It is a long-standing tradition that I use Excel to try and read the tea leaves at the end of each season, although this is easier in some years than others. When going through the data I concluded as a player that I didn't play enough games. I then proceeded to conclude the same as a statistician! But there are still some insights from the numbers, so let's delve in.
This was the second season running in which I played exactly 30 games - somewhat less than in the past, although I find it hard to understand how I played as many as 88 in 2011-12 in my last year at school. Last year I concluded that "playing strong opponents in higher quality formats has been good for my chess, even though I have had fewer opportunities to play". In a way, my results this time around have confirmed it, as parts of the recipe were missing this time and the outcome was a disappointment.
Starting with an openings breakdown, I had the regulation similar performance with White (+7 =2 -5 at 2028) and Black (+7 =5 -4 at 2013). As White, I did somewhat better with 1. e4 (+5 =1 -2 at 2065) than other moves (+2 =1 -3 at 1975) but I won't read much into a population of that size. It is maybe worth noting though that the average rating of the people I played 1. e4 against was 1894, and for other moves 2012. On the Black side, despite having nominally become a 1. e4 e5 player I only actually used it once, and scored an imposing +6 =1 -0 with the Sicilian. That's partly because I was only getting the Sicilian lines I wanted, but lots of those were to avoid d3 Italians against weaker opponents which is why that huge score was still only a 2103 performance. I only scored one other win with Black, and even that was from a lost position.
Last year I was disappointed that I generated totally won positions out of the opening in five games, but still lost two of them. I got five again and put them all away this time around (incidentally, they all arose from White 1. e4 games) and had at least a better position out of the opening in 12 of my 14 White games, and even 8 of my 16 Blacks. These are improved figures from a year ago (though see the next paragraph) but I still only made a slight plus score out of the better positions (+7 =3 -5) and the paltry +0 =1 -2 I returned from worse positions in no way makes up for it. And one final observation on openings that caught my interest: when I had prepared and got the right line on the board, I performed at 2112 as White compared to 1980 in other games; as Black it made virtually no difference. My White repertoire is a lot wider, but therefore needs more work to keep on top of; perhaps these numbers reflect that.
Now onto the opposition - if I had to explain the decline from 2023-24 in one paragraph, it would be this one. Last season the average rating of my opponents was 2071; this year it was 1929. The latter number is very similar to the averages for 2021-22 and 2022-23, and in all three cases I had a considerable plus score but performed at just over 2000, whereas the outlier year was closer to 2100. I picked up on this 12 months ago, and the reversion wasn't intentional: the decline in Epsom's 4NCL results, playing for the strongest county team in the land, and being forced into the U-2000 section at the London Chess Classic all took their toll. I had a rating advantage in 21 of my 30 games, and whilst I scored +14 =5 -2 from them (one of those losses a time trouble disaster in a won position), I didn't beat anyone rated over 2000 all season and +0 =2 -7 when outrated was fairly disastrous, especially given that I scored remarkably well in those games last season. It's not surprising that I lost an edge from playing weaker opponents most of the time.
Looking at game length, the mean number of moves came down a little from 42 to 39 - the median was also 39, up one from last year. In the past there seemed to be some link between number of moves and results, but - in keeping with the surer touch in technical positions which I've noted elsewhere - I scored +2 =1 from the three games to exceed 50 moves and the R value for game length and performance is -0.09, meaning that there is essentially no correlation.
Lastly, what about game format? I played no league games and only three county games, so my appearances consisted almost entirely of a full 4NCL season and two big tournaments - the London Chess Classic U-2000 and the Kingston Invitational Swiss. I scored +5 =3 -3 in the 4NCL, a 2005 performance, and +7 =3 -6 in my two tournaments which equates to a 1998 performance. However it's worth distinguishing London from Kingston - the former was one game a day, had lots of preparation time and was played when I had recent games under my belt; none of those applied to the latter. I performed at 2095 and 1930 respectively.
What then should I conclude about a difficult season? My overall performance of 2020 is a step back from last season even if +14 =7 -9 looks good on paper. I think that I was led astray by doing well from only 30 games in 2023-24, which was probably achieved in spite of the game count rather than because of it. It's noticeable that my results only really dropped off from February, after which I only played the final two 4NCL weekends before the Kingston Invitational in August. More games, and particularly more games against strong opponents, are needed, and my quest to diversify my White openings has made me dangerously reliant on good preparation. To end on a positive note though, my old difficulties with longer games seem to have subsided, for now at least.
To try and act on my conclusions (and because I unknowingly moved to within walking distance of a club) I've begun playing in the Oxfordshire League, and top board in the first division has also ticked the strong opposition box: between that and the Coventry International, I've played five titled opponents in my first eleven games! We'll pick up the action with my league debut, for which I was immediately thrown in at the deep end.
No comments:
Post a Comment