OK, so, how do you go about making a decision like this?  We haven't followed Charlie's thought processes move by move, so we need to warm up a little, get ourselves into the game... A quick once-over the available tactical ideas tells us that the position cannot be 'solved' by a combination, so we have to see what is our proper strategy here.

Jeremy Silman talks about a pawn-pointing rule, saying you should look to expand or attack on the side where your pawns are pointing.  I don't think pawns always point quite so obviously as this suggests, nor is it a rule that cannot be broken, but if you are stuck it will give you a shove in the right direction.

Charlie's game, for example, seems a very good example of the pawn-pointing rule.  White's pawn chain c3-d3-e4 points forward towards the King's-side, and Black's e7-d6-d5 chain points to the Queen's-side.  This is quite clear and also rather cheering, for it seems that Black's earlier manoeuvring on the Queen's-side has resulted only in a weak square on b5, which is occupied by the piece which benefits most from such outposts, the Knight.  So,

So, to play on the King's-side... White can try to attack mainly with pieces with moves like Qd2 or moving both Bishops (er, somewhere) and playing Re1-e3-g3 and maybe (groan) Na3-b1-d2-f1-h2-g4... I can't see that going anywhere at all, certainly not at that speed.  Imagine putting those pieces there now, what threats would White have, even if Black did nothing for 10 moves? However, there is a familiar plan here of playing f2-f4-f5 and opening a King's-side file, where Black has less space.  So, f4 seems a very natural, even an obvious choice.

The trouble with the march of the f-pawn is that White's pieces don't fit in with that right now.  I would rather prefer the Re1 to be on f1, and the Be2 to be on d3, and the Nb5 to be on g3, to lend their weight to a discussion of the f-file.  So, pointing pawns notwithstanding, I wouldn't be pushing the f-pawn to f5. 

Let's see if we can go through this from scratch.  First, make a quick comparison of the two positions: compare pairs of pieces, see which one is better.  Then have a go at the pawn structures, what are they telling us?

Both Kings are castled and secure.  The positions of the Queens seem equivalent, centralised and poised but not yet active.  White's Rooks are on good squares but on closed files; so that's our first hint, White's Rooks will need an open or half-open file to work on before long.  Black's Rooks on the 'bishop' files are not badly placed but I don't know what Black is planning to do in order to open a file (the c-pawn cannot be moved and perhaps the f-pawn should not be).  And the minor pieces: White's good Knight and Black's good Bishop are very much more active than the rest, which look rather tame at the moment.  If the position opens up, White's Bishop pair will enjoy life much better. 

If we want to (half-)open a file, where can White look?  Black's advanced a- and c-pawns allow White to lever open either the b-file or the d-file by advancing and exchanging the pawn on those files; Black, faced with b2-b4 or d3-d4 may choose to (half-)open the a-file or the c-file by exchanging his pawns on those files instead.  All Queen's-side files, I note.  Now, at the moment, most of Black's pieces look well-placed to take advantage of open lines on the Queen's-side: they're either on that side or pointing that way.  Nonetheless, I can imagine someone pointing to the weakness of b5 and that handsome knight, and declare, White should be attacking on the Queen's-side.  There are prospects there, I expect, but I can't see it happening very quickly.  I'd want to get the Be2 into the weak light squares, maybe the Queen... I have a feeling that by the time I'd arranged all my pieces for attack (Rb1 Bf3 Qc2-c4 maybe) Black would have re-arranged their pieces for defence (...Nc7 threatening to exchange on d4 is one immediate idea).  Hmm, Na3-c4 and b2-b4 might make the backward b6-pawn feel sorry for itself... OK, let's not forget all those ideas, but nothing there is compelling.

Well, if neither side is quite appealing, White can turn to the centre.  The 'crouching' white central pawns do suggest a central expansion with d3-d4.  That immediately gives a sniff of the wild to three of the currently tame White pieces, namely, the Rooks and the Be2, all of whom suddenly have new prospects.  An argument against d4, of course, is that Black has sensibly parked a Rook on the same file as the White Queen, so that after d3-d4, ...c5xd4, the white c-pawn is pinned against the Queen... Currently, White still has three other units covering d4, so d3-d4 doesn't lose anything, but it would be nice to threaten to take over the centre by d3-d4 and be able to recapture c3xd4 with two pawns abreast.  Having had this idea, I can see that I can get the white Queen off the c-file with gain of tempo:Qd2 forces some sort of response from Black, like Kh7, when most of our earlier options are still available.  After which, I think taking space in the centre with d4 seems very logical.  I also think that is White's best short-term goal; long-term, it is hard to tell what a general space advantage can lead to, but this push seems to put pressure on Black's game and gives a definite (modest) advantage to White.  Medium-term... well, I can imagine playing d4-d5 (again with tempo) squashing Black even more; I can imagine playing e4-e5, opening lines in the centre... Just at the moment, White doesn't have enough covering e5 to make that much of a threat, but Bf4 or f2-f4 would definitely make it a contender.

Charlie's actual move was f4.  So far, I've listed two separate if not overwhelming reasons for this move: to open a file with f4-f5, and to support a pawn advance after d3-d4 and e4-e5.  Having thought of playing it, I can see another virtue in the move: parking that fidgety light-squared Bishop on f3, waiting for the liberating e4-e5.  This is the Bishop without an opposite number on the Black side, it's the Bishop which make use of the weakened light squares on Black's Queen's-side... so it should be doing more in life than getting in the way of the Rooks!

Overall, then, 22.f2-f4 was probably worth more comment than I gave it, but it was plausible, unobjectionable and had more than one point... And there were more important things to mention!  If you fancy a go yourselves, what would you comment on this whole game?  I'll show you what I sent Charlie in a week or two.

22.f4 Nc7 23.d4 cxd4 24.Nxd4 Nde6 25.e5 Nd5 26.Qe4 Nxe3 27.Qxe3 Qxa4 28.Bb5 Qa2 29.exd6 exd6 30.Nxe6 fxe6 31.Bd7 Qxb2 32.Qxe6+ Kh8 33.Bxc8 Qxc3 34.Bb7 Qc5+ 35.Kh1 Rxf4 36.Rc1 Qb5 37.Qxg6 Rf6 38.Re8+ Rf8 39.Rxf8+ Bxf8 40.Be4 Qd7 41.Rc8 Qxc8 42.Qh7# 1-0

[C] Charlie's notes. My notes on Charlie's game

Just to complete the story; were they anything like yours?

Complete game without notes for you to annotate:

[Event "Exeter vs. Met. Office"]
[Site "?"]
[Date "2008.01.23"]
[Round "?"]
[White "Keen, C."]
[Black "Ward, G."]
[Result "1-0"]
[ECO "B50"]
[PlyCount "83"]

1. e4 c5 2. Nf3 d6 3. c3 Nf6 4. d3 g6 5. Be2 Bg7 6. O-O O-O 7. Bg5 Re8 8. Re1
Qb6 9. Qc2 Nc6 10. Nbd2 Be6 11. h3 Rac8 12. Rad1 a5 13. Be3 Nd7 14. Ng5 Nf8 15.
Nxe6 Nxe6 16. Nc4 Qc7 17. a4 b6 18. Bg4 Ncd8 19. Na3 h6 20. Nb5 Qd7 21. Be2 Rf8
22. f4 Nc7 23. d4 cxd4 24. Nxd4 Nde6 25. e5 Nd5 26. Qe4 Nxe3 27. Qxe3 Qxa4 28.
Bb5 Qa2 29. exd6 exd6 30. Nxe6 fxe6 31. Bd7 Qxb2 32. Qxe6+ Kh8 33. Bxc8 Qxc3
34. Bb7 Qc5+ 35. Kh1 Rxf4 36. Rc1 Qb5 37. Qxg6 Rf6 38. Re8+ Rf8 39. Rxf8+ Bxf8
40. Be4 Qd7 41. Rc8 Qxc8 42. Qh7# 1-0

Complete game with notes for you to compare:

[Event "Exeter vs. Met. Office"]
[Site "?"]
[Date "2008.01.23"]
[Round "?"]
[White "Keen, C."]
[Black "Ward, G."]
[Result "1-0"]
[ECO "B50"]
[Annotator "DR"]
[PlyCount "83"]

1. e4 c5 2. Nf3 d6 3. c3 Nf6 4. d3 (4. Be2 Nbd7 (4... Nxe4 $4 5. Qa4+) 5. d3 b6
6. O-O Bb7 7. Nbd2 g6 8. d4 cxd4 9. cxd4 Nxe4 10. Nxe4 Bxe4 11. Ng5 d5 12. Bb5
Bg7 13. f3 Bf5 14. g4 h6 15. gxf5 hxg5 16. fxg6 a6 17. gxf7+ Kxf7 18. Ba4 Rh5
19. Be3 Nf6 20. Qd2 Qd6 21. Rf2 Rah8 22. Rg2 Rh3 23. Rf1 R8h4 24. Bc2 Nh5 25.
Bf5 Nf4 26. Bxh3 Nxh3+ 27. Kh1 Qf6 28. Rg3 Qf5 29. Bxg5 Nxg5 30. Rxg5 Qh3 31.
Rg2 Bf6 32. Qd3 Rxd4 33. Qg6+ Ke6 34. Qe8 Rc4 35. Qd8 Qf5 36. Re1+ Be5 37. Qb8
{Svidler,P-Kasparov,G/Fontys Tilburg NED 1997/1-0 (37)}) 4... g6 5. Be2 Bg7 6.
O-O O-O 7. Bg5 {Not sure what the idea was there: you're not going to take it
to remove the pressure on e4, are you? Probably best leave it where it was.} (
7. Re1 Nc6 8. Nbd2 Ne8 9. Nf1 f5 10. Ng5 e6 11. f4 h6 12. Nf3 Nf6 13. Ng3 Kh7
14. h3 14... Nd7 {Hernandez Perez,J-Cruz Abdou,J/TCh-ESP Div 2 2001/1-0 (27)})
(7. Nbd2 b6 8. Re1 Bb7 9. Bf1 Qc7 10. a3 Nbd7 11. d4 Rae8 12. b4 e5 13. d5
13... h6 {Cyborowski,L-Lagowski,P/59th ch-POL 2002/1-0 (28)}) (7. h3 Nc6 8. Be3
b5 9. Qc1 Re8 10. Bh6 Bh8 11. Ng5 a6 12. Qf4 Ra7 13. Nd2 Bg7 14. Qh4 e6 15. f4
Qe7 16. Bxg7 Kxg7 17. e5 dxe5 18. fxe5 Nxe5 19. Rxf6 Qxf6 20. Rf1 Nf3+ 21. Rxf3
Qe5 22. Qxh7+ {1-0 Rodriguez,D-Fong,J/St John 1988}) 7... Re8 8. Re1 Qb6 9. Qc2
Nc6 10. Nbd2 Be6 11. h3 Rac8 (11... Rad8 {Black should take advantage of
White's slow play and go for some sort of pawn break. ...b5-b4 is thematic but
he has a Queen on b6, so ...d5 looks like the one.}) 12. Rad1 a5 13. Be3 {
# "I started (too) slowly but safely" [CK] White's huddled pieces are solid
but not really creating any problems for Black, but over the next few moves
they start to uncoil from their shell. I've included some examples of more
aggressive treatments by both sides above.[DR]} 13... Nd7 14. Ng5 Nf8 15. Nxe6
Nxe6 16. Nc4 Qc7 17. a4 b6 18. Bg4 Ncd8 19. Na3 h6 20. Nb5 20... Qd7 {
# White has been making little gains here and there, much like Steinitz used
to do.} 21. Be2 {That was happy where it was?} (21. Qd2) 21... Rf8 22. f4 {
(sic)} 22... Nc7 {#} 23. d4 {"d4 was long delayed but eventually achieved at
exactly the right moment" [CK] "Better late than never! I'd prefer to make it
in a position when I could recapture with the c-pawn, but it's fine as it is.
[DR]} 23... cxd4 24. Nxd4 Nde6 25. e5 Nd5 26. Qe4 {Not sharp enough.} (26. Bg4
$1 {threatens the g-pawn}) 26... Nxe3 27. Qxe3 27... Qxa4 {
# Every game has its crisis...} 28. Bb5 ({I thought} 28. Nxe6 28... fxe6 29.
exd6 {immediately was simple enough; your move gives Black a chance to cover
e6 with the Q.}) 28... Qa2 29. exd6 (29. Nxe6 $1 $11) 29... exd6 $2 (29... Nxd4
$1 30. cxd4 30... exd6 $1 $17) 30. Nxe6 $1 {winning} 30... fxe6 $2 31. Bd7 $1
31... Qxb2 (31... e5) 32. Qxe6+ Kh8 33. Bxc8 Qxc3 34. Bb7 Qc5+ 35. Kh1 Rxf4 36.
Rc1 Qb5 37. Qxg6 (37. Rc8+ $1 37... Rf8 38. Rxf8+ Bxf8 39. Bd5 {
is slightly crisper, says Fritz, but it doesn't really matter}) 37... Rf6 38.
Re8+ Rf8 39. Rxf8+ Bxf8 40. Be4 Qd7 41. Rc8 Qxc8 42. Qh7# 1-0