Daily iPhone chess programs releases

[Kaj Soderberg]

There is also another issue in analysis mode. The 3-minute barrier is lifted with the update, very good. But, when having passed those 3 minutes, no new plies are being displayed. I have tried this with a few positions, running them for over an hour. And yes, i would agree that multi-pv probably slows down the search which is not handy on slowish hardware. Food for thought for Christophe.

My best,
Kaj

[M ANSARI]

... selling 25 applications at .99 is much easier than 1 application at 25...

For the vast majority of businesses the reverse is actually true. It's much easier to sell one product at a higher price than multiple at a lower. Trust me, I've studied it. Think about it, would it be easier to sell 4 copies at 0.25c than 1 at 0.99c? I don't think so but it's the same logic.

Steve

With the iphone things are different. The customer base is gigantic, so much so that you get exposed to people that would have never thought they would buy a certain product. I would agree that it might be a good idea to put more thought into what would be the optimal sales strategy to take advantage of the "window shopping" customer and the more targeted customer. Maybe a lite version which has the main items functioning at one price point, and a full version with more capabilities at a higher price point. For most, simply playing chess against their phone would be enough ... for others, sophisticated database capabilities and optimized opening books and ability to play online or some other features might push them to pay more.

[tiger]

There is also another issue in analysis mode. The 3-minute barrier is lifted with the update, very good. But, when having passed those 3 minutes, no new plies are being displayed. I have tried this with a few positions, running them for over an hour. And yes, i would agree that multi-pv probably slows down the search which is not handy on slowish hardware. Food for thought for Christophe.

My best,
Kaj

I am going to look at this. I have not noticed it yet, but thank you for pointing this out Kaj.


// Christophe

[Cubeman]

There is also another issue in analysis mode. The 3-minute barrier is lifted with the update, very good. But, when having passed those 3 minutes, no new plies are being displayed. I have tried this with a few positions, running them for over an hour. And yes, i would agree that multi-pv probably slows down the search which is not handy on slowish hardware. Food for thought for Christophe.

My best,
Kaj

I am going to look at this. I have not noticed it yet, but thank you for pointing this out Kaj.


// Christophe

I have not noticed this problem, I can get 10 minutes analysis in all the positions I tested it.You can tell if the engine is still calculating by the spinning circular pattern symbol to the left or right of the flip board icon.Have you gone in to Settings/Analysis time limit option 10min per move?
I can't understand the reason for the time limit, the reason given is to save battery life but some people might not care about battery or have the unit connected to mains power so an infinite option would be welcomed.Also the number of lines should be optional if user wants to just show the best line.
One thing I notice with CT is that if unit is powered off with the physical on/off switch and then powered back on then CT does not resume the previous analysis info.Also if message comes on that battery is low then that also stalls the engine output.
Where as with StockFish I can power the unit off and StockFish is still calculating in the background so that when I turn back on at a later time I have all this new and deeper search results and have not had the screen on which drains the battery more than the app itself.

[tiger]

The App Store is a very special case. Usual economic constraints and rules do not apply.

Actually most of the REAL rules are not even known.

You've done an experiment by launching at one price and reducing, so you have a real data point. I'd be surprise if you've seen a x5 increase in unit sales.

I would also agree that the App Store's integration into the iPhone platform (and now the Mac) is phenomenal. It reduces the friction of buying. And I think the $1 economy is fine for products people stumble across. People think, "it's only a dollar - why not". But I think it's different for products that people search for, or ones that are somewhat specialized. The game of chess certainly has a "stumbled upon" market but it also has a "I'd like a decent chess program" market. IMO the first market is super crowded (and served well by Stockfish) while there is room for Tiger in the second. And I doubt the second group of customer would complain about a $5 or even $10 price point. Just my 2c.

BTW - quick question - in analysis mode there are multiple PVS. Is there a way to only have one PV? Am I right in assuming that by showing multiple PVS the search is slower.

Cheers,

Steve

I'm still experimenting and I'm just starting to understand a little bit about it. So I still do not know what the optimal recipe is.

I'm afraid that the optimal recipe may just be to spit a good looking functional product in 2 weeks, sell it at the smallest price point and move to something else. Tic-tac-toe, Four-in-a-row, whatever. If this happens to be true then I was wrong to spend all this time trying to make the best product I could. We will see. I hope I'm wrong.

Stockfish fulfills the needs of an infinitesimal fraction of people. It's a technical product made by a technical guy for people with the same set of mind. It does not work for people who want to discover chess or improve, it does not work for people with real chess needs. It works for people who read CCC (and to be polite I'm not going to mention what kind of percentage it represents relative to iOS devices users). It is downloaded because it's free. Easy. Put a price point on it and watch it get some impressive depth. In term of page ranking I mean.

I keep on reading CCC to keep in touch with the nerdy computer chess community but it is just a BAD habit of mine. Being influenced by what I read here is counterproductive in the sense that normal people have needs that have nothing to do with what is said here. I'm still trying to make everybody happy but this might be a mistake.


To your question about multiple PVs:
- having to compute multiple PVs is a disaster in term of search depth. The search is slowed down by a factor of 4 (approximately) when you display the 4 best moves. Or maybe it's just 3 times slower because hash tables speed things up a little bit, but you get the picture.
- You are right, I should add an option for a one-line only analysis. I add this to my todo list immediately. Thanks.


// Christophe

[tiger]

There is also another issue in analysis mode. The 3-minute barrier is lifted with the update, very good. But, when having passed those 3 minutes, no new plies are being displayed. I have tried this with a few positions, running them for over an hour. And yes, i would agree that multi-pv probably slows down the search which is not handy on slowish hardware. Food for thought for Christophe.

My best,
Kaj

I am going to look at this. I have not noticed it yet, but thank you for pointing this out Kaj.


// Christophe

I have not noticed this problem, I can get 10 minutes analysis in all the positions I tested it.You can tell if the engine is still calculating by the spinning circular pattern symbol to the left or right of the flip board icon.Have you gone in to Settings/Analysis time limit option 10min per move?
I can't understand the reason for the time limit, the reason given is to save battery life but some people might not care about battery or have the unit connected to mains power so an infinite option would be welcomed.Also the number of lines should be optional if user wants to just show the best line.
One thing I notice with CT is that if unit is powered off with the physical on/off switch and then powered back on then CT does not resume the previous analysis info.Also if message comes on that battery is low then that also stalls the engine output.
Where as with StockFish I can power the unit off and StockFish is still calculating in the background so that when I turn back on at a later time I have all this new and deeper search results and have not had the screen on which drains the battery more than the app itself.

I think it's because I save the current game state (position, etc.) and settings in this case, and in order to do this I must stop the engine.

However, and please accept my apologies in advance for this, I don't think I'm going to improve on this for some time.

I may add an option for a one-line analysis though. Steve have just asked exactly this, and both of you are right.


// Christophe

[Tord Romstad]

The name is not important. Look, you can have a horrible and nonsensical name like "Stockfish" and still be popular.

True, but actually the iOS version is named "Stockfish chess", in order to make it easier to find accidentally when searching for chess programs on the app store. I have several friends with iPhones who wanted a chess program, downloaded the most promising looking free program they could find, and discovered to their surprise that I am the author (most of my friends don't know I'm a chess programmer).

Thanks a lot for the new version! The only thing I'm missing now is iPad support.

Actually I'm wondering why you use Chess Tiger. You have developped your own chess program for the same platform...

Just wondering.

For the same reason that most of us on this message board uses more than one chess program: They are all different, and it is fun to play against a variety of opponents. In addition to Stockfish, I use Chess Tiger, Chess Genius, Hiarcs and Kakinoki Shogi regularly on my iPhone and iPad. I also have Shredder and tChess Pro installed, but I don't like them as much as the other programs, and rarely use them.

[Tord Romstad]

The game of chess certainly has a "stumbled upon" market but it also has a "I'd like a decent chess program" market. IMO the first market is super crowded (and served well by Stockfish)

If you could provide a list of features you would like to see added to Stockfish to make it satisfy the "I'd like a decent chess program" market, I would be very happy.

Of course I can't promise to implement everything people want, but I'll consider everything you suggest.

while there is room for Tiger in the second.

Absolutely. I'm as happy as anyone to see Chess Tiger for iOS, and I hope it will be a huge success. But we also need some good free and open source options, and at the moment there are far too few.

[Tord Romstad]

I'm afraid that the optimal recipe may just be to spit a good looking functional product in 2 weeks, sell it at the smallest price point and move to something else. Tic-tac-toe, Four-in-a-row, whatever. If this happens to be true then I was wrong to spend all this time trying to make the best product I could. We will see. I hope I'm wrong.

You are not wrong. Trying to make the best possible product is far more important and satisfactory than trying to make the biggest possible amount of money anyway.

Stockfish fulfills the needs of an infinitesimal fraction of people. It's a technical product made by a technical guy for people with the same set of mind. It does not work for people who want to discover chess or improve, it does not work for people with real chess needs.

Ironically, the impression I get by feedback from users is precisely the opposite. Casual players who know nothing about chess apart from the rules love the app. Serious players complain about the lack of advanced database functions, variations in PGN games, and multi-PV mode.

And by the way, as a user, I'm hardly a technical guy. I mainly use Stockfish to play against, and adjust the strength to such a level that it is a challenging, but beatable opponent to a patzer like myself.

It works for people who read CCC (and to be polite I'm not going to mention what kind of percentage it represents relative to iOS devices users). It is downloaded because it's free. Easy.

No argument there. Being free is a huge advantage, of course.

I keep on reading CCC to keep in touch with the nerdy computer chess community but it is just a BAD habit of mine.

Perhaps it is, but I am sure I am not alone when I say that I hugely appreciate that you are here. I hope you won't leave.

[tomgdrums]

I'm still experimenting and I'm just starting to understand a little bit about it. So I still do not know what the optimal recipe is.
.

Stockfish fulfills the needs of an infinitesimal fraction of people. It's a technical product made by a technical guy for people with the same set of mind. It does not work for people who want to discover chess or improve, it does not work for people with real chess needs. It works for people who read CCC (and to be polite I'm not going to mention what kind of percentage it represents relative to iOS devices users). It is downloaded because it's free. Easy. Put a price point on it and watch it get some impressive depth. In term of page ranking I mean.

I keep on reading CCC to keep in touch with the nerdy computer chess community but it is just a BAD habit of mine. Being influenced by what I read here is counterproductive in the sense that normal people have needs that have nothing to do with what is said here. I'm still trying to make everybody happy but this might be a mistake.



// Christophe

Wow. Tord has been nothing but complimentary to you and Chess Tiger, and unless I am missing some hidden sarcasm in your posts you have been rather impolite to him. In fact most of your responses to honest questions seem to be snotty and aimed to show how "put upon" you are by having to make your product. If it is that much of a downer for you, don't make it. Move on to Tic-tac-to as you say.