Rybka Chess Community Forum
For all the Freestylers here in this forum, please check the dates on InfinityChess
. The other tournaments you can check here
Need some feedback from someone who played in the first "welcome" tournament. Was it run smoothly? Any technical issues?
Nick Carlin played in it and told me of some problems:
>Unfortunately the server was beset by technical problems,
>even crashing once, and so we could not have the playoffs (Flyingfatman,
>Spaghetti Chess, Hoshad and Sniper) until another (TBA) day.
He also told me the GUI only works on 32 Bit Windows.
Right Lukas, I think from a Freestyling perspective the biggest problem is that currently games cannot be resumed following a disconnect by either party. Arno is aware of this and I'm sure they'll fix their technical issues. It wasn't a bad experience overall.
We're really pulling for Arno and the gang to get it together. We probably won't make our playing debut until 64-bit is up. I'm not even sure if Anson has downloaded yet.
Call me paranoid, but I suspect that lack of 64-bit support may be intentional, along the same lines as discouraging centaur and pure engine play. Remember, true freestyle players only use computers to validate their GM's moves, so an old laptop running XP-32 or earlier and Fritz 7 should be more than sufficient! :-D
Heh, that IS paranoid. So Freestyle must be exclusive preserve of titled players, eh? The chess aristocracy must prevail, eh? The chess barbarians must be held down; every obstacle must be put in their path!
Such OLD WORLD concepts!
If the sponsors could figure out a way to correlate title with freestyle results, I think they would do it in a heartbeat. The concept that freestyle requires additional skills, mastery of which can compensate for reduced OTB skills, has been grudgingly accepted, but not embraced.
Alan.. No, this your theory it is simply disproven of this:
you can play freestyle with Infinity GUI load on a laptop and engine (x64) on your desktop PC, in centaur fashions.
If you are poor as me, and you cannot buy yourself 2 PC, you made to lend the second PC from an friend or as I have served me from the my job firm.
All of my systems are running XP-64 or Linux 64, but I have borrowed an old laptop from my daughter with 32-bit Windows OS so that I can compete.
I still have the feeling though, that the lack of 64-bit support may not be accidental. Its hard for me to believe that a team of software engineers could have overlooked this aspect at a time when 32-bit processors are no longer being sold.
It's actually hard to write an interface which will work in 32 bits but not in 64. Anyway, practically, it's not a big issue. Installing a 32 bit OS is a tiny amount of work compared to everything else.
It's actually hard to write an interface which will work in 32 bits but not in 64.
I suspect they've accomplished this feat by restricting themselves to 32-bit DLLs. My office security system, office surveillance system, and DVRs had the same issues. I blame this all on Microsoft for not making 64-bit applications backward compatible with 32-bit DLLs.
Anyway, it would have been nice if while they were shamelessly ripping off the Chessbase GUI, they had also kept the CB GUI's 64-bit compatibility.
Installing a 32 bit OS is a tiny amount of work compared to everything else.
Yes, including getting myself to buy another copy of this PoS OS or trying to convince some Microsoft employee in India that it was necessary to upgrade one of my ancient single processor AMD systems to an Intel Quad core with all new components so that I can reuse one of the dozen XP32 licenses that have been gathering dust for the last three years. Its much easier to try to use my 12-year old daughter's laptop and hope I don't disconnect because of the huge IM volume.
Actually, the whole situation is interesting.
Generally, in a market with a strong existing #1, you get a foothold by differentiating yourself. Apple/Mac is a textbook example. Convekta is also going this route.
It seems that the InfinityChess guys will differentiate themselves by offering decent prizes. It's not really a business plan, but it might very well work and is excellent news for freestyle.
Very interesting and encouraging.
The "Grand Prix points" are an interesting way to discourage using multiple nicks and dropping the ones which score badly. However, it's not clear what those points are good for.
However, it's not clear what those points are good for.
Grand prix points are good for grand prix prizes! :-) For Freestyle World Cup it is qualification of #1-6 for Super tournament in 2010; for Freestyle Masters and Computer Challenge you have grand prix prizes for the best three.
Yes, I saw that, but there are no details (that I have seen) about this tournament, so it's hard to attach value to these Grand Prix points.
Most of the suspense will center on the question: What method will be used to justify not giving Vas his prize money? As I recall, there were no less than three tournaments where this was discussed, with three different justifications. And in the last tournament we learned that changing team members is a capital offense, but throwing games is ok! :-D
Wait until 31st october, when InfinityChess will publish its current Freestyle Rules! I´m sure, w´ll see a Vasik Rajlich chapter. :-)
Well come on, do YOU want to see Vas get his prize money? :)
So far, the rationals given have been nonsense. I would recommend disenfranchising Vas with a dress code, making it strictly illegal to compete while wearing any type of Che Guevara apparel.
Now you're talking. We need web-cams on all players to ensure that they abide by the no-Communist-clothing injunction. No Mao caps, no t-shirts with Che or Lenin, no Orders of the Red Banner medals, etc. Otherwise prizes shall be impounded and contributed to Radio Liberty.
This might not be sufficient. I would recommend that PAL stand up a "Freestyle Police" regiment, empowered to travel to the residence of the winner, and go through their closets and drawers looking for contraband!
Very good suggestion from the land of the free
I see no problem here. Policeman #1 of the world will become jobless in january. He will like the job. Contraband is his issue. :-)
You are certainly right about that. Also poking his nose where it shouldn't be.
Actually that's what made those tournaments espacially interesting :)
>What method will be used to justify not giving Vas his prize money? As I recall, there were no less than three tournaments where this was discussed, with three different justifications.
I think you are not quite right here: in the 8th Freestyle, the discussion was whether to give Iweta
her prize money...
Oh what's the difference? :-)
Multiple nicks are a completely non-viable strategy in an age of Linux clusters.
Wishful thinking on your part. For any scenario with a reasonable number of entries (~100) in a Swiss tournament, a fixed number of computers, and enough friends to use the computers individually or as clusters, I suspect that the optimal strategy always involves multiple nicks.
Yes, assuming you can assemble such hardware as is implied in "enough friends to use computers individually or as clusters". Assuming four entries and each one a smallish 16-core cluster, that's 64 processors. Anyone who can pull all that together is just going to have an advantage no matter what you do.
Well, if half of the entries in the tournament are Italian, you will know the fix is in! :-D
Seriously now, you have to give the ruthless Italian bastards grudging respect. Anybody who wins, after going through everything they had to go through, is definitely a bastard. Doesn't matter how they warped the rules, cut throats, gamed the system. Bottom line, they won. You can't take that away from them. Unless they are damned Bolsheviks.
As I've stated before, I admired the audacious plan and the execution was flawless. I was very surprised though, by the willingness to share some of the details and the feeble response from the organizer. But I have to admit this is not the first time where I've been in competitions where I've been surprised that my fellow competitors were not playing by the spirit of the rules.
Is the spirit of the rules vodka, tequila, rum, gin or something else?
Bombay, Tanqueray, Booth's, Seagrams, Beefeater, Steinhäger or something more exotic?
Beefeater for starters. After that, it hardly matters...
Always Martini of course! Shaken, not stirred! :-)
All of the above, plus a little more...a little mint flavored moonshine.
Classic! "Good taste is relative"...
"Good taste is relative"...
Maybe, but this is absolutely revolting! :-)
hahaha :) :) :) that's a good one. creativity can be important :)
> For any scenario with a reasonable number of entries (~100) in a Swiss tournament, a fixed number of computers, and enough friends to use the computers individually or as clusters, I suspect that the optimal strategy always involves multiple nicks.
If Grand Prix points are valuable enough (or, to put it another way, if the 'real' tournament is long enough), this may not be true for more than a tiny percentage of the overall number of rounds.
OK, I disagree and think a few half points could be very decisive, so its time for a simulation. I'll suggest a framework below, get inputs from others, and run a MATLAB simulation using Monte Carlo analysis based on a large number of trials with a given trinomial distribution. Finally, I'll publish the results for all to ridicule.
Here is my first cut at parameters for this simulation:
- 10 round Swiss system
- 128 contestants playing, all competing for every half point except for 4 contestants that are on a team which makes all of its decisions to maximize the good of the team. For clarity, lets assign this team a name, say Team Italy.
- Draw percentage of 60%, a 25% chance of a white win, and a 15% chance of a black win. (Let's start with equal probabilities for each team, but there would be no problem putting in separate probabilities for each team, given a method for resolving what this meant when two teams with different percentages competed).
I hope people will feel free to suggest improvements, particularly to the draw/win/lose percentages I've given, based on past freestyle events. I'll construct and run the simulation on Sunday.
This initial simulation won't address the Grand Prix situation, but it will still be useful as a model for the first freestyle event.
o_o... people could get a feeling that you have a lot of spare time...
Actually, I don't have any spare time so I hope this simulation won't take long. It doesn't strike me as being very complicated as its currently formulated...
So, are you going to throw dice to decide game results? Or a true random number service (as opposed to pseudo random) like random.org
I'll use MATLAB's rand function, which generates pseudo random numbers with uniform distribution over the range 0 to 1. If the simulation isn't accurate, it won't be the fault of the random number generator. :-)
Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill