pawn game

LOL. Ahh, now I see where your position came from...  I completely spaced out on tonydal's analysis and the refutation of e3, which is Ke2. All is clear now. It really is an interesting ending! And all my earlier analysis wasn't entirely wasted (although tonydal's line is the critical one). I feel so much better, really!

But wait... now I'm back to having to admit that my intuition leading me to e3 was bad. I can't win.

Fascinating endgame and discussion :)

Yes, interesting endgame... this is why king and pawn endgames make me nervous OTB, you know one wrong move and you may be throwing the game away.  Even if you didn't find the best moves JG27Pyth, I'm sure in all that analysis you were solving legitimate positions and coming up with ideas that will help you in future king+pawn endgames.

Thanks orangehonda. I think my job is to make relevant mistakes which are interesting enough to inspire good players to correct them.

Certainly a niche,  JG :p

JG27Pyth, what's wrong with White taking a straightforward approach in your post 14 line? I think White has more than one way to win after e3.

40. Kg4 might have been better.

ADK

Oh, thanks! I completely missed that.

Although it's an intestesting line mine is irrelevant early... Tonydal's Ke2 line is an improvement for White that demostrates e3 fails for Black. 

But Tonydal's comment: "(and btw in JG27Pyth's 43 c5 would also win)" has me puzzled. When I first looked at it I thought I saw what Tonydal meant... but looking it over just now time i can't for the life of me see how c5 wins for anyone... now with my best analysis it draws... Tonydal, can you point out the c5 win, please? 

With c5, the black king has to retreat to stay inside the square, and the white king picks off the black pawns.

Yes, that's what I thought the first time too, but now I can't make it work... the Black king has just enough to defend... 

If you can demonstrate the forced win I'd like to see it.

Unfortunately, I think that 7. ... d4 draws. 

39 f3 has a few benifits

Yes. You've recreated exactly what I did, I first investigated 7...Ke6 and thought Black was cooked, but I think 7...d4 draws -- however some of the draws get down to K v K opposition and in another line I looked at the draw had both players promoting queens... so maybe there's some clever manuver for White that I've missed. 

@JG27Pyth      Enjoyed trading ideas with you.  I picked up a few pointers in the process.

What happens after 7...d4+ 8.Kf3? For example, 8...Kf5 9.c6 Ke6 10.Ke4 or 8...Kd5 9.Kf4 d3 10.Ke3 Kc4 11.c6 Kc3 12.c7 d2 13.c8(Q)+. I'm sure you've looked at this, but I can't see the draw.

Hey David, I think you're right (at least about the first line, I didn't go further)

... I remember looking very specifically at the line highlighted above and I didn't find the win.  But it seems pretty clear that it does win for White. It's not like it's my first mistake, but it still aggravates me! The win/draw hangs on single tempo ( I was able to validate this authoritatively with Nalimov Tablebases) ...  I have no idea where I threw away a White tempo but I must have (it's not that hard to do).... hmmm... actually, looking at the last position diagrammed below... I remember where I went wrong...

So I got to the last position diagrammed above (white to move and win/black to move and draw) and I analyzed White's Kd5 with the wrong plan! It's an instructive ending (for me and anyone else) I looked at Kd5 thinking White goes and wins the h pawn... (that line draws unless I screwed up there too!) The correct plan is for to use the opposition to win the b pawn. I've studied exactly that kind of king and pawn ending win... argh... so, maybe this discussion will help me apply what I've studied next time around.

OK. So it seems we pretty much know enough to get an accurate evaluation after each move and say which moves were mistakes, right?

OK, I just went and put it into Fritz. It found 38.Kf1 which I believe is a much better try.