Stepping Up vs. Stepping Down

A meta-analysis of variance in play against stronger and weaker players

Opening

Playing chess without consistent analysis is one of the surest ways not to improve. Delving into mistakes, honing opening theory, and sharpening tactical play are essential to getting better independent of level. Drawing inspiration from classical games and deepening your understanding of fundamental principles will be more typical of Chess for Life, but not in this first piece.

An often-overlooked form of improvement is meta-analysis — examining patterns across hundreds, sometimes thousands of games — to determine factors that might be dictating your play.

I’ve always had a hunch that my chess varied against opponenets of different strengths, but the differences were starker than I would have thought — and may surprise you if you run similar self-analysis. In chess and in life, pulling your head out of the sand to see the imprints you’ve been making on a wider scale is critical to getting better — no matter the skill, no matter the timeframe.

Middlegame

One of lichess.org’s best features (chess.com has something similar) is “Chess Insights”, allowing you to explore various questions about your games to uncover otherwise hard-to-see patterns. I was particularly interested in the following:

How does my play vary depending on the strength of my opponent?

The following chart (dynamic link here) reveals a potentially surprising result, but is something I found pretty illuminating:

The far-left shows that, on average, I drop significant amounts of rating against “much weaker players” (those rated 200+ points below my own rating), with a steadily improving curve as my opponents get stronger. Against “stronger” or “much stronger” players, I actually find myself picking up a good amount of rating.

Given chess operates on an ELO scale, this does not mean that I’m losing a larger quantity of games against lower-rated opponents (as the next chart will show). What it does mean is that if I play a large sample of games against “much weaker players”, my dynamic ELO rating is very likely to decrease. If I win a game against a weaker player, I’ll pick up anywhere from [0.1 - 2.0] ELO points per game, but if I lose, I often drop [9.0 - 13.0] points per loss (even a draw is typically [-4.0 or -5.0]). The exact opposite effect is happening when I play against “much stronger” players; I lose the lion’s share of games (as we all do, since that’s pretty much the prescription when playing much stronger opponents), but I often find a way to win enough of those games to pick up my ELO. Here’s how my raw win-loss ratio breaks down across players of each type:

I clearly still win the vast majority of games against lower-rated players and take many more losses against higher-rated players. But the fact that I’m losing 31% of games to players 100+ points weaker than me tells me I’m losing just about all the ELO I’m picking up from winning 32% of my games against players rated 100+ points above me. And taking wins 18% of the time against players who are rated 2500+ (oftentimes National Masters and above), tells me that my own rating could be significantly higher if I tighten up my game against weaker tiers.

Endgame

Psychology changes how you play — and I clearly am the type of player who takes things for granted against supposedly “weaker” players. My biggest personal takeaway is that if I approach my own games against lower-rated players the same way I approach games against higher-rated players, the immediate benefits of raising my floor will be even more substantive than those from raising my ceiling. Both can be done, but to reach my goals I can’t afford such swings in play. That means focusing on every move, calculating with the same level of rigor as I would against a master, and not assuming my opponent is ever operating under shallow logic. There’s nothing to lose (except some immediate-term effort); playing more accurately, more often can only raise my game over the long-term.

As an individual, I think it’s equally important to be the type of person who takes care of the little things. Making sure your habits are in order and you’re doing the small, compounding work day in and day out makes a dramatic long-term difference. Even if you’re someone who feels they’ve solved some of the larger things in life—building a career, getting healthy, paying off a loan or mortgage—it’s still important to remember your keys and not leave your laundry in the wash (again). Once you’ve gotten back to basics and dotted your i’s, you’ll be 10x more effective when it’s time for you to lean into your strength: stepping up in big moments.

And if you’re the reverse — the type of person who gets psyched out when you have a big presentation or an important conversation, there’s opportunity to lean into your own strength: the basics. Breathe, exercise, get good sleep. Diligently prep and take your time to make sure you’re as ready as can be for the stuff that matters most. It won’t always work out, but each time it does, you’ll slowly build the confidence to take on something more.

Ultimately, the critical step is performing a self-focused meta-analysis in the first place. Only when you’ve accurately assessed your strengths and weaknesses from a more global vantage point can you build a gameplan for improvement. If you’re struggling with something, step back and look for ways to evaluate yourself objectively across relevant dimensions to unlock a path forward.

---

For the purposes of this substack series, I’ve decided to open-share insights across all my games on lichess.org (username is infinite_chess_bdl), and you can do the same under “Preferences —> Privacy —> Share your chess insights data —> With everybody”.

---