“Then there was the man who drowned crossing a stream with an average depth of six inches.”
- W.I.E. Gates

What would we do without statistics?  Probably, think before crossing a stream or, more relevantly, making a move.  However, I am here not to bury statistics, but to offer them… for whatever purposes the unthinking masses out there might wish to use them… just remember… do no harm to others.

Before his  round 5 game, Jorge Sammour-Hasbun was debating whether to play 1.e4… or 1.d4… Sympathetically, I offered the following advice… “You know, of course, that in the league 1.e4… has the highest winning percentage.”  Whereupon, he promptly played 1.e4… and won his second Game of the Week this season… We also serve who only stand and watch.

General Statisitics

Ok, let’s get to the nitty-gritty of this article and take an uninformed, uneducated, and unnecessary look at what it all means.  Here are the initial figures to look at…

Based on 196 games played through round seven in 2008

(Diagram 1)

Opening   Scoring%   Win%   Loss%   Draw%   Games   Wins   Draws   Losses
1.e4         58%      42%    26%     32%      90     38      29      23
2.d4         55%      40%    31%     29%      77     31      22      24
3.Nf3        54%      29%    21%     50%      14      4       7       3
4.c4         50%      36%    36%     27%      11      4       3       4
5.g3         83%      67%     0%     33%       3      2       1       0
6.f4        100%     100%     0%      0%       1      1       0       0

Of course, you can get this kind of break down from any database… no big deal.  At first glance, my recommendation to Jorge was a bold faced lie.  1.e4… does not have the highest winning percentage… I should have said “most wins” but, I’m sure Jorge will forgive me, as we all know that the smaller the sample, the less accurate the statistics.

Also, we must consider that in team match play these numbers will not reflect the strength of the opening played, as many team oriented players have thrown themselves on their sword when the team could not afford a draw.

And what about transpositions?  Looking at the chart above, we could say that if you want a sure win, play the Bird’s Opening… never mind that there was only one game in our sample… and that game transposed to a King’s Gambit (1.f4 e5 2.e4…).

So, how can we make sense of any of this?  For my purposes, I have decided…

1. Ignore the team match aspect and assume it all washes out in the end.
2. Ignore transpositions.  My reasoning here is a little more valid.  A first move such as 1.Nf3… for example, precludes certain defenses.  That is, you could play 1.Nf3 c5 2.e4… and you have transposed to the Sicilian, but cleverly avoided the 1.e4 e5 trap.  So, no matter how the game ends, the outcome is a result of the first move.  It may be that 1.Nf3… is the best move possible because it avoids certain defenses.  That’s my claim, and I’m sticking to it.
3. It is important to look at sample size.  Ergo, 1.f4… can be dropped from any consideration for best move.

Given the above, we can now say with preliminary innaccuracy that 1.e4… is the best opening for White, 1.d4… is second best, 1.Nf3… and 1.c4… are about tied for third best considering the small sample size and 1.g3…, although it actually scores best, due to small sample size, I will consider this irrelevant and rank it fourth and best in the category of others, which includes 1.f4… (the anomaly).

That’s my judgement looking at the initial picture above.  Therefore, Jorge played the correct move and I deserve half of any reward for the GOTW he won.

The Analysis

Now, I will begin my analysis with groups 3, 4, 5 and 6… Why?  Because we all know the only two valid openings are 1.e4… and 1.d4… just kidding… I want to look for trends in the small sample first and see if there is any useful knowledge in this before I waste my time on the larger sample… here goes…

Let’s first classify openings into the following first move (for White) categories…
e-pawn, d-pawn, c-pawn, Nf3, other (g3, f4, a3, etc…).

In the first seven rounds, there were a total of  29 games in which White began by playing Nf3, c4, g3 or f4.  The record was (11-7-11)  (win-draw-loss)… or, an overall score of 50%… good for Black… and less than expected for White.
**************************************************

1.Nf3…

Fourteen games began 1.Nf3… the overall record was (4-7-3)… again meaning (4 wins, 7 draws, 3 losses).  Quite a good record.  But, we all know that 1.Nf3 openings tend to transpose to those of the e, d or c-pawn variety.  The breakdown by transposition is shown below…

(Diagram 2)
Nf3 opening transpositions

Opening Record

Maroczy Bind              0-1-0
Queen Pawn game           0-1-0
Queen’s Gambit Declined   0-0-1
Slav                      1-0-2
Old Indian                0-1-0
King’s Indian             0-1-0
English                   2-2-0
Reti                      1-1-0

All this is according to ECO classifications as determined in my database.  However, I tend to think of the English as an Nf3 type opening, but to keep things simple, I’ll go along with ECO.

If we now group these according to the five categories, we get…

(Diagram 3)
Opening Record

e-pawn                    0-1-0
d-pawn                    1-3-3
c-pawn                    2-2-0
Nf3                       1-1-0

Just glancing at diagram 3, tells me that Nf3 is probably a good opening, (1-1-0) and fairly safe, although I would need to see a bigger sample.  Also, when it transposes to a c-pawn opening, English (2-2-0), it seems extremely effective.  But, the eye-opener above seems to be that when it transposes to a d-pawn opening it… well… it just stinks (1-3-3).

Of course, there was only one e-pawn transposition to the Maroczy… a draw is nothing much to speak of… so let’s ignore it.

Initial judgement here… don’t transpose to d-pawn.

If you think about it, the d-pawn transpositions will be of the type… 1.Nf3 d5 2.d4… or Indian defenses like 1.Nf3 Nf6 2.d4 g6 where White has blocked the f-pawn and the Samisch is not available.  This does seem to say that committing the King Knight to f3 early allows Black to direct things in a more favorable way.

Conclusion: 1.Nf3… is safe and strong… stick to the English or Reti opening. (3-3-0)
**************************************************

1.c4…

Eleven games began 1.c4… with an overall record of (4-3-4)… respectable, but a split is not the greatest result for White.  Again, c-pawn openings often transpose… let’s take a look…

(Diagram 4)
1.c4… opening transpositions

Opening Record

Maroczy Bind (e-pawn)     0-1-0
King’s Indian  (d-pawn)   1-0-1
English (c-pawn)          3-2-3

Obviously, not as many transpositions as 1.Nf3… The English has always had a reputation of being drawish.  However, here the wins and losses dominate the stats even though the overall outcome is level.  Even when it transposes to a KID the outcome is 50% scoring.  Does this tell us anything… well, it’s not drawish… just an equal unbalanced game that gives good winning, as well as, losing chances.  Not the drawish opening we all thought to find.

Surprising is the lack of Queen’s gambit transpositions… it is not really that hard, for example, 1.c4 e6 2.Nc3 d5 should do it.  But, it seems that USCL Black players don’t like this transposition.  At first, this seems strange since the statistics in 1.Nf3 show that the Slav opening works quite well.  But, the Slav is difficult to get to as 1.c4 e6 2.e4… would lead to the Caro-Kaan Panov-Botvinnik… an e-pawn transposition.  So, maybe the English takes away the strongest response to 1.Nf3… very interesting… but silly.

Conclusion: The English is not drawish… just equal, with good chances for both sides.
**************************************************

1.g3…

Three games began 1.g3 scoring (2-0-1) a small sample but impressive score until you realize that the two wins were scored by GM Nakamura, the highest rated player in the USCL by far, about +70 points.  So, maybe this stat doesn’t mean much, but, let’s ignore the facts and continue blindly as before.

(Diagram 5)
1.g3… opening transpositions

Opening Record

English                   0-1-0
Bird’s (1.f4…)            1-0-0
Other                     1-0-0

The English opening was from Round 1, Boston-New York, IM Jay Bonin vs. SM Denys Shmelov and, oddly enough, reached the same Position as FM John Readey (SEA) vs. NM Mackenzie Molner (NJ) in round 3… for the heck of it, let’s take a quick peak…

IM Bonin started the game 1.g3… whereas, FM Readey began his 1.Nf3… both reaching this English position.  So, what’s the point.  Well, FM Readey continued with an immediate 7.d4… while IM Bonin kept things closed for now with 7.b3… and a later 9.d4… In both games it seemed possible to reach a Maroczy Bind (e-pawn) game… but, neither White player opted for this.  Both games ended in draws.

My intent is to point out the flexibility of both 1.Nf3… and 1.g3… The transpositions are numerous and the knowledge required to play such openings must be wide spread as the game can turn down paths that seem to be unconnected, i.e., the English vs Maroczy.  And it seems that this position my be required knowledge for players of both first moves.

What can we learn from this?  Well, if your going to play 1.g3… it’s good to have the skill of GM Nakamura… and you should also understand the 1.Nf3 and English transpositions.

Conclusion:  An unbalanced game with good winning chances for White… Study Bonin-Shmelov and don’t play this (or anything else) against GM Nakamura.
**************************************************

1.f4…

The only game that actually began 1.f4… transposed to a King’s Gambit after 1.f4 e5 2.e4…  see NM Joel Johnson (ARZ) vs. FM Jon Bick (TEN), round 4.  Nothing much to learn here… except never… never… allow your opponent to play the King’s Gambit.

Conclusion: 1.f4… wins
**************************************************

And last, but not least, what would the statistics look like if we grouped all four moves (Nf3, c4, g3, f4) together by their transpositions… here goes…

(Diagram 6)
Grouping after transposition for all Nf3, c4, g3, f4 Openings

Opening   Scoring%   Win%   Loss%   Draw%   Games   Wins   Draws   Losses
e-pawn       66%      33%     0%     66%       3      1      2        0
d-pawn       39%      22%    33%     44%       9      2      3        4
c-pawn       50%      38%    23%     38%      13      5      5        3
other       100%     100%     0%      0%       2      2      0        0

Conclusion:  Transposing to d-pawn is the pits.  But, the English is impressive.

Well, that’s the end of this tale… at least the Nf3, c4, g3, f4 part.  I should explain the statistics above a bit.  If you have a won-draw-loss record of 1-1-1. The Scoring % is 1.5 out of 3 or 50%… the Winning % is 1 out of 3 or 33.333% rounded to 33%… the losing % is also 33%… and the drawing % is 33% adding to 99% not 100% due to rounding error.  Just in case you were wondering why things don’t add up.

Anyway, I also should explain the title of my article, or series of articles, if I have enough energy…  You see…it all has to do with statistics… there’s an old saying…

“Statistics… are like a lamppost to a drunk… they’re used more for support than illumination.”
- Alfred Edward Housman

1.e4… next…  To be continued…