Which engine will surpass Stockfish in the future

[抽刀断水水更流]

Although Stockfish is a program with joint effort,and it has been dominanting for many years,but top engines are also improving tremendously fast in recent years. Any engines from the poll is more likely to break Stockfish's barrier ? I would like to hear your thoughts.

[swami]

A new engine, not from the list above, will turn up out of blue one day and become the strongest. It's what had always happened in the past, isn't it?

With so many derivatives around, truthfully being acknowledged and given credit where it is due, and released to public either with source or without it. I wouldn't be surprised if one of them become the strongest. But yeah, that will take a long time.

[Rebel]

Doing a scaling test currently.

Code: Select all

|-------------------------|----------------|----------------|----------------|-----------------|
|         TC=40/10    ELO | TC=40/10   ELO | TC=40/20   ELO | TC=40/40   ELO | TC=40/80   ELO  | 
|-------------------------|----------------|----------------|----------------|-----------------|
| Engine     1 CPU        |    20 CPU      |    20 CPU      |    20 CPU      |    20 CPU       | 
|-------------------------|----------------|----------------|----------------|-----------------|
| SF15       61.7%    +82 |   58.3%    +58 |   54.8%    +33 |                |                 |
| SF14.1     54.1%    +31 |   53.3%    +23 |   51.8%    +12 |   running      |                 |
| SF14       48.5%    -10 |   46.3%    -26 |   48.7%     -9 |                |                 |
| SF13       35.7%   -100 |   42.0%    -56 |   44.7%    -37 |                |                 |
|-------------------------|----------------|----------------|----------------|-----------------|
| Dragon 2.5 66.2%   +113 |   58.0%    +56 |   58.7%    +61 |   running      |                 |
| Dragon 2.0 33.8%   -113 |   42.0%    -56 |   41.3%    -61 |                |                 |
|-------------------------|----------------|----------------|----------------|-----------------|
|                         |     Equals     |     Equals     |     Equals     |     Equals      |
|                         |  40m in 3m/20s |  40m in 6m/40s |  40m in 13m    |  40m in 26m     |
|-------------------------|----------------|----------------|----------------|-----------------|

Hence my vote went to Komodo.

[abgursu]

I pick Komodo, I would always pick Komodo between those 12 options. I feel like time of the leela has passed but who knows?
But I really feel like none of them. Maybe some new engine which will be revealed from the shadows... like AlphaZero.

[towforce]

Where's the option for "SF will be the last engine to ever be the strongest one"?

IMO we will reach the upper limit of the elo scale before a stronger chess engine comes along.

[DrCliche]

With the caveat that I've made some very wrong chess-related predictions in the past, my guess is it's going to be the first team to figure out a quick and efficient neural network based search algorithm.

[CornfedForever]

Where's thhe option for "SF will be the last engine to ever be the strongest one"?

IMO we will reach the upper limit of the elo scale before a stronger chess engine comes along.

+1
Badly designed poll.

SF has a much bigger team behind it than any other engine...and Dragon ( updates ) only ever seems to get close to the 'official' version of SF these days so it's often unfair to even compare a Dragon update with the current 'official version' of SF. More fair is to compare the strongest 'current version' with the strongest 'current version'. NNUE draws engines closer...but for the foreseeable future, it's going to be SF leading the way.

[Frank Quisinsky]

Engine with most aggressiness in early mid-games which can hold late mid-game and endgame Stockfish has.
No doubt about it ... that is Dragon by Komodo but many others have strong statistics here.
The SF advantage melts slowly away the higher the time control.

For beat Stockfish I think it need a new team of strong programmers.
Elo from Berserk, rofChade, RubiChess, Ehtereal, Seer, Koivisto ... others ... are sensational for programmers working alone or in small teams. How big can the advantage if programmers are building a cooperation? Team work is not a magic word, the advantage from team-work is fact.

[Uri Blass]

Engine with most aggressiness in early mid-games which can hold late mid-game and endgame Stockfish has.
No doubt about it ... that is Dragon by Komodo but many others have strong statistics here.
The SF advantage melts slowly away the higher the time control.

For beat Stockfish I think it need a new team of strong programmers.
Elo from Berserk, rofChade, RubiChess, Ehtereal, Seer, Koivisto ... others ... are sensational for programmers working alone or in small teams. How big can the advantage if programmers are building a cooperation? Team work is not a magic word, the advantage from team-work is fact.

I think that the stockfish's advantage is mainly hardware advantage.
For some reason that I do not understand people prefer to donate computer time for stockfish and not for different engines so stockfish has thousands of cores that work on tests every second.

There are things that I clearly dislike about stockfish but developers of stockfish do not care about them.

They do not care about mating in the fastest possible way and for them mate in 50 is the same as mate in 80(not in static evaluation but in evaluation of the final result of game when they test) so I still get stupid analysis from stockfish.

I know that white is winning but I want to see logical attacking lines and not stupid passive play that is slow in mating the opponent.

I would like to see a rating list when you start with many winning positions when mate in 20 gives more points than mate in 60 so stockfish is not going to be the leader and it does not deserve to be the leader with the slow winning lines that it gives.

My opinion that I cannot prove is that if stockfish change their testing method and give more points for winning faster then they could also earn some elo points and there is no benefit with the slow winning lines(I guess that in some positions stockfish may fail to make progress and get a draw from a winning position because of this stupid behaviour but these positions are not common enough for the stockfish team to pay attention to them).

[fen]1nb1kbn1/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQ - 0 1[/fen]

Stockfish_22073111_x64_avx2:
Found 145 tablebases
NNUE evaluation using nn-ad9b42354671.nnue enabled
1/1 00:00 20 20k +21.96 e2-e3
2/2 00:00 77 77k +21.96 e2-e3
3/2 00:00 367 367k +21.96 e2-e3
4/2 00:00 634 634k +21.96 e2-e3
5/2 00:00 907 454k +21.96 e2-e3 a7-a6
6/5 00:00 1k 626k +21.90 Nb1-c3 Ng8-f6 Ng1-f3
7/6 00:00 4k 1,301k +21.87 Ng1-f3 Ng8-f6 d2-d3 d7-d6
8/8 00:00 15k 1,876k +21.82 Ng1-f3 e7-e6 h2-h3 Ng8-f6
9/10 00:00 44k 2,088k +21.75 Nb1-c3 Nb8-c6 Ng1-f3 Ng8-f6 Nc3-b5 Ke8-d8 Nf3-g5
10/12 00:00 82k 1,961k +21.75 Nb1-c3 Nb8-c6 Nc3-d5 Ke8-d8 Ng1-f3 Ng8-f6 Nd5-e3 e7-e6
11/15 00:00 123k 1,988k +21.68 Nb1-c3 Nb8-c6 Nc3-b5 e7-e6 Nb5xc7+ Ke8-d8 Nc7-a8 Ng8-f6 e2-e3 Nf6-d5 Bf1-d3
12/16 00:00 305k 2,242k +21.53 d2-d3 Nb8-c6 Nb1-d2 d7-d6 e2-e4 Ng8-f6 h2-h3
13/14 00:00 409k 2,323k +21.53 d2-d3 Nb8-c6 Nb1-d2 d7-d6 e2-e4 e7-e6 Ng1-e2 Ng8-f6 Rh1-g1 Bf8-e7 f2-f4 Ke8-f8 g2-g4
14/17 00:00 814k 2,415k +21.41 d2-d3 Nb8-c6 e2-e3 Ng8-f6 Nb1-d2 e7-e6 Ng1-f3 d7-d5 Bf1-e2 Bf8-d6 O-O Ke8-f8 h2-h3 Nc6-e5 Nf3-d4
15/19 00:00 1,044k 2,457k +21.61 d2-d3 Nb8-c6 e2-e3 Ng8-f6 Ng1-e2 e7-e6 g2-g3 Bf8-e7 Bf1-h3 Ke8-f8 Nb1-d2
16/19 00:00 1,322k 2,485k +21.55 d2-d3 Nb8-c6 e2-e3 Ng8-f6 Ng1-e2 g7-g6 g2-g3 d7-d6 h2-h3 Bf8-h6 Bf1-g2 Ke8-f8 O-O
17/22 00:00 1,721k 2,421k +21.61 g2-g3 Nb8-c6 e2-e3 d7-d6 Bf1-g2 Ng8-f6 Ng1-e2 g7-g6 Nb1-c3 Bc8-g4 f2-f3 Bg4-e6 Nc3-b5
18/22 00:00 1,896k 2,422k +21.62 g2-g3 Nb8-c6 e2-e3 d7-d6 Bf1-g2 Ng8-f6 Ng1-e2 g7-g6 O-O Bc8-g4 f2-f3 Bg4-e6 e3-e4 Bf8-h6 a2-a3 Ke8-f8 Nb1-c3 a7-a6 Rf1-e1
19/24 00:00 2,159k 2,437k +21.62 g2-g3 Nb8-c6 e2-e3 d7-d6 Bf1-g2 g7-g6 Ng1-e2 Ng8-f6 d2-d3 Bc8-e6 e3-e4 Bf8-g7 Nb1-c3 a7-a6 O-O Nf6-d7
20/24 00:01 2,722k 2,450k +21.71 g2-g3 Nb8-c6 e2-e3 d7-d6 Bf1-g2 g7-g6 Ng1-e2 Bf8-h6 O-O Ng8-f6 Nb1-c3 Bc8-g4 f2-f3 Bg4-e6 b2-b3 a7-a6 e3-e4 Ke8-f8 Bc1-b2 Kf8-g8 Kg1-h1
21/25 00:01 3,299k 2,453k +21.77 g2-g3 Nb8-c6 e2-e3 d7-d6 Bf1-g2 g7-g6 Ng1-e2 Bf8-h6 d2-d4 Ng8-f6 c2-c4 Ke8-f8 Nb1-c3 a7-a6 O-O Kf8-g7 Qd1-b3 e7-e6 Rf1-d1 Kg7-g8 f2-f3 Kg8-g7
22/26 00:01 3,993k 2,470k +21.72 g2-g3 Nb8-c6 e2-e3 d7-d6 Bf1-g2 g7-g6 Ng1-e2 Bf8-h6 d2-d4 Ng8-f6 c2-c4 a7-a6 O-O e7-e5 d4-d5 Nc6-e7 Qd1-a4+ Nf6-d7 Nb1-d2
23/28 00:01 4,643k 2,444k +21.75 g2-g3 Nb8-c6 e2-e3 d7-d6 Bf1-g2 g7-g6 Ng1-e2 Bf8-h6 O-O Ng8-f6 d2-d3 a7-a6 c2-c4 Ke8-f8 Nb1-c3 Bc8-g4 f2-f3 Bg4-e6 b2-b3 Kf8-g8 h2-h4 Nc6-e5 d3-d4
24/29 00:02 5,055k 2,448k +21.75 g2-g3 Nb8-c6 e2-e3 d7-d6 Bf1-g2 g7-g6 Ng1-e2 Bf8-h6 O-O Ng8-f6 d2-d3 a7-a6 c2-c4 Ke8-f8 f2-f3 Bc8-e6 Qd1-d2 Nc6-e5 b2-b3
25/30 00:03 7,789k 2,445k +21.71 g2-g3 Nb8-c6 e2-e3 d7-d6 Bf1-g2 Ng8-f6 Ng1-e2 e7-e5 O-O Bc8-g4 f2-f3 Bg4-e6 d2-d3 a7-a6 c2-c4 Nf6-d7 b2-b3 g7-g6 Nb1-c3 Bf8-h6 h2-h4 Ke8-f8 g3-g4 Kf8-g8 Qd1-d2 Nd7-c5 g4-g5
26/32 00:03 8,663k 2,444k +21.71 g2-g3 Nb8-c6 e2-e3 d7-d6 Bf1-g2 Ng8-f6 Ng1-e2 e7-e5 O-O Bc8-g4 f2-f3 Bg4-e6 d2-d3 d6-d5 Nb1-d2 Bf8-c5 Nd2-b3 Bc5-d6 c2-c3 Ke8-f8 Rf1-f2 Nf6-d7 d3-d4 g7-g6 g3-g4 Kf8-g8
27/33 00:04 10,451k 2,454k +21.75 g2-g3 Nb8-c6 e2-e3 d7-d6 Bf1-g2 Ng8-f6 Ng1-e2 e7-e5 O-O Bc8-g4 f2-f3 Bg4-e6 d2-d3 d6-d5 h2-h3 Bf8-e7 d3-d4 Ke8-f8 h3-h4 e5xd4 e3xd4 Kf8-g8 c2-c3
28/36 00:05 12,853k 2,446k +21.67 g2-g3 Nb8-c6 e2-e3 d7-d6 Bf1-g2 Ng8-f6 Ng1-e2 e7-e5 O-O Bc8-g4 f2-f3 Bg4-e6 d2-d3 d6-d5 h2-h3 Bf8-d6 d3-d4 g7-g6 d4xe5 Bd6xe5 Qd1-d3 a7-a6 Qd3-b3
29/33 00:06 15,579k 2,444k +21.67 g2-g3 Nb8-c6 e2-e3 d7-d6 Bf1-g2 Ng8-f6 Ng1-e2 e7-e5 O-O Bc8-g4 f2-f3 Bg4-e6 d2-d3 d6-d5 h2-h3 Bf8-d6 d3-d4 g7-g6 Rf1-e1 e5xd4 e3xd4 Ke8-f8 Nb1-d2 Nc6-a5 b2-b3 Na5-c6 g3-g4 Bd6-b4 c2-c3

[Werewolf]

Where's the option for "SF will be the last engine to ever be the strongest one"?

IMO we will reach the upper limit of the elo scale before a stronger chess engine comes along.

^^ This.