Four wild handicaps for FM level player.

[lkaufman]

But I don't know what seven moves are best, or how to force an advantage greater than knight odds with it.

I imagine it would not be that hard to modify the Komodo search to have the first 7 moves be for white and going to the first 4 ranks. It would give you a decent estimate.

-Sam

If anyone wants to modify his own engine or Stockfish in this manner, let us know the results. I tried a few setups and got evals in the low 200s cp. Seems like White should be able to get more than this somehow.

[Jesse Gersenson]

Ke5 is hilarious.

This I'd like to see. I just called a local 2100-2200, very aggressive and a very strong bughouse player, also he's computer saavy. His first question to me was, "how long do I have to prepare?".

I would run just two rounds but against several opponents at once. So Komodo would give Ke5 odds and be playing a simul. Would give you some data points and be very entertaining.

[lkaufman]

Ke5 is hilarious.

This I'd like to see. I just called a local 2100-2200, very aggressive and a very strong bughouse player, also he's computer saavy. His first question to me was, "how long do I have to prepare?".

I would run just two rounds but against several opponents at once. So Komodo would give Ke5 odds and be playing a simul. Would give you some data points and be very entertaining.

I can only handle a single opponent here myself, but if you want to organize such an event where you live just let me know. I agree that the Ke5 handicap would be most entertaining, I'm just not sure if careful preparation might insure the win with no need for skillful play over the board by the human. They would have at least some days to prepare.

[Jesse Gersenson]

The other guys on my club would help run the tournament. We put on 8-10 events a year and I wouldn't be short of skilled hands.

Limiting it to 2200 players lowers the number of people I can get. If this is (mostly) for entertainment, what further comic concessions could be made for 1700-2100 players?

If you really want to put on a show (and get rich), have Komodo play from the regular starting position against a team of 1600s at 5+0 and draw!

[lkaufman]

The other guys on my club would help run the tournament. We put on 8-10 events a year and I wouldn't be short of skilled hands.

Limiting it to 2200 players lowers the number of people I can get. If this is (mostly) for entertainment, what further comic concessions could be made for 1700-2100 players?

If you really want to put on a show (and get rich), have Komodo play from the regular starting position against a team of 1600s at 5+0 and draw!

Any event like this that you want to organize is fine, but there is no commercial value in proving we can beat ordinary amateurs in normal chess just with time odds. For the king on e5 handicap it might be reasonable to lower the minimum rating to say 2000, since Komodo probably won't be using very many cores per game. The king can also be on d5, I don't think it matters much. If you make it any worse for Black it might just be a trivial win. We can offer some small prize per win.

[Aser Huerga]

A very interesting handicap would be a simultaneous chess match, in which a modified Komodo capable of run several games at once in a given time, could fight against a given number of players. I bet this would be of mass appeal for any chess club.

I think this idea would be a great product promotion for the engine involved, and increase significantly engine sales with proper media campaign.

I encourage any top engine author to put it into practice, I predict great success all over the world.

[Laskos]

A very interesting handicap would be a simultaneous chess match, in which a modified Komodo capable of run several games at once in a given time, could fight against a given number of players. I bet this would be of mass appeal for any chess club.

I think this idea would be a great product promotion for the engine involved, and increase significantly engine sales with proper media campaign.

I encourage any top engine author to put it into practice, I predict great success all over the world.

Against 2400 players and below Komodo can play >1,000 games simultaneously, and the main problem is time allocation close to clock tick and overhead. Is this really interesting?

There is an issue with the funny and enjoyable to watch king on e5 handicap. It seems no larger than f7 pawn (practically lost) and castling rights handicap, more adapted against GM than FM or weaker.

[Jesse Gersenson]

The other guys on my club would help run the tournament. We put on 8-10 events a year and I wouldn't be short of skilled hands.

Limiting it to 2200 players lowers the number of people I can get. If this is (mostly) for entertainment, what further comic concessions could be made for 1700-2100 players?

If you really want to put on a show (and get rich), have Komodo play from the regular starting position against a team of 1600s at 5+0 and draw!

Any event like this that you want to organize is fine, but there is no commercial value in proving we can beat ordinary amateurs in normal chess just with time odds. For the king on e5 handicap it might be reasonable to lower the minimum rating to say 2000, since Komodo probably won't be using very many cores per game. The king can also be on d5, I don't think it matters much. If you make it any worse for Black it might just be a trivial win. We can offer some small prize per win.

My point was not clear, perhaps and also it has nothing to do with this thread. But, the idea was to substantially alter Komodo such that it'll play like a 'typical 1600'. It's not a weekend project but it is a gold mine.

If it's entertainment you want, what about Ke4?

[lkaufman]

A very interesting handicap would be a simultaneous chess match, in which a modified Komodo capable of run several games at once in a given time, could fight against a given number of players. I bet this would be of mass appeal for any chess club.

I think this idea would be a great product promotion for the engine involved, and increase significantly engine sales with proper media campaign.

I encourage any top engine author to put it into practice, I predict great success all over the world.

Against 2400 players and below Komodo can play >1,000 games simultaneously, and the main problem is time allocation close to clock tick and overhead. Is this really interesting?

There is an issue with the funny and enjoyable to watch king on e5 handicap. It seems no larger than f7 pawn (practically lost) and castling rights handicap, more adapted against GM than FM or weaker.

I know that's what the engines say, it's just very hard to believe. I think most human masters would prefer White even giving knight odds with the Black king starting on e5. I don't know who is really right, but I'm sure that many people would be amazed to see Komodo beat even an FM at this handicap.

[thekingman]

Limiting it to 2200 players lowers the number of people I can get. If this is (mostly) for entertainment, what further comic concessions could be made for 1700-2100 players?

1. If you don't reveal all the handicaps in advance, you could put the king on e4 instead of e5 (evals too high to do with any prep, but without prep, I bet Komodo could do it against a 1700).

2. Give them many more moves, but specify them yourself (for example, start with "rnbqkbnr/pppppppp/8/8/1PPPPB2/P1NB1N1P/3Q1PP1/2R1R1K1 b kq b3 0 14", which can be reached after 14 moves by white).

3. Replace Komodo's non-royal pieces with all knights (nnnqknnn/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQ - 0 1). This gives up both exchanges and the bishop pair, as well as castling rights.

4. Komodo plays without d, e, f, or g pawns.

5. White starts without the b1 knight, Komodo starts without *any* pawns. The complete lack of pawns may hurt Komodo's long term chances too much for this to be playable, though.

6. This is getting just silly, but Komodo plays with an extra set of pawns instead of a queen (rnb1kbnr/pppppppp/pppppppp/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1 )

All of these evaluate to roughly +4, fairly suitable odds for an A-class player, in my opinion. Some of them may be easy to play for the human after an initial sequence that could be found by a computer. But if the goal is to have a fun match with the chess club, maybe you don't need to give them the positions in advance. Just tell them the starts will evaluate as roughly +4, and they'll have 30 minutes (without computers) to analyze the start positions (perhaps collectively) before the clock starts.