Dominoes on ESPN

People need to accept that not everything has mass-market appeal. And to bring chess to TV in any meaningful way would require changing chess more than anyone who cares about the game would want. Yes, it is unfortunate that this means there isn't much money in the game. It's also unfortunate that computers can now regularly beat the best in the world at the game. c'est la vie.

Hey Mig,

Dominoes? Seriously Dominoes? On ESPN? I mean I've played quite a bit of dominoes in my day... you dont have to 'comb the streets of NY' to find the best players, dominoes is luck of the draw and being awake. If ESPN can put on dominoes, why can they not put the world's most popular game on TV? Next thing you know there will be ESPN Euchere or Pinocle... You laugh now... Just wait and see.

I've been fighting for years to get my family's snakes and ladders league telivised here in the UK.

For some reason both ITV and the BBC refuse to answer my calls, even when I go by the name of Anomini Anonumous the third.

How about solitare?

Simply put... chess is not understood by the mainstream audience. It has an image problem of being an elitist board sport and hard to understand where the others do not carry that stigma.

What's wrong with that? Several attempts have been made, the fact remains: chess is not popular enough to be televised. Poker is. Scrabble is. So what?

Dominoes? Is this a late April fools gag? Little children play with dominoes; the only verbal sparring I can imagine is with Mommy when it's time to take a nap. Bragging and bluffing? What happened to urban folk shooting craps or playing street hockey like adults?

Little kids play WITH Dominoes, but PLAYING Dominoes is done by a widespread audience, probably mostly of spanish descent, and of an older generation.

But, no matter, ANY game that breaks into ESPN coverage, no matter how BAD ESPN has gotten (RE: Please get your ESPN head out of Barry Bonds rear - he has cheated for years) can only help those other great games that have yet to find the stage there.

I think it is good for chess. Eventually I see Chess For TV developing into a 1-hour show with 4-game blitz matches, players acting like those punks in Poker games slandering each other:

"My Name is Ivanchuk. I Must Break You."

I'll watch the Domino matches if they make it to TV.

I venture to say that not everyone knows how to play poker(definitely not with any sort of strategy) and that the rules are about as complex as chess rules.

I'd love to see chess televised. But obviously, it would have to be time limited. Add some trash talking and cash prizes to lessen that 'elitist stigma' and you could have something. I'm not sure if commentators would be any help though.

DP said:
>> I venture to say that not everyone knows how to play poker (definitely not with any sort of strategy) and that the rules are about as complex as >>Chess rules

However, with Poker, it is immediately obvious to 99.99999% of the people watching it on television who is winning the hand. If someone has pocket aces is going up against someone with pocket kings, you know who has the highest likelihood of winning the game. This is not true with Chess and never will be.

Do you think Poker would be popular on television if each hand required 2 commentators to analyze each card for 5 minutes so the audience could understand what is happening?

Can you name ANY sport that is popular on television where the average viewer can't immediately look at the television and decide who is winning? I can't.

Although the complexity of chess is indeed an obstacle to mass viewership of chess events as a sport, the last comment from Archie Bunker is nevertheless off the mark in arguing that chess can't succeed on TV because "the average viewer can't immediately look at the television and decide who is winning."

If you tune in to a baseball or football game in the middle, the only way you can tell who is winning is that the score up to that point is shown constantly or frequently on the screen.

It's already a trivial matter to do this for chess at weak levels of play. That is, if kids' or other low-level amateur events were being televised, just have software tote up each side's remaining men, and then display the good old material point-count as a "score" in the manner of baseball or football.

As for high-level chess competition, many of you reading this probably believe (wrongly) that software can already show "who is winning." The kibitizers' comments I've seen in various forums -- while doing live commentary on the US Championship on ICC for example -- indicate to me that the mass of woodpushers who follow chess on the 'net, think they can simply consult Fritz about any given position and if it spits out, "+0.65 +/-", that evaluation is at least as reliable as any human assessment of the position (NOT).

Still, it might not be long before chess software does achieve the degree of sophistication that woodpushers now wrongly attribute to it. If so -- if the "point counts" of POSITIONAL factors that underly engines' evaluation functions ultimately become as straightforward and reliable as counting material -- then simply displaying at all times the machine-calculated "score" of a game in progress, would eliminate the problem raised by Archie Bunker.

Still, I don't think that would make live chess coverage any more palatable to mass audiences, though.

a couple years ago my kid dragged me over to a TV set and said "look at this!"..hhmm. It was rugby. I didn't know the rules or understand the jargon but I immediately got into the beauty of the flow of play. The cameras picked up the REACTIONS (crucial!) of the players to the physical contact and the progression of the game. Between phases (?) of the game the commentators (Australian if that matters) were shown discussing what we had seen in a pub setting where they were quaffing down brews. I hate both American football and "soccer"..but now I go out of my way to watch rugby on the tube. It's a good scrap...and the producers are smart enough to let the expressions of the players tell the story. Some chess players would just naturally "sell" the action of a TV game and others would be boring to watch...just like poker.

Jon,

You just posted that your ICC rating is about 1800, and yet you are scathing of ignorant woodpushers who follow computer evaluations.

My experience watching commentary on ICC is that IM level players are usually, not sometimes but usually, confused and contradictory when giving opinions during the course of a game.

My experience is also that if Fritz has a self-claimed +.65 against you in the middle game, its pretty good at increasing that to a win.

The rating (playing strength) of computer programs indicates that their judgement deserves a little respect.

gg,

1) My comment about the unreliability of computer evaluations is made not on my own authority, but rather is a received opinion I took from professional-level players and writers of books about computers and chess. For example, reading Chapter 6 of How to Use Computers to Improve Your Chess, by Christian Kongsted, may help deepen your understanding of why I said what I said.

2) Doing live commentary is very challenging even for titled players. On ICC at least, just one or two commentators covering multiple games at the same time is the rule rather than the exception. Having to look at a position for only a few seconds and make a comment before moving on to the next game and the next, produces analysis of a quality comparable to other contexts where the time constraint is similar (such as playing 1-minute games). However, when the commentators discuss plans and general features of positions, I think most viewers tend to find their observations instructive and valuable. If they were as bad as the Dirt-bags say, then surely the broadcasts would not be as popular as they are.

3) If Fritz has a self-claimed +0.65 against me in any part of a game, I'm sure it would be more than "pretty good" at increasing that to a win. Likewise if Fritz evaluated its position at -0.65, -1.65 or even -2.65. It is widely accepted among people knowledgeable about these matters, that computers' playing strength has little to do with their ability to evaluate positions. Rather, that particular area (evaluations) is still recognized as one where the engines are still weak relative to their playing strength, and relative to the strength of top human players; an area where there is much room for the engines to improve.

4) Another recent thread saw the question of the relative strengths of computers vs humans get debated in great depth. I suggest you find and read through it, because one of the few things that both sides there seemed to agree on, is that computers' ability to overwhelm even the best humans rests on superior calculation. Planning, intuition, and judgment remain the domain of humans. Put simply, humans and computers think and play in entirely diffferent ways. Because computers beat humans, that may mean that computers' style -- heavy on calculation, light on principles -- is "superior" to the human style, on some overarching level (this was the key point of contention in that debate). Whether such "superiority" has any practical significance for humans trying to improve their chess -- that is, whether in general it is physically possible for human chess knowledge and teaching to shift to the computer style -- is another matter entirely. I forget which post began the thread that contains that debate, but the computer-vs-human issue was initially off-topic when introduced and it eventually came to dominate the thread.

Bottom line, computers' MOVE CHOICES and ANALYSIS (suggested/preferred lines) do "deserve a little respect." Computer EVALUATIONS do not.

Jon,

You are totally confused about basic vocabulary.

A computer chess program is an evaluation function, a weighted addition of the importance place don all sorts of aspects of the position, and supported by tactical analysis. This is not so different to human evaluation, but usually better.

The computer's choices are driven by its analysis resulting from its evaluation function.

"Discuss plans and general features" is not a synonym for "evaluate".

I am not sure you are right Jon. I just spoke with a GM about one position which I was sure should be better for white. The GM says Fritz 9 "is giving me ...Qe6 and saying it is equal and I tend to believe him in this type of position." I think there are certain positions where the computer still is poor at evaluation, but in most positions I guess that its evaluation is at least as good as an ELO 2600+ GM.

Not to knock America, but to state what I believe to be the truth:

It's not that chess is not marketable. It's that the way in which chess IS marketable (battle of intelligence between titans) is not appealing in the US. People here I think are not as interested in, and are a lot more comfortable with not, thinking.

I would think that a country full of people who are more comfortable not thinking would never be successful.

gg, I did not use "discuss plans and general features" as a synonym for "evaluate" -- which you will see if you go back and re-read my comment. That phrase was used within my specific response (#2) to your claim that "IM level players are usually, not sometimes but usually, confused and contradictory when giving opinions during the course of a game." And that particular response of mine made no mention of computers. Next time, please read before you write.

DP, I appreciate your point, and your judgment about these matters deserves some respect because I know you understand chess at my own level or better. gg's strength is unknown to me. Although his comment I have just quoted verbatim would seem to imply that he thinks he is stronger than most IM's, I'm going to regard that as a hollow boast on his part until proven otherwise.

However, I fail to see how the fact that a GM told you he trusts Fritz evaluation in one particular position where the computer's evaluation differed from your), is any evidence that "in most positions I guess that its evaluation is at least as good as an ELO 2600+ GM." It seems like a huge leap to go from the first proposition, to the second.

First of all, did the GM, before he saw Fritz's evaluation of the position, agree with yours (that it should be better for White)? Second, and more important, did the GM himself tell you that he thought that in MOST positions he would tend to trust Fritz's evaluation over his own or that of other GMs? Even if he did, I suspect that opinion would be in the minority amongst GMs generally; there are matters, after all, about which even GMs differ.

Again, reading a book about the strengths and weaknesses of computers my help clarify a lot of these issues for readers. I found the Kongsted book (mentioned in my prior comment) enlightening.

For example, that book's detailed description of how computer evaluation functions are constructed, makes it clear that an engine's evaluation process is "not so different to human evaluation" (as gg said) -- but, rather than being "usually better" (as gg said), the engine's result is usually WORSE.

That's because the engine evaluation by nature relies on a formulaic toting up of values the programmer has pre-assigned to various factors: doubled pawns, open files, "bad bishop" (in whatever way that is defined, which is a huge problem in its own right for a programmer), king safety, and so on down the line.

Human evaluation, of course, is far more flexible, thereby more realistic in that a human can value a seemingly identical factor differently in one position than another - perceiving important distinctions that are very hard for a programmer to allow for.

I would add the caveat that software continues to make progress, so that it's possible that some points in a book published as late as 2003 (i.e. Kongsted) might be no longer valid.

I just re-read two chapters of great relevance to this discussion: Chapter 3, "The Blind Spots of the Computer," and Chapter 6, "Computer-Assisted Analysis". I see that the most revealing illustrations of computer weaknesses presented in Chapter 3 were played in the year 2000 (i.e., 6 years ago): Anand-Tissir, Shirov-Fritz, Fritz-Kramnik, van Wely-Fritz, Scherbakov-Rebel Century. So it's possible that the types of problems that got the computers in trouble in those games have since been overcome due to improved software, ever-faster hardware, or both.

Note also I do admit the possibility (which I mentioned both in this thread and the prior one) that contined tweaking of engine software might in time raise the quality of both computer evaluations and computers' "understanding" of chess principles, to a level equal or superior to that of the top humans.

Upon checking FIDE ratings I see that Kongsted (the author of the computers-and-chess book I quoted at length in the preceding comment) has no title, is rated only 2245 at present, and his peak rating apparently is 2280.

gg may take the low rating as reason to reject anything Kongsted says. For me, however, it's enough that the guy put in the time and work to write the book, and a good publisher (Gambit) thought enough about the quality of the effort to publish it. Not that I'm taking it as gospel; the message I got from Kongsted is in complete agreement with what I've heard from the handful of strong players I discussed these subjects with.

Jon, His statement was that "Fritz 9 is normally very close to the objective truth in most positions." I haven't spent too much time with Fritz 9 so I don't know to what extent it is a real improvement by way of evaluation, which as you point out doesn't linearly increase its strength(the machine's biggest asset is still brute force). Among the few people I have consulted they have claimed that it is drastic.

Jon,

On other threads you pontificate about legal issues then say you are not a lawyer. Now you law down the law about computer chess and say you are 1800 on ICC.

You then say you don't actually have your own opinions so your low rating doesn't matter, you are just quoting someone who was 2280.

You can choose to believe Mr Kongsted, but that does not make the two of you final authorities.

In my opinion, there is a legitimate case for claiming the outcome of a computer evaluation (a numerical +/- score and a preferred move) is normally, with a good program, better than depending on all but a handful of the best human players.

Some positions, computers are still quite bad, many positions they are a lot better than humans.

Now you did twist things a bit:

** an engine's evaluation process is "not so different to human evaluation" (as gg said) -- but, rather than being "usually better" (as gg said), the engine's result is usually WORSE. **

The PROCESS can be similar (summation of general principles then do some calculation) but the RESULT can be different.

It is very clear than the engine's result is better, because it come to conclusions that beat humans.

The reference to Fritz 9 is relevant - this program is good - it plays like a top human without the blunders.

gg,

Unlike you, at least I do quote at least one authority other than myself. (Nor am I relying exclusively on Kongsted. I've spoken with titled players who voiced views identical to those in his book; I quoted him because the book deals exclusively with this topic.)

Nor do I pretend to be strong enough to issue global judgments about the quality of IMs' commentary, as you freely do. By now the legion of fans you are attracting on this thread (NOT) must be wondering whether you are an IM or a GM.

In fact I see a pattern here, that leads me to a surprising conclusion:

gg is really Bobby Fischer! That would explain a lot.

Now that you're been unveiled here, Bobby, will you join Mig's lineup of Chessninja commentators?

Jon,

You are a funny guy. First you present something as your opinion. When you are challenged you say its actually someone elses opinion.

The simple fact is that computers play good, so their evaluations must be good.

Go and ask your titled friends (dukes and lords no doubt) about that.

And now it comes out - no one is worthy to challenge Jon Jacobs except Bobby Fischer.

Actually, I pretty much take it for granted that all of us here base our about technical subjects largely on what we have learned about the opinions of others who we can safely presume are well informed (i.e., chess authorities, such as titled players).

All of us that is except gg, who makes quite clear in his latest post (and his earlier posts) that he bases his opinions on absolutely nothing.

I guess maybe that (rather than him being a title-holder himself, as I mistakenly speculated) is why he finds it so easy to ridicule IMs and their commentary, as he did earlier in this thread.

Typo in the above: First sentence should say "....all of us here base our OPINIONS about technical subjects..."

SUDOKU - A mind game for the masses....for people who are too stupid to play chess....

Play dominoes with the stars