|Feb-16-05|| ||Sizzle: Okay please tell me I'm missing something. Doesn't 28...Rxf7 win the rook?? |
|Feb-16-05|| ||sneaky pete: <Sizzle> It does, so obviously the score is wrong and white played 28.R7xf2 .. etc. |
|Dec-23-14|| ||poorthylacine: Here we see it, the difference between the match Rubinstein-Salwe and this one: in this game the inaccuracies are swarming...
Even so, the battle is of course fierce and interesting...|
|Mar-13-16|| ||zanzibar: <Sizzle>,<sneaky> yes, the game score here, and at 365chess is wrong.|
Using Vazquez's book on Blackburne, G-3, p15, we have
<28—TD x D> for White.
One have to properly account for Spanish Descriptive Notation, but the QR does the recapture = 28.R7xf2.
Wonder if Steinitz's <ICM> gets this right?
He uses 28.QRxQ so the same problem arises if one isn't careful distinguishing QR from KR.
|Mar-13-16|| ||luftforlife: According to Dr. Tim Harding:
"27. Rxf7 Qxf2 28. R7xf2 <Astonishingly ChessBases's database here has 28. R1xf2 which should lose a rook.> 28. . . . Rf8 . . . . "
Joseph Henry Blackburne: A Chess Biography (Jefferson, North Carolina: McFarland & Co. 2015) at 307.
Here's a link:
|Mar-14-16|| ||zanzibar: <luft> these games are easy to trace back. |
The original complete source is A.C. Vazquez's <"Chess in Cuba - Mr. Blackburne in Havana (1891)">.
Of course we might also assume the few games which show up in Graham's Blackburne book come from Blackburne.
All the games seem also to have been published in Steinitz's <International Chess Magazine - v7>, but I think it safe to assume Steinitz obtained the scores from Vazquez.
Vazquez is truly a seminal figure in the Cuban/Mexican chess world, similar to Staunton, in basically forging the first high-quality Spanish chess publications in the New World.
(I have to qualify that statement, since I don't have an exhaustive knowledge on the subject, but that's what I've learned so far. And Vazquez's influence is undeniable.)
OK, so how did <ChessBase> get it wrong?
Aside- once <ChessBase> got it wrong, we know how the rest world got it wrong.
The problem isn't the inaccuracy of the scores, it's the error-prone notation used. And it's not particular to Spanish either, as Steinitz's English notation suffers from the same effect, as does some of the French notation I've read.
The problem is using Q- and K-side designation of the pieces to distinguish ambiguity. As I can attest from first-hand experience typing in old scores, it is very easy to mess up.
And if I had the time, I'd hack SCID to make a visual hack on Q-side vs K-side for R+B+N. Then it would be much easier to correctly type in the moves.
(Note that <retireborn> suggested using real pieces and a bit of tape for this - which isn't a bad idea. But why can't <ChessBase> or <SCID> do the same? Easy enough to add an extra option.)
So, Dr. Harding is astonished? Well, I can't say I am. I see it as all too easy to happen.
Normally, when one screws-up making a mistake like that, you soon reach a point in the game where a move doesn't make sense. Then you have to backtrack and find the error.
You can get away with this approach more than you might imagine. But not always, as here, where the divergent movelists merge back together after 30.Rf6.
Of course, one would like to ensure absolute correctness as the moves are entered. But who's actually going to set up a board to play over the game. Or actually hack SCID.
No, you sit down and get to work, hoping for the best.
Of course, I like to run an engine as I enter the moves. But this isn't always good for picking up errors, as sometimes masters "leave moves on the board", especially the old masters. Plus, it's difficult to actually look at every move with an engine, since it slows one down. Generally, even when using an engine, you bang out a series of moves, occasionally stopping to access.
Well, it's a long sidenote, this post. But actually an important, and under-appreciated topic, imo.
|Mar-14-16|| ||zanzibar: Next, let's examine the option of using Harding as a source, if one has access to his material.|
First of all, one should mention three obvious benefits.
First, Harding uses algebraic notation, thus avoiding some of the big problems mentioned above.
Second, Harding lists his sources rather well.
Third, his level of accuracy is one of the highest around.
But, there's a disadvantage in the last point. Since he's so reliable, one might assume him 100% reliable. And that would be a mistake.
Even Harding makes mistakes.
And, in a sense, once you know the primary sources, you're in the same position that Harding is. (Of course, Harding is a professional, and will often have access to sources beyond someone like myself - at least today.)
There was an old post I wrote about epsilon-square accuracy. It's a simple idea - basically, humans make mistakes. Some are much better than others, but two average people can easily out-perform an expert if properly coordinated.
Have the two non-experts independently enter the games. Then, assuming the errors are non-systematic, compare the results.
Suppose Harding makes errors at 0.5% level, and is ten times better than the average worker's 5% rate.
When comparing the two worker's input, the two must simultaneously make the same error for it to go undetected. So, their combined error rate is 5% of 5% = 0.25%, which is half of the expert's error-rate.
So, instead of using Harding as a source, one really should use the primary source in order to enter the games. Then you can compare to <CG>, and if a discrepancy shows up, use Harding as arbiter.
Sure, it's more work than using Harding to check <CG>, but using the primary sources has the advantage of checking Harding too, or at least, being independent of him.
(The thing is that you don't really know the source of <CG>'s games; in principle they could have some from Harding. So again, if primary sources are available use them. Then you can add a proper Source tag to the pgn too!)
|Mar-14-16|| ||luftforlife: <zanzibar>: Thanks for your thoughtful and informative comments. |
As far as my providing some additional support goes, I simply figured that since you had compared the most venerable and reliable source(s) to the score here and confirmed that <Sizzle> and <sneaky pete> were correct and the score was incorrect, I'd include the quotation from Dr. Tim Harding not because of its tone (or because of his criticism of ChessBase), but because it confirmed <Sizzle's> analysis and your conclusion, based on your diligent research. (I'm definitely a firm believer in locating and examining the best sources available, starting with the primary sources.) I also figured that if Dr. Harding recently confirmed what you had found in the older, primary sources, and he saw fit to point out and to correct this particular, persistent error, as you and others have done, then the scales had tipped even further in favor of correcting the score.
I'm always happy to offer whatever help I can provide. Kind regards.
|Mar-14-16|| ||luftforlife: <zanzibar>: You're right on target in your comments about the difficulty in description notational schemas of tracking king's versus queen's pieces (in this case the rook). Algebraic solves that problem nicely, but primary sources in descriptive notation require transcription, and that can be tricky in this regard. Thanks again.|
|Mar-14-16|| ||offramp: A very interesting discussion that probably deserves to be on a higher-profile page.|
|Mar-15-16|| ||luftforlife: <zanzibar>: Your points about primary sources and protocols for score-verification are well-taken. Although I didn't upload the PGN header or moves for this game, I did take the approach you advocate (before you had posted your helpful message), and I did submit a correction slip using Dr. Harding as the arbiter, so to speak. I'm not sure whether there is an accepted standard of proof in this area, so I paid heed to journalistic and historiographic standards. I considered Dr. Harding's published annotation as providing the sort of corroboration or ratification of your diligently discovered seminal primary sources (though perhaps not corroboration sufficiently independent of them to constitute fresh evidence) that would constitute additional legitimate evidence weighing in favor of score-correction. Not sure where I may have gone wrong, if I did, but I'm always eager to learn.|
Your valuable and wholesome suggestion about one's adding a "Source" tag to the PGN header begs the question I've been asking myself since the beginning of my time here: how does one do that?
Here's a link to the PGN Upload page:
PGN Upload Utility
Although certainly I understand that providing a hard source (a reliable, published source, as opposed to a soft source, such as one from an Internet database without verifiable attribution to a hard source) would maximize likelihood of compliance with Guideline No. 9 ("No games of which the authenticity is questionable."), and is the best way (maybe the only way) to proceed, the sample PGN header shown for Bird-Steinitz shows no bracketed entry for the source. How might one enter the source in standard PGN format?
In a similar vein, should one actually insert into the correction slip itself a quote from, or link to, a specific hard source or sources (I would say a link to a published volume excerpted or previewed online would qualify), or is directing the administrators' attention to such hard sources quoted or linked in the readers' commentary sufficient?
I do see our discussion as productive and instructive (at least I'm learning much from it!), and as on-topic, for the rectitude vel non of the posted score always seems to me to be centrally important. If I've made any mistakes in submitting my correction slip, either in form or in content, I do apologize; same goes for citing to Dr. Harding, who, as I wrote previously, seems just the sort of arbiter to whom you reasonably and helpfully suggest one should advert to resolve conflict between the score posted here and the score recited in the primary sources. Thanks again.
|Mar-15-16|| ||luftforlife: Glad to see the Editor Notes!|
|Mar-17-16|| ||luftforlife: I should mention briefly (and at large) that while I am willing to attorn to anyone's characterization of a source as primary (or as secondary) when I've not seen the source, the temporal priority or contemporaneity of a source might make it original, but does not by itself make it primary. Neither temporal priority nor contemporaneity to an historical event by itself renders a source primary; the relationship between the describer and the described is the key. Of course primary sources need not be autobiographical, but they must consist of first-hand accounts or evidence of direct contemporaneous experience or direct acquisition of knowledge, accurately recorded and credibly recounted. This definition of primary source satisfies historiographic and journalistic standards as I understand (and do my best to follow) them.|
Here's a handy précis from a worthwhile institution of higher learning (not my own alma mater):
For examples of venerable secondary sources on Golmayo, including some featuring game annotations, see, e.g., Edward Winter, Chess Note 9137, March 5, 2015, "Golmayo v. Loyd":
|Mar-17-16|| ||zanzibar: <Lust> I missed seeing your elaborations, sorry but that sometimes happens.
I don't have time at the moment to give it the attention it merits, but I think it safe to say that Dr. Harding is one of the most trustworthy of sources.|
I've deferred pushing on <CG> on the issue till after the Candidates, but a post by <MissS> caught my eye on this subject. I intend a post later.
At the moment, I suggest keeping a local copy of all your submissions to <CG>, and to begin training yourself to add a Source tag to the PGN.
As for someone as reliable as Harding, he almost always (or even always?) quotes his sources. So if I got a game from him I might add a Source tag that looks like this:
<[Source "Harding - Blackburne / The Field 10 Aug 1872"]>
Actual example from <Z-base>. So you see I got it from Harding, and he got it from "The Field".
A good system, or at least, the best I could come up with for now.
Let me defer replying further for the moment. Cheers and thanks.
|Mar-17-16|| ||zanzibar: And oh yeah, please do submit your games to <CG> with the Source tag in it.|
Time to start training them too.
(They'll simply just strip it out, without complaint, for now. And later, if they continue to strip it out, you can expect some complaints - from me!)
|Mar-18-16|| ||luftforlife: <zanzibar>: Thanks for your posts. No need to reply further; we each of us have other irons in the fire. I appreciate your comments and suggestions. |
I always seek the best sources I can find, and, in keeping with my practice in all areas of endeavor, vocational and avocational, I always provide them whenever possible in my posts and correction slips, either directly (by quotation or citation and link, if available) or by ready reference. For certain games, hard sources are sometimes impossible to find online, and so resort offline to publications, including many that are rare and hard-to-find, becomes necessary. (For example, hard sources for certain of GM Planinc's games don't seem to exist online, and so I'll have to track them down by finding published sources offline -- which has already proved to be quite a challenge, and promises to remain quite a challenge for some time to come, given the obscurity of his later years and the scarcity of certain publications in which his games are most likely to appear.) If I had a healthy chess library (as you probably do), or local access to one, I'd be thrilled to dive into actual print sources in hard copy, but unfortunately I haven't those resources at my disposal just now, and so I make do with the computer. Not the best way to conduct research, and I respect those who do have or who do lay hands on (or who do receive legitimate digital copies for fair use of) hard sources.
Since I posed the question about "Source" tags in PGN headers, I've come to understand from my reading in the Bistro that <cg> doesn't use "Source" tags in PGN headers, though they are used (though perhaps without standardized format) in other quarters. Indeed, I've seen them in games inserted into kibitzes here (but not necessarily uploaded) by seasoned members. So, as I'd always wanted to do (but didn't quite know how to do), I'll just insert "Source" tags with the best formatting I can come up with (and your sample template is sensible and helpful in this regard) into the PGN headers of any games I seek to have uploaded (or just include them with games I insert into kibitzes if I don't deem the games likely to be uploaded), and let the chips fall where they may. Even if the "Source" tag is stripped out (a matter of shipping-and-handling surely beyond my control), at least the person reviewing my PGN-upload game-submission will see source information that might prove convincing enough, or at least sufficiently persuasive, to satisfy the relevant criteria for inclusion, whatever those criteria might be beyond those stated in the guidelines on the PGN Upload Utility page. I appreciate your contributions in this connection. Thanks for them.
P.S.: My handle is a play on words, but I do go by "luft" (or "lufty"). I chose my handle in part because I do believe strongly in luft-for-life to avoid back-rank mates. While I'm dimly aware of the antics of James Newell Osterburg Jr., I try to keep it G-rated here. No worries. Thanks again; good health to you; and kind regards.
|Mar-18-16|| ||zanzibar: Sorry about the typo, but you did pick a name that plays on <lustforlife>... like the Van Gogh movie (not exactly G, more like PG):|
|Mar-18-16|| ||luftforlife: <zanzibar>: Quite right! Best to you.|
|Mar-18-16|| ||luftforlife: As an aside, I do recommend Irving Stone's 1934 novel about Vincent van Gogh (upon which the 1956 motion picture starring Kirk Douglas was based). I had had that in mind when I chose my moniker, as well as the 1977 song with lyrics by Iggy Pop and music by the late great David Bowie. But sorry to stray off-topic.|
|Mar-18-16|| ||luftforlife: The Chess Player's Chronicle, No. 410, March 14, 1891 (London: W. Wray Morgan Jr.), Game 2016, p. 396, likewise recites "28 R (B 7) takes Q." Annotations from the "Daily News" appear on the following page; this move is not annotated. |
Here's a link:
|Mar-18-16|| ||luftforlife: Here is a link to the primary source <zanzibar> found and cited above, viz. Andrés Clemente Vasquez, El Ajedrez en Cuba: Mr. J.H. Blackburne en la Habana (Habana: La Universal 1891), Articulo Quinto, at 15:|
This volume is a collection of articles by A.C. Vasquez originally published in the "Diario de la Marina."
Thanks to <zanzibar> for this seminal and deeply enriching source, and for first clarifying the game score.
|Apr-09-16|| ||luftforlife: Lest there be any remaining doubt, the quoted game-score as Dr. Tim Harding recited it is correct -- just as is the quoted game-score as A.C. Vasquez reported it. The game score is also correct in the other corroborative sources offered. |
While games should of course be optimally sourced, concordance between a contemporaneous, reliable, primary source and a retrospective, reliable, secondary source merely confirms in reassuring fashion the rectitude of the consistently recited score.
If the score as Dr. Harding recited it in 2015 in his authoritative work on Blackburne had been used as a source, the score above would read correctly -- just as it would read correctly if the source had been the score as it was recited in 1891 by A.C. Vasquez in his seminal reportage. (I do not know if Vasquez enjoys temporal priority in this connection; if he was not the first to recite the correct game-score, he certainly was one of the first to do so.)
Without intending any criticism in the following observation, the Editor Notes no longer refer to Harding or to Vasquez -- the sources I had cited in the first correction slip I had submitted. I did not mean to occasion their inadvertent and seeming displacement by my submission of a seoond correction slip citing to The Chess Player's Chronicle from 1891 as an additional corroborative (and contemporaneous) source justifying correction of the game-score.
The opacity of sourcing and of the sourcing process can be and would be ameliorated by the use of source tags -- and perhaps by enunciation in the PGN Upload guidelines of which sorts of sources would be deemed most preferable or reliable. Certainly primary sources sounding the ring of authenticity are desirable; so are secondary sources bearing the imprimatur of reliability. Although sources for games, be they hard or soft, old or new, primary or secondary, may contain errors, it is meet to adduce a recently published reliable secondary source bearing a game-score at odds with one posted on a game page at <cg> -- especially in support of an originally or contemporaneously published primary source bearing the identical game-score at odds with one posted here.
|Apr-09-16|| ||zanzibar: <Luft> I think I followed your post up to the last paragraph...|
Let's break it down:
<The opacity of sourcing and of the sourcing process can be and would be ameliorated by the use of source tags ...>
Yes, the Source tag has the potential to be quite useful, agreed.
<-- and perhaps by enunciation in the PGN Upload guidelines of which sorts of sources would be deemed most preferable or reliable. Certainly primary sources sounding the ring of authenticity are desirable; so are secondary sources bearing the imprimatur of reliability.>
A topic for the Bistro, but I think contemporaneous sources, where available - either newspaper or periodical, are to be preferred. Ideally.
Especially those by a respected chess columnist.
Experienced biographers will know who (e.g. Hoffer, Steinitz, Helms, Stuanton, etc., etc.).
If no such source is available, or if exigency, a reliable contemporary like Harding is perfectly acceptable. And being in algebraic notation, might even be preferred(?).
A contemporary source, like Harding etc., will certainly be preferred if the contemporaneous is known to be incorrect. Of course, a Bistro biographer would what to know the details - will might require a comment in the game forum.
But the Source tag could be used to indicate such a case, similar to the above quoting of a source within a source, e.g.
<[Source "Harding Citation <replaces> Old Citation"]>
Or some-such. This is assuming someone is doing a tremendous amount of work tracking each game.
<Although sources for games, be they hard or soft, old or new, primary or secondary, may contain errors, it is meet to adduce a recently published reliable secondary source bearing a game-score at odds with one posted on a game page at <cg> -- especially in support of an originally or contemporaneously published primary source bearing the identical game-score at odds with one posted here.>
OK, now that I broke down your paragraph, I understand the last point.
Of course, having a Source is also useful for interested readers who might like to see the original annotations, or even to know what journal might have additional side stories or tournament description, etc.