It causes a serious regression hanging a simple fixed
depth search. Reproducible with:
position fen q1B5/1P1q4/8/8/8/6R1/8/1K1k4 w - - 0 1
go depth 13
The reason is a search tree explosion due to:
if (... && depth < 3 * ONE_PLY)
extension = ONE_PLY;
This is very dangerous code by itself because triggers **at the leafs**
and in the above position keeps extending endlessly. In normal games
time deadline makes the search to stop sooner or later, but in fixed
seacrch we just hang possibly for a very long time. This is not acceptable
because 'go depth 13' shall not be a surprise for any position.
This patch reverts commit 76f1807baa.
and fixes the issue https://github.com/official-stockfish/Stockfish/issues/2091
Bench: 3243738
Introduce a new search extension when pushing an advanced passed pawn is
also suggested by the first killer move. There have been previous tests
which have similar ideas, mostly about pawn pushes, but it seems to be
overkill to extend too many moves. My idea is to limit the extension to
when a move happens to be noteworthy in some other way as well, such as
in this case, when it is also a killer move.
STC:
LLR: 2.96 (-2.94,2.94) [0.50,4.50]
Total: 19027 W: 4326 L: 4067 D: 10634
http://tests.stockfishchess.org/tests/view/5cac2cde0ebc5925cf00c36d
LTC:
LLR: 2.94 (-2.94,2.94) [0.00,3.50]
Total: 93390 W: 15995 L: 15555 D: 61840
http://tests.stockfishchess.org/tests/view/5cac42270ebc5925cf00c4b9
For future tests, it looks like this will interact heavily with passed
pawn evaluation. It may be good to try more variants of some of the more
promising evaluations tests/tweaks.
Bench: 3666092
Instead of looping through kfrom,kto, rfrom, rto, we can use BetweenBB. This is less lines of code and it is more clear what castlingPath actually is. Personal benchmarks are all over the place. However, this code is only executed when loading a position, so performance doesn't seem that relevant.
No functional change.
The kingDanger term is intended to give a penalty which increases rapidly in the middlegame but less so in the endgame. To this end, the middlegame component is quadratic, and the endgame component is linear. However, this produces unintended consequences for relatively small values of kingDanger: the endgame penalty will exceed the middlegame penalty. This remains true up to kingDanger = 256 (a S(16, 16) penalty), so some of these inaccurate penalties are actually rather large.
In this patch, we increase the threshold for applying the kingDanger penalty to eliminate some of this unintended behavior. This was very nearly, but not quite, sufficient to pass on its own. The patch was finally successful by integrating a second kingDanger tweak by @Vizvezdenec, increasing the kingDanger constant term slightly and improving both STC and LTC performance.
Where do we go from here? I propose that in the future, any attempts to tune kingDanger coefficients should also consider tuning the kingDanger threshold. The evidence shows clearly that it should not be automatically taken to be zero.
Special thanks to @Vizvezdenec for the kingDanger constant tweak. Thanks also to all the approvers and CPU donors who made this possible!
STC:
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 141225 W: 31239 L: 30846 D: 79140
http://tests.stockfishchess.org/tests/view/5cabbdb20ebc5925cf00b86c
LTC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 30708 W: 5296 L: 5043 D: 20369
http://tests.stockfishchess.org/tests/view/5cabff760ebc5925cf00c22d
Bench: 3445945
The sed command is a bit different in Mac OS X (why not!).
The ‘-i’ option required a parameter to tell what extension to add for the
backup file. To fix it, just add extension for backup file, for example ‘.bak’
Fix broken Trevis CI test
No functional change.
The current update only by main thread depends on the luck of
whether main thread sees any/many changes to the best move or not.
It then makes large, lumpy changes to the time to be
used (1x, 2x, 3x, etc) depending on that sample of 1.
Use the average across all threads to get a more reliable
number with a smoother distribution.
STC @ 5+0.05 th 4 :
LLR: 2.95 (-2.94,2.94) [0.50,4.50]
Total: 51899 W: 11446 L: 11029 D: 29424
http://tests.stockfishchess.org/tests/view/5ca32ff20ebc5925cf0016fb
STC @ 5+0.05 th 8 :
LLR: 2.96 (-2.94,2.94) [0.50,4.50]
Total: 13851 W: 2843 L: 2620 D: 8388
http://tests.stockfishchess.org/tests/view/5ca35ae00ebc5925cf001adb
LTC @ 20+0.2 th 8 :
LLR: 2.95 (-2.94,2.94) [0.00,3.50]
Total: 48527 W: 7941 L: 7635 D: 32951
http://tests.stockfishchess.org/tests/view/5ca37cb70ebc5925cf001cec
Further work:
Similar changes might be possible for the fallingEval and timeReduction calculations (and elsewhere?), using either the min, average or max values across all threads.
Bench 3506898
Simplification which removes the pawns connected array.
Instead of storing the values in an array, the values are
calculated real-time. This is about 1.6% faster on my machines.
Performance:
master ave nps: 159,248,672
patch ave nps: 161,905,592
STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 20363 W: 4579 L: 4455 D: 11329
http://tests.stockfishchess.org/tests/view/5c9925ba0ebc5925cfff79a6
Non functional change.
Shuffle detection procedure :
Shuffling positions are detected if
the last 36 moves are reversible (rule50_count() > 36),
the position have been already in the TT,
there is a still a pawn on the board (to avoid special endings like KBN vs K).
The position is then judged as a draw.
An extension is realized if we already made 14 successive reversible moves in PV to accelerate the detection of the eventual draw.
To go further : we can still improve the idea. The length of the tests need a lot of ressources.
the limit of 36 is logic but must be checked again for special zugzwang positions,
this limit can be decreased in special positions,
the limit of 14 moves for extension has not been tuned.
STC
LLR: -2.94 (-2.94,2.94) [0.50,4.50]
Total: 32595 W: 7273 L: 7275 D: 18047 Elo +0.43
http://tests.stockfishchess.org/tests/view/5c90aa330ebc5925cfff1768
LTC
LLR: 2.95 (-2.94,2.94) [0.00,3.50]
Total: 51249 W: 8807 L: 8486 D: 33956 Elo +1.85
http://tests.stockfishchess.org/tests/view/5c90b2450ebc5925cfff1800
VLTC
LLR: 2.96 (-2.94,2.94) [0.00,3.50]
Total: 137974 W: 20503 L: 19983 D: 97488 Elo +1.05
http://tests.stockfishchess.org/tests/view/5c9243a90ebc5925cfff2a93
Bench: 3548313
Adding a clamp function makes some of these range limitations a bit prettier and removes some #include's.
STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 28117 W: 6300 L: 6191 D: 15626
http://tests.stockfishchess.org/tests/view/5c9aa1df0ebc5925cfff8fcc
Non functional change.
always use the implementation of gives_check in position, no need to
hand-inline part of the implementation in search.
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 57895 W: 12632 L: 12582 D: 32681
http://tests.stockfishchess.org/tests/view/5c9eaa4b0ebc5925cfffc9e3
No functional change.
This is a non-functional code style change.
If we add an accessor function for SquareBB we can consolidate all of the asserts. This is also a bit cleaner because all SquareBB accesses go through this method making future changes easier to manage.
STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 63406 W: 14084 L: 14045 D: 35277
http://tests.stockfishchess.org/tests/view/5c9ea6100ebc5925cfffc9af
No functional change.
This is covered by the line just before. If we would like to protect
against the piece value of e.g. a N == B, this could be done by an
assert, no need to do this at runtime.
No functional change.
This is a non-functional simplification/speedup.
The truth-table for popcount(support) >= popcount(lever) - 1 is:
------------------lever
------------------0-------1---------2
support--0------X-------X---------0
-----------1------X-------X---------X
-----------2------X-------X---------X
Thus, it is functionally equivalent to just do: support || !more_than_one(lever) which removes the expensive popcounts and the -1.
Result of 20 runs:
base (...h_master.exe) = 1451680 +/- 8202
test (./stockfish ) = 1454781 +/- 8604
diff = +3101 +/- 931
STC
LLR: 2.94 (-2.94,2.94) [-3.00,1.00]
Total: 35424 W: 7768 L: 7674 D: 19982
Http://tests.stockfishchess.org/tests/view/5c970f170ebc5925cfff5e28
No functional change.
While looking at pruning using see_ge() (which is very valuable)
it became apparent that the !extension test is not adding any
value - simplify it away.
STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 56843 W: 12621 L: 12569 D: 31653
http://tests.stockfishchess.org/tests/view/5c8588cb0ebc5925cffe77f4
LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 78622 W: 13223 L: 13195 D: 52204
http://tests.stockfishchess.org/tests/view/5c8611cc0ebc5925cffe7f86
Further work could be to optimize the remaining see_ge() test. The idea of less pruning at higher depths is valuable, but perhaps the test (-PawnValueEg * depth) can be improved.
Bench: 3188688
On OS X threads other than the main thread are created with a reduced stack
size of 512KB by default, this is dangerously low for deep searches, so
adjust it to TH_STACK_SIZE. The implementation calls pthread_create() with
proper stack size parameter.
Verified for no regression at STC enabling the patch on all platforms where
pthread is supported.
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 50873 W: 9768 L: 9700 D: 31405
No functional change.
This is a non-functional simplification / code-style change.
This popcount16 method does nothing but initialize the PopCnt16 arrays.
This can be done in a single bitset line, which is less lines and more clear. Performance for this code is moot.
No functional change.
This is a functional simplification that removes the FutilityMoveCounts array with a simple equation using only ints.
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 14175 W: 3123 L: 2987 D: 8065
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 9900 W: 1735 L: 1597 D: 6568
Bench: 3380343
This is a non-functional patch which shrinks the reductions array.
This saves about 8Kb of memory.
The only slow part of master's reductions array is the calculation
of the log values, so using a separate array for those values and
calculating the rest real-time appears to be just as fast as master.
STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 63245 W: 13906 L: 13866 D: 35473
http://tests.stockfishchess.org/tests/view/5c7b571f0ebc5925cffdc104
No funcional change.
This removes the skipQuiets variable, as was done in an earlier round by
@protonspring, but fixes an oversight which led to wrong mate
announcements. Quiets can only be pruned when there is no mate score, so
set moveCountPruning at the right spot.
tested as a fix at STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 66321 W: 14690 L: 14657 D: 36974
http://tests.stockfishchess.org/tests/view/5c74f3170ebc5925cffd4b3c
and as the full patch at LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 25903 W: 4341 L: 4228 D: 17334
http://tests.stockfishchess.org/tests/view/5c7540030ebc5925cffd506f
Bench: 3292342
This is a somewhat different patch. It fixes blindspots for
two knights vs pawn endgame.
With local testing starting from random KNNvKP positions where the
pawn has not advanced beyond the 4th rank (thanks @protonspring !)
at 15+0.15 (4 cores), this went +105=868-27 against master. All except
two losses were won in reverse.
The heuristic is simple but effective - the strategy in these endgames
is to push the opposing king to the corner, then move the knight that's
blocking the pawn in for the checkmate while the pawn is free to move
and prevents stalemate. This patch gives SF the little boost it needs
to search the relevant king-cornering mating lines.
See the discussion in pull request 1939 for some more good results for
this test in independant tests:
https://github.com/official-stockfish/Stockfish/pull/1939
Bench: 3310239
This is a functional simplification of the Outposts array
moving it to a single value. This is a duplicate PR because
I couldn't figure out how to fix the original one.
The idea is from @31m059 with formatting recommendations by @snicolet.
See #1940 for additional information.
STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 23933 W: 5279 L: 5162 D: 13492
http://tests.stockfishchess.org/tests/view/5c3575800ebc596a450c5ecb
LTC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 41718 W: 6919 L: 6831 D: 27968
http://tests.stockfishchess.org/tests/view/5c358c440ebc596a450c6117
bench 3783543
This is non-functional. These 5 arrays are simple enough to calculate real-time and maintaining an array for them does not help. Decreases the memory footprint.
This seems a tiny bit slower on my machine, but passed STC well enough. Could someone verify speed?
STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 44745 W: 9780 L: 9704 D: 25261
http://tests.stockfishchess.org/tests/view/5c47aa2d0ebc5902bca13fc4
The slowdown is minimal even in 32 bit case (thanks to @mstembera for testing):
Compiled using make build ARCH=x86-32 CXX=i686-w64-mingw32-c++ and benched
This patch only:
```
Results for 40 tests for each version:
Base Test Diff
Mean 1455204 1450033 5171
StDev 49452 34533 59621
p-value: 0.465
speedup: -0.004
```
No functional change.
A simple idea, but it makes sense: in current master the search is extended
for checks that are considered somewhat safe, and for for this we use the
static exchange evaluation which only considers the `to_sq` of a move.
This is not reliable for discovered checks, where another piece is giving
the check and is arguably a more dangerous type of check. Thus, if the check
is a discovered check, the result of SEE is not relevant and can be ignored.
STC:
LLR: 2.96 (-2.94,2.94) [0.50,4.50]
Total: 29370 W: 6583 L: 6274 D: 16513
http://tests.stockfishchess.org/tests/view/5c5062950ebc593af5d4d9b5
LTC:
LLR: 2.95 (-2.94,2.94) [0.00,3.50]
Total: 227341 W: 37972 L: 37165 D: 152204
http://tests.stockfishchess.org/tests/view/5c5094fb0ebc593af5d4dc2c
Bench: 3611854
There was a simplification attempt last week for the tropism
term in king danger, which passed STC but failed LTC. This
was an indirect sign that maybe the tropism factor was sightly
untuned in current master, so we tried to change it from 1/4
to 5/16.
STC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 28098 W: 6264 L: 5990 D: 15844
http://tests.stockfishchess.org/tests/view/5c518db60ebc593af5d4e306
LTC:
LLR: 2.95 (-2.94,2.94) [0.00,3.50]
Total: 103709 W: 17387 L: 16923 D: 69399
http://tests.stockfishchess.org/tests/view/5c52a5510ebc592fc7baea8b
Bench: 4016000
Remove overlapping safe checks from kingdanger:
- rook and queen checks from the same square: rook check is preferred
- bishop and queen checks form the same square: queen check is preferred
Increase bishop and rook check values as a compensation.
STC
LLR: 2.95 (-2.94,2.94) [0.50,4.50]
Total: 27480 W: 6111 L: 5813 D: 15556
http://tests.stockfishchess.org/tests/view/5c521d050ebc593af5d4e66a
LTC
LLR: 2.95 (-2.94,2.94) [0.00,3.50]
Total: 78500 W: 13145 L: 12752 D: 52603
http://tests.stockfishchess.org/tests/view/5c52b9460ebc592fc7baecc5
Closes https://github.com/official-stockfish/Stockfish/pull/1983
------------------------------------------
I have quite a few ideas of how to improve this patch.
- actually rethinking it now it will maybe be useful to discount
queen/bishop checks if there is only one square that they can
give check from and it's "occupied" by more valuable check. Right
now count of this squares does not really matter.
- maybe some small extra bonus can be given for overlapping checks.
- some ideas about using popcount() on safechecks can be retried.
- tune this safecheck values since they were more or less randomly handcrafted in this patch.
Bench: 3216489
