction between PV / NonPV
We simplify the razoring logic by applying it to all nodes at depth 1 only.
An added advantage is that only one razor margin is needed now, and we treat
PV and Non-PV nodes in the same manner.
How to continue?
- There may be some conditions in which depth 2 razoring is beneficial.
- We can see whether the razor margin can be tuned, perhaps even with a
different value for PV nodes.
- Perhaps we can unify the treatment of PV and Non-PV nodes in other parts
of the search as well.
We are opting for this patch as the more logical and simple of the two,
and it appears to be no less strong. Thanks in particular to @DU-jdto
for input into this patch.
Miguel Lahoz [Thu, 9 Aug 2018 14:04:36 +0000 (22:04 +0800)]
Remove Condition For Passed Pawns
Currently, we do not consider pawns passed if there is another pawn of
the same color in front of them. It appears that this condition is not
necessary. The idea is that the doubled pawns are likely to be weak and
one of them will be likely captured anyway. On the other hand, if we do
somehow manage to promote a pawn, then the pawn behind it becomes passed
as well. In any case, the end result is we end up with an extra
potentially passed pawn. The current evaluation for passed pawns already
handles this case by also scaling down this effect.
Unify the "quiet" and "non-quiet" reduction rules for use at any kind of moves.
The idea behind it was that both rules reduce at similiar cases in master:
one directly for late previous moves and the other indirectly by using a
bad stat score which is used for most move sorting and so approximates the
late move condition.
For captures/promotions the old rule was triggered in 25% but the new
rule only for 3% of all cases (so now more reductions are done, whereas
for quiet moves reductions keep the same level).
Currently, we first calculate some bitboards at the top of Evaluation::space()
and then check whether we actually need them. Invert the ordering. Of course this
does not make a difference in current master because the constexpr bitboard
calculations are in fact done at compile time by any decent compiler, but I find
my version a bit healthier since it will always meet or exceed current implementation
even if we eventually change the spaceMask to something not contsexpr.
FauziAkram [Wed, 8 Aug 2018 15:48:06 +0000 (17:48 +0200)]
King Psqt Tuning
After a session of tuning for King Psqt I got some new values, which was later
tweaked manually by me Fauzi, to result in an Elo-gain patch which seems to scale
pretty well:
Introduce voting system for best move selction in multi-threads mode.
Joint work with Stefan Geschwentner, based on ideas introduced by
Michael Stembera.
Moves are upvoted by every thread using the margin to the minimum score
across threads and the completed depth.
First thread voting for the winner move is selected as best thread.
Passed STC, LTC. A further LTC test with only 4 threads failed with positive
score. A LTC with 31 threads was stopped with LLR 0.77 after 25k games to
avoid use of excessive resources (equivalent to 1.5M STC games).
Similar ideas were proposed by Michael Stembera 2 years ago #507, #508.
This implementation seems simpler and more understandable, the results
slightly more promising.
Further possible work:
1) Tweak of the formula using for assigning votes.
2) Use a different baseline for the score dependent part: maximum score
or winning probability could make more sense.
3) Assign votes in `Thread::Search` as iterations are completed and use
voting results to stop search.
4) Select best thread as the threads voting for best move with the highest
completed depth or, alternatively, vote on PV moves.
Marco Costalba [Sat, 28 Jul 2018 13:33:39 +0000 (15:33 +0200)]
Improve Stats definition
Use operator const T&() instead of operator T() to avoid possible
costly hidden copies of non-scalar nested types.
Currently StatsEntry has a single member T, so assuming
sizeof(StatsEntry) == sizeof(T) it happens to work, but it's
better to use the size of the proper entry type in std::fill.
Note that current code works because std::array items are ensured
to be allocated in contiguous memory and there is no padding among
nested arrays. The latter condition does not seem to be strictly
enforced by the standard, so be careful here.
Finally use address-of operator instead of get() to fully hide the
wrapper class StatsEntry at calling sites. For completness add
the arrow operator too and simplify the C++ code a bit more.
Same binary code as previous master under the Clang compiler.
This is the first patch teaching Stockfish how to use the 7-pieces
Syzygy tablebase currently calculated by Bujun Guo (@noobpwnftw) and
Ronald de Man (@syzygy1). The 7-pieces database are so big that they
required a change in the internal format of the files (technically,
some DTZ values are 16 bits long, so this had to be stored as wide
integers in the Huffman tree).
Here are the estimated file size for the 7-pieces Syzygy files,
compared to the 151G of the 6-pieces Syzygy:
```
7.1T ./7men_testing/4v3_pawnful (ongoing, 120 of 325 sets remaining)
2.4T ./7men_testing/4v3_pawnless
2.3T ./7men_testing/5v2_pawnful
660G ./7men_testing/5v2_pawnless
117G ./7men_testing/6v1_pawnful
87G ./7men_testing/6v1_pawnless
```
Some pointers to download or recalculate the tables:
Location of original files, by Bujun Guo:
ftp://ftp.chessdb.cn/pub/syzygy/
Bench: 5591925 (No functional change if SyzygyTB is not used)
----------------------
Comment by Leonardo Ljubičić (@DragonMist)
This is an amazing achievement, generating and being able to use 7 men syzygy
on the fly. Thank you for your efforts @noobpwnftw !! Looking forward how this
will work in real life, and expecting some trade off between gaining perfect
play and slow disc Access, but once the disc speed and space is not a problem,
I expect 7 men to yield something like 30 elo at least.
-----------------------
Comment by Michael Byrne (@MichaelB7)
This definitely has a bright future. I turned off the 50 move rule (ala ICCF
new rules) for the following position: `[d]8/8/1b6/8/4N2r/1k6/7B/R1K5 w - - 0 1`
This position is a 451 ply win for white (sans the 50 move rule, this position
was identified by the generator as the longest cursed win for white in KRBN v KRB).
This patch adds the tropism measure as a new term in the king danger variable.
Since we then trasform this variable as a Score via a quadratic formula, the
main effect of the patch is the positive correlation of the tropism measure
with some checks and pins information already present in the king danger code.
Miguel Lahoz [Fri, 27 Jul 2018 16:23:53 +0000 (00:23 +0800)]
Increase the mg->eg gradient for the PawnlessFlank malus
Just a change of value to S(19, 84). Also somewhat of a follow up
to the recent tweak in definition of KingFlank.
I tried a lot of other values before this, increasing and decreasing
but with little success, and before giving up I wanted to try tweaking
the middlegame and endgame values in the opposite directions. I guess
this is somewhat lucky.
The previous commit wouldn't compile on the Microsoft Virtual Studio C++ compiler. So use a more compatible style for the same idea (which we already use in numerous places of evaluate.cpp, for instance in line 563).
Under the Clang compiler, both versions generate exactly the same machine code (same md5 signatures for the two binaries).
Remove a popcount for HinderPassedPawn, and compensate by doubling
the bonus from S(4,0) to to S(8,0).
Maybe it was pure luck, but we got the idea of this Elo gaining patch by
seing the simplification attempt by Mike Whiteley in pull request #1703.
This suggests that whenever we have a passed evaluation simplification,
we should consider the possibility that the master bonus has become
slightly out of tune with time, and we should try a few Elo gaining [0..4]
tests by hand-tuning the master bonus.
Miguel Lahoz [Wed, 25 Jul 2018 16:11:51 +0000 (00:11 +0800)]
Tweak KingFlank when king is on edge files
This tweak excludes files D and E from the KingFlank bitboard when our
king is on the A or H files respectively. As far as I can tell, this
affects two things: the calculation for CloseEnemies and PawnlessFlank.
Aside from filtering out slightly less relevant attacks in the flank,
I suspect this helps with king prophylaxis, avoiding attacks and moving
towards the center when the pawns start to come off.
How to continue: Tweaking the two bonuses mentioned might give some
gain, although as far as I can tell, CloseEnemies is very sensitive to
even small changes.
I don't measure any speed-up on my system, with two parallel benches at depth 22:
Total time (ms) : 74989
Nodes searched : 144830258
Nodes/second : 1931353
master
Total time (ms) : 75341
Nodes searched : 144830258
Nodes/second : 1922329
testedpatch
And anyway, like Stefan Geschwentner, I don't think that a 0.3% speed-up would
be enough to pass a [0..5] LTC test -- as a first approximation, we have this
rule of thumb that 1% speed-up gives about 1 Elo point.
However, considering the facts that the reformatting by itself is interesting,
that this is your first green test and that you played by the rules by running
the SPRT[0..5] test before opening the pull request, I will commit the change.
I will only take the liberty to change the occurrences of safe in lines 590 and
591 to b, to make the code more similar to lines 584 and 585.
How to continue after this patch: there is now a slight semantic
overlap between the ThreatByPawnPush and the ThreatBySafePawn bonuses,
so hand-tuning either of these, or both at the same time, is natural.
This patch implements some idea by Alain Savard and Mike Whiteley taken from the perpertual renaming/reformatting thread.
This is a pure code cleaning patch (so no change in functionality), but I use it as a pretext to correct the bogus bench number that I introduced in the previous commit.
2) So there was a suspicion that the original fast passing LTC test which lead
us to accept the patch may have been a statistical accident, so we organized
a match against the previous master at LTC to get an Elo estimate for the
patch:
Replace the depth part in the reduction formula for higher depths
with a slower growing linear function. So for depth > 3 less reductions
are used.
What we can try next:
- move the break point to even higher depths
- tweak the slope for lower and higher depth
- even possibly use a further higher depth threshold for a another
slower growing function
Remove stronglyProtected Queen for ThreatByRook. Idea is that in the
current master the SliderOnQueen bonus and the see_ge() function do
something similar as ThreatByRook for Queen, so this patch removes
some redundancy, in that sense.
This is a non-functional simplification. We change replaces an 'OR'
and a lookup (rank_bb(ksq)) with a bitwise ~. This is fewer operations
and is probably faster.
This is a follow-up of the previous pull request (#1686) by Miguel.
We simplify the "Overload" bonus condition by re-using the "weak"
variable, which captures well the essence of the overload condition.
This may also be a small speed optimization because the weak variable
is in a register at this point of the code.
Miguel Lahoz [Thu, 19 Jul 2018 11:18:30 +0000 (19:18 +0800)]
Remove connectivity.
There seems to be some strange interaction between Overload and Connectivity.
Overload encourages us to not have too many defended and attacked pieces,
as this may expose us to various tactics. This feels somewhat like it is in
conflict with Connectivity, where pieces are defended preemptively.
Here I take the "pick one or the other" approach and just remove connectivity,
while strengthening the effect of Overload to compensate. The reasoning is that
if we defend our pieces preemptively, then it does get attacked, we want to do
something about it so we don't get penalized by Overload. On the other
hand, if it doesn't get attacked, then there's no need to defend it.
Use in addition the counter move history table for sorting quiet
check evasion moves in main and quiecence search. Also rename
"contHistory" to "continuationHistory" while there.
Miguel Lahoz [Mon, 16 Jul 2018 16:51:43 +0000 (00:51 +0800)]
Use single value for KingProtector.
After some recent big tuning session, the values for King Protector were
simplified to only be used on minor pieces. This patch tries to further
simplify by just using a single value, since current S(6,5) and S(5,6)
are close to each other. The value S(6,6) ended up passing, although
S(5,5) was also tried and failed STC.
Extend the bonus for Overload to cases where our side
has more than one attacker to a non pawn piece.
Based on an idea by Bryan in the forum. For instance,
now black gets the overload bonus in this position:
8/5R1k/6pb/p6p/P1N4P/1Pp5/2K3P1/2N4r b - - 6 46
because two black pieces are attacking the knight on c1
that is defended only by the king.
This was broken when cuckoo hashing was added, and
subtly broke (for example) lichess' Android application,
thus illustrating the original judgement was sound.
I think I reached the limit of the fishtest framework. It frequently
crashed at 2 million games already. The small values also moved a lot
throughout the entire tuning session though with smaller margin. The
passed danger and close enemies values seems the most sensitive (changing
close enemies alone to 6 failed before but now it passes), whether or not
they are close to optimal I don't know, but it seems some parameters are
also correlated to others.
In the current master, ThreatByKing is an array of two Scores, one for
when we have a single attack and one for when we have many. The latter
case is very rarely called during bench and was recently given a strange
negative value during a tuning run, as pointed out by @candirufish on
commit efd4ca2. Here, we simplify away this second case entirely, and
increase the remaining ThreatByKing to compensate.
Although I derived the parameter tweak independently, with the goal of
preserving the same average bonus, I later noticed that a very similar
Score had already been derived by an ongoing SPSA tuning session.
I therefore recognize @candirufish for first discovering these values.
I would also like to thank @Rocky640 for valuable feedback that pointed
me in the direction of ThreatByKing.
Ondrej Mosnáček [Sat, 9 Jun 2018 11:45:49 +0000 (13:45 +0200)]
Move PSQ score to Position
This patch simplifies Position::do_move() by moving the PSQ score from
StateInfo to Position and updating it inside the put/remove/move_piece
functions.
The downside is that there is now slightly more computation done in
Position::undo_move(), but the fishtest results are Elo neutral.
protonspring [Sun, 17 Jun 2018 02:26:25 +0000 (20:26 -0600)]
Remove make_bitboard()
In current master, the function make_bitboard() does nothing apart from
helping initialize the SquareBB[] array. This seems like an unnecessary
abstraction layer.
The advantage of make_bitboard() is we can define a bitboard, in a simple
and general way, not only from a single square but also from a list of
squares. It is more elegant, faster and readable than combining multiple
SquareBB explicitly, but the last complex use case in evaluation was
simplified away a few months ago.
If make_bitboard() becomes useful again to define complicated bitboards,
it will be easy enough to reintroduce it using this pull request as
an implementation reference.
Alain SAVARD [Sun, 24 Jun 2018 22:06:13 +0000 (18:06 -0400)]
Simplify HinderPassedPawn bonus
Make sure each piece is not scored more than once as a passed pawn "hinderer",
by scoring only the blockers along the passed pawn path. Inspired by TCEC Game 29.
Passed STC as a simplification
http://tests.stockfishchess.org/tests/view/5b3016d00ebc5902b2e58552
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 75388 W: 16656 L: 16641 D: 42091
Passed LTC as a simplification
http://tests.stockfishchess.org/tests/view/5b302ed90ebc5902b2e587fc
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 49157 W: 8460 L: 8386 D: 32311
Current master was also counting the number of attacks along a passed pawn path,
which might be misleading:
a) a defender might be counted many times for the same pawn path. For example a
White rook on a1 attacking a black pawn on a7 would score the bonus * 6 but
would be probably better placed on a8
b) a defender might be counted on different pawn paths and might be overloaded. For
example a Ke4 or Qe4 against pawns on d6 and f6 would score the bonus * 6.
Counting each blocker or attacker only once is more complicated, and does not help
either: http://tests.stockfishchess.org/tests/view/5b2ff1cb0ebc5902b2e582b2
After this small simplification, there might be ways to increase the HinderPassedPawn
penalty.
Give more incentive to king activity in the endgame by increasing the weight
of the "outflanking" variable from 8 to 12 in the function evaluate_initiative().
Michael An [Thu, 21 Jun 2018 01:05:27 +0000 (21:05 -0400)]
Fix GCC 8 cast warnings
Silences the following warnings when compiling with GCC 8.
The fix is to use an intermediate pointer to anonymous function:
```
misc.cpp: In function 'int WinProcGroup::get_group(size_t)':
misc.cpp:241:77: warning: cast between incompatible function types from 'FARPROC' {aka 'long long int (*)()'} to 'fun1_t' {aka 'bool (*)(_LOGICAL_PROCESSOR_RELATIONSHIP, _SYSTEM_LOGICAL_PROCESSOR_INFORMATION_EX*, long unsigned int*)'} [-Wcast-function-type]
auto fun1 = (fun1_t)GetProcAddress(k32, "GetLogicalProcessorInformationEx");
^
misc.cpp: In function 'void WinProcGroup::bindThisThread(size_t)':
misc.cpp:309:71: warning: cast between incompatible function types from 'FARPROC' {aka 'long long int (*)()'} to 'fun2_t' {aka 'bool (*)(short unsigned int, _GROUP_AFFINITY*)'} [-Wcast-function-type]
auto fun2 = (fun2_t)GetProcAddress(k32, "GetNumaNodeProcessorMaskEx");
^
misc.cpp:310:67: warning: cast between incompatible function types from 'FARPROC' {aka 'long long int (*)()'} to 'fun3_t' {aka 'bool (*)(void*, const _GROUP_AFFINITY*, _GROUP_AFFINITY*)'} [-Wcast-function-type]
auto fun3 = (fun3_t)GetProcAddress(k32, "SetThreadGroupAffinity");
^
```
DU-jdto [Wed, 13 Jun 2018 05:22:52 +0000 (15:22 +1000)]
Remove lmrDepth restriction on quiet see pruning
And tweak the threshold value. With this threshold and the current piece
values, this permits see pruning on quiets to be done up to an lmrDepth
of 9 (beyond that the threshold is below -QueenValueMg and see_ge will
pass unconditionally).
protonspring [Mon, 11 Jun 2018 08:17:03 +0000 (10:17 +0200)]
Optimize an expression in endgame.cpp
I believe using foward_file_bb() here is fewer instructions.
a) Fewer instructions and probably more clear (debatable).
b) Possible that a lookup is slower than a few local operations, but the
forward_file_bb table is probably used often enough that it is always
cached.
candirufish [Sun, 10 Jun 2018 08:53:05 +0000 (01:53 -0700)]
Simplify capture pruning margin formula
Using just `PawnValueEg * depth` as Capture Prune Margin. There was a bunch
of patches that passed recently regarding captures, maybe this part of the
master code redundant? The patch was tested as a simplification:
After several tests it seems best to increase contempt from 12 to 21. This does
not regress against contempt=0 and gives a gain of around 7-8 elo against SF 7
in comparison to current default contempt.
STC: Test for non-regression contempt=21 against contempt=0
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 71250 W: 13956 L: 13926 D: 43368
http://tests.stockfishchess.org/tests/view/5b19a58d0ebc5902ab9c3bfa
STC: Test contempt 21 against SF 7
ELO: 190.06 +-2.8 (95%) LOS: 100.0%
Total: 40000 W: 22608 L: 2676 D: 14716
http://tests.stockfishchess.org/tests/view/5b19a6520ebc5902ab9c3c0e
STC: Test master against SF 7 for comparison
ELO: 182.95 +-2.7 (95%) LOS: 100.0%
Total: 40000 W: 21905 L: 2595 D: 15500
http://tests.stockfishchess.org/tests/view/5b16f5bc0ebc59214346d5ca
LTC: Test for non-regression contempt=21 against contempt=0
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 47666 W: 6914 L: 6832 D: 33920
http://tests.stockfishchess.org/tests/view/5b1a170b0ebc5902ab9c3fde
LTC: Test contempt 21 against SF 7
ELO: 203.92 +-2.6 (95%) LOS: 100.0%
Total: 40000 W: 22447 L: 1340 D: 16213
http://tests.stockfishchess.org/tests/view/5b1a174b0ebc5902ab9c3fe1
LTC: Test master against SF 7 for comparison
ELO: 196.08 +-2.6 (95%) LOS: 100.0%
Total: 40000 W: 21639 L: 1191 D: 17170
http://tests.stockfishchess.org/tests/view/5b1a17e40ebc5902ab9c3fe4
candirufish [Wed, 6 Jun 2018 08:13:08 +0000 (10:13 +0200)]
Quiet move soft fail high bonus
Extra bonus for quiet move creating a huge soft fail high (triggered
in 21% of quiet bestmoves on a normal bench run). Pb00067 original idea
using PawnValueMg.
This has the property of raising alpha before calling qsearch(), thus
maybe giving some more cuts during qsearch(). The patch is equivalent
to the use of cycle detection inside qsearch() at depth 0, but is in
fact implemented by re-ordering code inside search(), which explains
the [0..4] bounds in the following tests.
STC (interrupted after 124250 games, with LLR=0.87):
http://tests.stockfishchess.org/tests/view/5b1500bd0ebc5902a8b420bf
LLR: 0.87 (-2.94,2.94) [0.00,4.00]
Total: 124250 W: 24973 L: 24470 D: 74807
After a helpful suggestion from AppVeyor support staff, moving the Stockfish
execution from ps to cmd seems to work. Alternative to PR #1624 tested in PR #1637.
Guenther Demetz [Mon, 4 Jun 2018 07:10:30 +0000 (09:10 +0200)]
Remove a superfluous subtrahend
The '- 1' subtrahend was introduced for guarding against null move
search at root, which would be nonsense. But this is actually already
guaranteed by the !PvNode condition. This followed from the discussion
in pull request 1609: https://github.com/official-stockfish/Stockfish/pull/1609
joergoster [Fri, 1 Jun 2018 20:35:23 +0000 (22:35 +0200)]
Bugfix of Position::has_repeated()
The function Position::has_repeated() is used by Tablebases::root_probe()
to determine whether we can rank all winning moves with the same value, or
if we need to strictly rank by dtz in case the position has already been
repeated once, and we are risking to run into the 50-move rule and thus
losing the win (especially critical in some very complicated endgames).
To check whether the current position or one of the previous positions
after the last zeroing move has already been occured once, we start looking
for a repetition of the current position, and if that is not the case, we
step one position back and repeat the check for that position, and so on.
If you now look at how this was done before the new root ranking patch was
merged two months ago, it seems quite obvious that it is a simple oversight:
https://github.com/official-stockfish/Stockfish/commit/108f0da4d7f993732aa2e854b8f3fa8ca6d3b46c
More specifically, after we stepped one position back with
```
stc = stc->previous;
```
we now have to start checking for a repetition with
Makes sure the potential benefit of first touch does not depend on
the order of the UCI commands Threads and Hash, by reallocating the
hash if a Threads is issued. The cost is zeroing the TT once more
than needed. In case the prefered order (first Threads than Hash)
is employed, this amounts to zeroing the default sized TT (16Mb),
which is essentially instantaneous.
Follow up for https://github.com/official-stockfish/Stockfish/pull/1601
where additional data and discussion is available.
Stockfish currently takes a while to clear the TT when using larger hash sizes.
On one machine with 128 GB hash it takes about 50 seconds with a single thread,
allowing it to use all allocated cores brought that time down to 4 seconds on
some Linux systems. The patch was further tested on Windows and refined with
NUMA binding of the hash initializing threads (we refer to pull request #1601
for the complete discussion and the speed measurements).
protonspring [Thu, 24 May 2018 16:46:38 +0000 (18:46 +0200)]
Simplify BlockedByPawn to one dimension
I was able to get this to pass which reduces BlockedByPawn to one dimension
with NO distance from edge offset.
GOOD) It's more simple and may provide additional clarity for further
simplifications. Facilitates migrating unblocked to one dimension as well.
BAD) If there is indeed a distance component to BlockedStorm (may or may
not be the case), this obfuscates this component into ShelterStrength and
UnblockedStorm. This may be more convoluted. Also, it may be more convenient
to have each of the three arrays (ShelterStrength, BlockedStorm, and UnBlocked)
be the same size.
As discussed with @pb00068, the condition to prevent recursive verification
was not completely correct. This patch corrects that condition, and adds an
assert. In the current implementation, recursive verification needs to be
avoided in order not to break the verification of the move closer to the
root (i.e. to not set thisThread->nmp_min_ply to zero prematurely).
This patch is tested as a bug fix, based on and tested against PR #1609 .
Simplifying away all the progressKey stuff gives exactly the same bench,
without any speed impact. Tested for speed against master with two benches
at depth 22 ran in parallel:
**testedpatch**
Total time (ms) : 92350
Nodes searched : 178962949
Nodes/second : 1937877
**master**
Total time (ms) : 92358
Nodes searched : 178962949
Nodes/second : 1937709
We also tested the patch at STC for no-regression with [-3, 1] bounds:
mstembera [Wed, 16 May 2018 21:38:13 +0000 (23:38 +0200)]
Fix MSVC errors in tbprobe.cpp
Default template parameters values and recursive functions do not play well
together. Fix for below errors that showed up after updating to latest MSVC.
````
tbprobe.cpp(1156): error C2672:
'search': no matching overloaded function found
tbprobe.cpp(1198): error C2783:
'Tablebases::WDLScore `anonymous-namespace'::search(Position &,Tablebases::ProbeState *)':
could not deduce template argument for 'CheckZeroingMoves'
Tom Truscott [Wed, 16 May 2018 20:47:41 +0000 (22:47 +0200)]
Use cycle detection to bound search value
A position which has a move which draws by repetition, or which could have
been reached from an earlier position in the game tree, is considered to be
at least a draw for the side to move.
How does the algorithm work in practice? The algorithm is an efficient
method to detect if the side to move has a drawing move, without doing any
move generation, thus possibly giving a cheap cutoffThe most interesting
conditions are both on line 1195:
This uses the position keys as a sort-of Bloom filter, to avoid the expensive
checks which follow. For "upcoming repetition" consider the opening Nf3 Nf6 Ng1.
The XOR of this position's key with the starting position gives their difference,
which can be used to look up black's repeating move (Ng8). But that look-up is
expensive, so line 1195 checks that the white pieces are on their original squares.
This is the subtlest part of the algorithm, but the basic idea in the above game
is there are 4 positions (starting position and the one after each move). An XOR
of the first pair (startpos and after Nf3) gives a key matching Nf3. An XOR of
the second pair (after Nf6 and after Ng1) gives a key matching the move Ng1. But
since the difference in each pair is the location of the white knight those keys
are "identical" (not quite because while there are 4 keys the the side to move
changed 3 times, so the keys differ by Zobrist::side). The loop containing line
1195 does this pair-wise XOR-ing.
Continuing the example, after line 1195 determines that the white pieces are
back where they started we still need to make sure the changes in the black
pieces represents a legal move. This is done by looking up the "moveKey" to
see if it corresponds to possible move, and that there are no pieces blocking
its way. There is the additional complication that, to match the behavior of
is_draw(), if the repetition is not inside the search tree then there must be
an additional repetition in the game history. Since a position can have more
than one upcoming repetition a simple count does not suffice. So there is a
search loop ending on line 1215.
On the other hand, the "no-progress' is the same thing but offset by 1 ply.
I like the concept but think it currently has minimal or negative benefit,
and I'd be happy to remove it if that would get the patch accepted. This
will not, however, save many lines of code.
• The code in search() that checks for cycles has numerous possible variants.
Perhaps the check need could be done in qsearch() too.
• The biggest improvement would be to get "no progress" to be of actual benefit,
and it would be helpful understand why it (probably) isn't. Perhaps there is an
interaction with the transposition table or the (fantastically complex) tree
search. Perhaps this would be hard to fix, but there may be a simple oversight.
VoyagerOne [Mon, 14 May 2018 04:52:16 +0000 (06:52 +0200)]
Update search.cpp
At PvNodes allow bonus for prior counter move that caused a fail low
for depth 1 and 2. Note : I did a speculative LTC on yellow STC patch
since history stats tend to be highly TC sensitive
Use the whole kingRing for pawn attackers instead of only the squares directly
around the king. This tends to give quite a lot more kingAttackersCount, so to
compensate and to avoid raising the king danger too fast we lower the values
in the KingAttackWeights array a little bit.
Credits to user @xoroshiro for the idea of using the kingRing for pawn attackers.
How to continue? It seems that the KingAttackWeights[] array stores values
which are quite Elo-sensitive, yet they have not been tuned with SPSA recently.
There might be easy Elo points to get there.
mstembera [Thu, 10 May 2018 20:49:56 +0000 (13:49 -0700)]
Include all blockers in king danger
Simplification: in king danger, include all blockers and not only pinned
pieces, since blockers enemy pieces can result in discovered checks which
are also bad.
I also incorrectly scheduled STC with [0,5] which it failed.
http://tests.stockfishchess.org/tests/view/5af283c00ebc5968e6523f33
LLR: -2.94 (-2.94,2.94) [0.00,5.00]
Total: 12338 W: 2451 L: 2522 D: 7365
Thanks to @vondele and @Rocky640 for a cleaner version of the patch,
and the following comments!
> Most of the pinned, (or for this pull request, blocking) squares were
> already computed in the unsafeChecks, the only missing squares being:
>
> a) squares attacked by a Queen which are occupied by friendly piece
> or "unsafe". Note that adding such squares never passed SPRT[0,5].
>
> b) squares not in mobilityArea[Us].
>
> There is a strong relationship between the blockers and the unsafeChecks,
> but the bitboard unsafeChecks is still useful when the checker is not
> aligned with the king, and the checking square is occupied by friendly
> piece or is "unsafe". This is always the case for the Knight.
protonspring [Thu, 10 May 2018 13:46:13 +0000 (15:46 +0200)]
Consolidate pawn storm types
Simplification: the Unopposed and Unblocked pawn storm types are mathematically
similar enough to combine with no Elo loss. This reduces the pawn storm types
to BlockedByPawn and UnBlocked.
candirufish [Wed, 9 May 2018 07:18:22 +0000 (09:18 +0200)]
Tuned some pawns and evaluation constants
Tuned values in pawns.cpp and evaluate.cpp after a SPSA session:
419k games 60sec 600nodetime. We have adjusted the PassedRank[]
output of the SPSA session to keep increasing values with rank,
and PassedFile[] output to keep the West <–> East symmetry of
the evaluation.
This patch simplifies the control flow in search(), removing an if
and a goto. A side effect of the patch is that Stockfish is now a
little bit more selective at low depths, because we allow razoring,
futility pruning and probcut pruning after a null move.
Alain SAVARD [Tue, 8 May 2018 09:00:51 +0000 (11:00 +0200)]
Drop the lever condition for backwards
We can view the patch version as adding some "undermining bonus" for
level pawns, when the defending side can not easily avoid the exchange
by advancing her pawn.
• Case 1) White b2,c3, Black a3,b3:
Black is breaking through, b2 deserves a penalty
• Case 2) White b2,c3, Black a3,c4:
if b2xa3 then White ends up with a weak pawn on a3
and probably a weak pawn on c3 too.
In either case, White can still not safely play b2-b3 and make a
phalanx with c3, which is the essence of a backward pawn definition.
There were some promising tests a couple of months ago about adding
a lever condition for king danger in evaluate.cpp, maybe it would
be time to re-try this after all the recent changes in pawns.cpp
Remove nine boolean arguments and the corresponding skipEarlyPruning variable.
Instead, skip early pruning only when there is an excluded move, and try null
move pruning only if the previous move was not itself a null move.
protonspring [Sun, 6 May 2018 07:42:49 +0000 (09:42 +0200)]
Simplify the backward pawns code
The two lines of code in the patch seem to be just as good as master.
1. We now only look at the current square to see if it is currently backward,
whereas master looks there AND further ahead in the current file (master would
declare a pawn "backward" even though it could still safely advance a little).
This simplification allows us to avoid the use of the difficult logic with
`backmost_sq(Us, neighbours | stoppers)`.
2. The condition `relative_rank(Us,s) < RANK_5` is simplified away.
• The new code flags some pawns on the 5th rank as backward, which was not the
case in the old master. So maybe we should test a version with that included?
• Further tweaks of the backward condition with [0..5] bounds?
Tweak the connected[] array value for pawns on rank 5
A recent tuning session by Jerry Donald Watson suggested that the
value for the pawns on the fifth rank in the connected[] array were
a little bit too high in master. We lower here this value from 75 to 65.
Correct a bug introduced by Stéphane in the previous patch.
When we are using the "Bitboard + Square" overloaded operators,
the compiler uses the interpediate SquareBB[s] to transform the
square into a Bitboard, and then calculate the result.
For instance, the following code:
```
b = pos.pieces(Us, PAWN) & s
```
generates in fact the code:
```
b = pos.pieces(Us, PAWN) & SquareBB[s]`
```
The bug introduced by Stéphane in the previous patch was the
use of `b = pos.pieces(Us, PAWN) & (s + Up)` which can result
in out-of-bounds errors for the SquareBB[] array if s in the
last rank of the board.
We coorect the bug, and also add some asserts in bitboard.h to
make the code more robust for this particular bug in the future.