When we have a fail-high of a quiet move, store it in
a Followupmoves table indexed by the previous move of
the same color (instead of immediate previous move as
is in countermoves case).
Then use this table for quiet moves ordering in the same
way we are already doing with countermoves.
These followup moves will be tried just after countermoves
and before remaining quiet moves.
Passed both short TC
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 10350 W: 1998 L: 1866 D: 6486
And long TC
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 14066 W: 2303 L: 2137 D: 9626
bench: 7205153
SMP case is very tricky and raises an assert in stage_moves():
assert(stage == KILLERS_S1 || stage == QUIETS_1_S1 || stage == QUIETS_2_S1)
So rewrite the code to just return moves[] when we are sure
we are in quiet moves stages.
Also rename stage_moves to quiet_moves to reflect that.
No functional change (but needs testing in SMP case)
Use MovePicker moves[] to access already tried
quiet moves. A bit of care shall be taken
to avoid calling stage_moves() when we are still
at ttMove stage, because moves are yet to be
generated. Actually our staging move generation
makes this code a bit more tricky than what I'd
like, but removing an ausiliary redundant
array like quietsSearched[] is a good thing.
Idea by DiscoCheck
bench: 9355734
This seems a die hard idea :-)
Passed both short TC
LLR: 2.97 (-2.94,2.94) [-1.50,4.50]
Total: 17485 W: 3307 L: 3156 D: 11022
And long TC
LLR: 2.97 (-2.94,2.94) [0.00,6.00]
Total: 38181 W: 6002 L: 5729 D: 26450
bench: 8659830
1/ eval margin and gains removed:
16bit are now free on TT entries, due to the removal of eval margin. may be useful
in the future :) gains removed: use instead by Value(128). search() and qsearch()
are now consistent in this regard.
2/ futility_margin()
linear formula instead of complex (log(depth), movecount) formula.
3/ unify pre & post futility pruning
pre futility pruning used depth < 7 plies, while post futility pruning used
depth < 4 plies. Now it's always depth < 7.
Tested with fixed number of games both at short TC:
ELO: 0.82 +-2.1 (95%) LOS: 77.3%
Total: 40000 W: 7939 L: 7845 D: 24216
And long TC
ELO: 0.59 +-2.0 (95%) LOS: 71.9%
Total: 40000 W: 6876 L: 6808 D: 26316
bench 7243575
1/ eval margin and gains removed:
- gains removed by Value(128): search() and qsearch() now behave consistently!
2/ futility_margin()
- testing showed that there is no added value in this weird (log(depth), movecount)
formula, and a much simpler linear formula is just as good. In fact, it is most
likely better, as it is not yet optimally tuned.
- the new simplified formula also means we get rid of FutilityMargins[], its
initialization code, and more importantly ss->futilityMoveCount, and the hacky
code that updates it throughout the search().
- the current formula gives negative futility margins, and there is a hidden interaction
between the move coutn pruning formula and the futility margin one: what happens is
that MCP is supposed to be triggered before we use the non-sensical negative futility
margins.
3/ unify pre & post futility pruning
- pre futility pruning (what SF calls value based pruning) used depth < 7 plies,
while post futility pruning (what SF calls static null move pruning) used depth < 4 plies.
- also the condition depth < 7 in pre futility pruning was not obvious, and it seemd
to be depth < 16 (futility_margin() returns an infinite value when depth >= 7).
Tested with fixed number of games both at short TC:
ELO: 0.82 +-2.1 (95%) LOS: 77.3%
Total: 40000 W: 7939 L: 7845 D: 24216
And long TC
ELO: 0.59 +-2.0 (95%) LOS: 71.9%
Total: 40000 W: 6876 L: 6808 D: 26316
bench: 10206576
This reverts commit 800410eef1 and instead increases
stack size.
I went through the old emails with Daylen that reported the
crash issue on Mac OS X and was fixed by 0049d3f337.
It was reported default stack size for a thread in Mac OS X is 8
megabytes while the patch that we are reverting allows to reduce
stack size at max of about 217KB, so the reason for the crash was
only marginal in MAX_MOVES value. On those emails Daylen also
hinted how to increase stack size for Mac OS X to 16MB.
So prefer to increase stack size to 16MB instad of re-inventing
the wheel and do our home grown stack as we did with the patch
that we are now reverting (it will remain anyhow in git history
for documentation purposes).
No functional change.
This greately reduces stack usage and is a
prerequisite for next patch.
Verified with 40K games both in single and SMP
case that there are no regressions.
No functional change.
And #ifdef instead of #if defined
This is more standard form (see for example iostream file).
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
A big simplification and removing of useless code.
Finished at 50% both at short TC (with SPRT) than
at long TC at fixed number of games:
ELO: -0.14 +-3.4 (95%) LOS: 46.8%
Total: 15206 W: 2836 L: 2842 D: 9528
bench: 5059948
Most of the time we cut-off earlier, at captures, so this
results in useless work.
There is a small functionality change becuase 'ss' can change
from MovePicker c'tor to when killers are tried due, for
instance, to singular search.
bench: 4603795
Very good at long 60"+0.05 TC
LLR: 2.95 (-2.94,2.94)
Total: 5954 W: 1151 L: 1016 D: 3787
[edit: slightly changed form original patch to avoid useless loop
across killers when killer is MOVE_NONE]
bench: 4327405
Handling of History and Gains is almost the same, with
the exception of the update logic, so unify both
classes under a single Stats struct.
No functional change.
Templetize MovePicker::next_move() member function instead. It
is easier and we also avoid the forwarding of MovePicker() c'tor
arguments in the common case.
Suggested by Rein Halbersma.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Instead of by SEE. Almost no ELO change but it is
a bit easier and is a more natural choice given
that good captures are ordered in the same way.
After 10424 games
Mod vs Orig 1639 - 1604 - 7181 ELO +1 (+-3.8)
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Integrate TT_MOVE step into the first state. This allows to
avoid the first call to next_phase() in case of a TT move.
And use overflow detection instead of the bunch of STOP_XX
states to detect end of moves.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Follow the suggested Qt style:
http://doc.qt.nokia.com/qq/qq13-apis.html
It seems to me simpler and easier to read.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
This avoids search explosion in qsearch for some
patological cases like:
r1n1n1b1/1P1P1P1P/1N1N1N2/2RnQrRq/2pKp3/3BNQbQ/k7/4Bq2 w - - 0 1
After 9078 games 20"+0.1 QUAD:
Mod vs Orig 1413 - 1319 - 6346 ELO +3 (+- 4)
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
It seems much worst in number of nodes seacrhed to reach
the depth and anyhow does not give any advantage to the
Onno's oroginal one.
So revert by now and perhaps readd when we find something
clearly better.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
After 12613 games at 20"+0.1 on QUAD
Mod vs Orig 1870 - 1863 - 8880 ELO +0 (+- 3.3)
So no performance change but it is a code semplification
and also is more easy to understand.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
Here the idea is to test probcut not only after bad
captures, but after any bad move, i.e. any move that
leaves the opponent with a good capture.
Ported by a patch from Onno, the difference from
original version is that we have moved probcut after
null search.
After 7917 games 4 threads 20"+0.1
Mod vs Orig: 1261 - 1095 - 5561 ELO +7 (+- 4.2) LOS 96%
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
