29... Ne4?

"This should let White escape with a draw. Correct is 29...g6 30. Qd3 Ne4 [lets White off the hook, 30...Qa3 is Black's best chance to play for a win here--KEG] 31. Rb3 [this gives away the game, 31. c4 is the only way for White to hold the position] f6 32. Nxg6 Kg7 [even better was 32...Rg7--KEG] (Tarrasch)

While there are some holes in Tarrasch's analysis, he did recognize that Lasker's 29...Ne4? was a mistake and that 29...g6 was much better. Moran also champions 29...g6 in his notes on the game.

In fact, the best way for Black to win here was 29...Qe6! leaving White without resource whether he trades Queens or not.

Contrary to NM JRousselle, it appears to me that Marshall fully understood the ramifications of the position after 29...Ne4 and played excellently until his blunders on moves 41 and 42. Had Marshall not gone hunting for a non-existent win on move 41, he likely would have held the game. But, as Sally Simpson aptly notes, Marshall almost always sought to attack whatever the position and whatever material remained on the board.

After 29...Ne4, the position was:

click for larger view

30. Nxf7!

"!"--(Tarrasch)

31... RxN!

"!"--(Tarrasch)(Moran)

"Nothing daunted, Lasker makes a virtue of necessity. The sacrifice of the exchange at any rate gives him some attack; but against correct defense it should have yielded a draw at most." (Tarrasch)

31. QxR+ Rf8

click for larger view

32. Qb7!

32. Qg4?? loses at once to 32...Rxf4 (after which White's Queen is lost to a discovered check wherever it moves) and 32. Qa6? loses to 32...Qxf4+ 33. Kg1 Q3+ 34. Kh1 Nf2+ 35. Kg1 [35. Kh2 Qf4+ 36. Kg1 Nxh3+ 37. gxN Qg3+ 38. Kh1 Qxh3+ 39. Kg1 Qg3+ 40. Kh1 Rf2 with mate to follow after a few spite checks by White] Nd3+.

Hard to believe on this open board, but Marshall's move was forced--anything else loses.

32... Qxf4+
33. Kg1!

Again forced. If 33. Kh1 White gets crushed after 33...Nf2+ 34. Kg1 Nxh3+ (Moran). If now 35. Kh1 (35. gxN Qe3+ 36. Kh1 Qxh3+ 37. Kg1 Qg3+ 38. Kh1 Qf3+ 39. Kh2 Qh5+ 40. Kg1 Qe4+ 42. Kg1 Rf3 43. Qb8+ Kh7 44. Qh2 Rxc3 45. RxR QxR+ 46. Kg2 Qb2+ 47. Kh1 QxR leaving Black three pawns up in the Queen and pawn ending) Nf2+ 36. Kg1 Ng4 37. Qxd5+ Kh7 38. Qh5 Qe3+ 39. Kh1 Rf4 40 Rf1 Nf2+ 41. Kh2 (41. RxN QxR is also hopeless for White) g6.

Marshall's sharp tactical prowess allowed him to avoid catastrophe on his last three moves.

The position after 33. Kg1! was:

click for larger view

In the diagrammed position with which I ended my last post, Lasker played:

33... Qe3+

As will be seen, after this Marshall had chances to hold the game. Yet this move--so far as I have been able to discover-- escaped any criticism until the excellent post on this site ten years ago by Honza Cervenka, who points out a likely win for Lasker with 33...Nxc3! If then 34. RxN (probably best), following Honza Cervenka's notes, then Qxd4+ 35. Kh1 QxR leaving Lasker a pawn ahead in the resulting Queen and Rook ending after 36. Qxa7 Qe3. Alternatively, if White plays the inferior 34. Kh1, then--again following Honza Cervenka's analysis--34...NxR 35. Qxd5+ Qf7 36. QxQ+ KxQ 37. RxN Ke6 38. Rc1 [38. a4 looks somewhat better, but this does not detract from Honza Cervenka's excellent analysis--KEG] Rf7 [38...Rf2 is probably even better] and Black is a big favorite to win the Rook and pawn ending.

Bravo <Honza Cervenka>

But Lasker's 33...Qe3+, though theoretically inferior to Honza Cervenka's 33... Nxc3, created practical problems for Marshall that--in the long run--he was unable to solve.

The position after 33...Qe3+ was:

click for larger view

34. Kh2

Another careful defensive move by Marshall. 34. Kh1, as Wilson has shown, would have lost to 34...Nf2+ 35. Kh2 Qf4+ 36. g3 Qf5 (and if 36. Kg1 Black wins by 36...Nxh3+ 37. gxh3+ [37. Kh1 Nf2+ 38. Kg1 Ng4] Qf2+ 38. Kh1 Qf3+ 39. Kg1 Qg3+ 40. Kh1 Qxh3+ 41. Kg1 Rf5).

34... Qg3+

Even better was 34...Qf4+ 35. Kg1 Nxc3 36. RxN Qxd4+ 37. Kh2 QxR leaving Black a pawn up in the Queen and Rook ending (after White captured either the Black a-pawn or the Black d-pawn.

35. Kg1 Nd2
36. Qxd5+

click for larger view

36... Kh8

36...Kh7 was probably better, but a win for Black is difficult on either move.

37. Kh1 Nf3!

click for larger view

Lasker continued to find ways to make life difficult for Marshall. After 37...NxR 38. RxN Qxc3 A draw was likely.

38. gxN

What else?

38... Qxh3+
39. Kg1 Qg3+
40. Kh1 Rf4

click for larger view

41. Qd8+?

"By this and the following move Marshall evidently hoped not merely to save the game but even to win it. [The text is] a grievous error. His only chance was 41. Qh5 whereupon Black has nothing better than forcing the draw by 41...Rh4+." (Tarrasch)

"Marshall had a chance to trade his Queen for the Rook with 41. Qh5. [The] Computer thinks the position [then] is dead drawn." (Volcach on this site in 2015).

"An easy draw is 41. Qh5 Rh4+." (Calli on this site).

As the above comments reflect, Marshall could almost certainly have saved the game by trading Queen for Rook. The resulting Queen versus two Rooks looks like a draw.

Remarkably, however, even after Marshall's 41. Qd8+, he could probably have saved the game.

41... Kh7

click for larger view

42. Rf1?

"?"--(Tarrasch)

"??" (Cali)(Soltis)(Moran)

This was indeed a blunder, but what should Marshall have played.

"42. Rc2 must be played." (Tarrasch)

"A blunder which loses at once. After 42. Rc2! winning would have been by no means an easy matter." (Moran)

In fact, 42. Rc2 was also a losing blunder. Lasker would then have won with 42...Qxf3+ [not Tarrasch's 42...Rh4+ which allows White to draw with 43. QxR] 43. Rg2 Qe4 44. Rc1 Rf2 45. Rcg1 Rf5 46. Ra1 Rh5+ 47. Kg1 Qe3+ 48. Rf2 [48. Kf1 Rh1+ and mate next move] Rg5+ 49. QxR QxQ+ 50. Kf1 e3 51. Rd3 Qh3+ 52. Ke2 Qxc3 or 43. Kh2 Qh5+ 44. Kg1 Qg6+ (this brutal fork only works because the Rook is on c2).

As suggested by my last note to 42. Rc2?, the best--and probably only--way to save the game for Marshall was 42. Rb2! Then, after 42...Qxf3+ 43. Rg2 Marshall would have had a decent chance to survive: 43...Qe4 44. Rcc2 [now this Rook is not hanging] Rh4+ 45. QxR QxQ+ 46. Kg1.

But after Marshall's awful 42. Rf1?, he was polished off with dispatch by Lasker:

click for larger view

42... Rf5!

"!"--(Moran)

click for larger view

43. Qe8

Marshall could have delayed mate by a few moves with 43. Rb5, but the outcome was now clear.

43... Qh4+

click for larger view

"The Rook must mate next move." (Tarrasch)

These posts are wasted here. Have you ever thought about getting these into a magazine.

KEG covers each move, mentions better moves, then moves on without getting bogged down on something that was not played; it's a chess game story from start to finish reminiscent of Chernev and the player series published by Chess Digest. He's not put-off by an imperfect game as so many chess authors are. He understands the value of a well-placed diagram.

Hi Fred,

Agree, I think they deserves a wider audience, to catch it properly here you need to copy each post, paste it in Word doc, then play it out OTB.

It does have a Chernevian feel, which is no bad thing.

Keg's liberal use of diagrams helps but even so I fear a lot of these posts go unread and to be honest they are the best thing 'chess wise' on the site second only to the graft the Bio Bunch put in. (not forgetting Once's posts in POTD.)

Hi perfidious:

It seems everyone likes his posts (you like so much you posted twice.)

The slight errors in analysis actually promote a healthy response and everyone joins in - which is what a chess forum should be all about.

Wish he would reply, I've laid some groundwork for a possible hard print outlet, now waiting to see if he up for it.

Just leave it, it's a short, if Alyerkupp had done it then we delete it to save 69 meg of net space.

There is thought, will there ever be a day when the Internet will stop. Is there be a future moment just waiting to happen when it will be impossible to type any further because it is full up.

Chess also overlaps. In my final climb yesterday, there were numerous choices on how to pull myself up the mountain. Looking ahead is a key. One route may look easiest, but it may lead to trouble two or three steps ahead.

Hi Keg,

I went to the ballet once, they should get taller girls, that lot had to dance on their tip-toes all night.

And I have never felt the need to climb a mountain or walk up a hill for a view. Good scenery to me is a well executed Kingside sac-attack.

If interested in getting into print go to https://www.redhotpawn.com/ join (it cost nothing and no software download) they have a private P.M. system. I'm greenpawn - (hence the green pawn here...I was also my alias when I wrote for fanzines in the 70's/80s.).

There are plenty of "tall" ballerinas these days. My favorite ballerina dances with a 6'6" partner, and is taller than 6'2" partners when she is on pointe.

As for scenery, I don't climb for the view (I'm not one for views anyway) but for the adventure of scaling a mountain or wall, which I find almost as exhilarating at a fine King-side attack.

FTB enjoys camping, hiking, trout fishing and other wildlife. Coming face-to-face w/a human on a mountain path gets your ticker pumping. FTB is afraid of heights, but he certainly enjoys caves and waterfalls.

The annual Pike's Peak chess tournament in downtown Manitou Springs, Colorado is this coming weekend. However, quality hotel reservations will be hard to come by at this point unless one is willing to pay top dollar. It's a trip best planned a year in advance. There is plenty to see and do in the area (the area is long-accustomed to heavy tourism).

http://www.coloradochess.com/newtou...

Hi Keg,

They use to climb the Flodden Wall which is right next to where I live.

https://upload.wikimedia.org/wikipe...

They would put chalk marks on the wall to show the good hand holds.

One night me and my son got a ladder and put chalk marks near the top of the wall. Next day we watched these people falling off the wall and skinning their knees and hands trying to reach them.

Thank you for your encouragement.

I have looked at redhotpawn and thus far am a bit mystified. My computer skills are feeble, and perhaps I am missing something very obvious.