--- title: How to Actually Review Your Games on Lichess (And Learn From Them) description: Read this post carefully as I've seen a lot of people analyzing their games completely the wrong way, but they miss out on some very important insights that actually lead to improvement image: https://image.lichess1.org/display?fmt=webp&h=550&op=thumbnail&path=ublog:u0Tqx2aM:0ETQvdkD.webp&w=880&sig=c6b6b66b2a3d2fd3faa0b09e7b44966cedce6f8b --- When most players finish a game on Lichess, they press “Analysis,” glance at the blunders and centipawn loss, maybe scan a few “??”s, and move on. But if you want to really *improve*, that kind of surface-level review won't cut it, and you need to treat each Rapid game as a training session. I’m currently rated around 2650 in Rapid, and I’ve found that my most meaningful progress didn’t come from blitz marathons or anything like that, instead it came from sitting down after a single Rapid game and studying it with full focus and intention. Not just looking at mistakes, but understanding decisions, going through alternative ways of playing it, and learning from the positions themselves. In this post, I’ll walk you through how I personally review my Rapid games. This is the time control I focus on most, because I believe it offers the best value in terms of its depth and amount of time spent on it. Shorter formats like blitz and bullet can be fun, but they rarely give you the space to reflect and grow. Let’s dive into the process, and then I’ll show you a full example from a recent Rapid game I played on Lichess. *** ### Step 1: *Don’t* Turn the Engine On Right Away Engines are a crutch. If you turn on Stockfish immediately, you’re outsourcing your thinking. First, try to understand *on your own*: * Where did the game start to become difficult? * When did I feel in control or out of control? * What psychological biases were influencing my decisions? * Was I playing the position, or just reacting to my opponent's moves? I often write down a few “memory impressions” from the game before reviewing anything. *** ### Step 2: Look at The Opening, But Don’t Obsess Over It Ask yourself: * Did I actually *understand* the ideas in the opening, or was I just following muscle memory? * Was there a better way to enter a familiar structure? In the game I was referring to, I faced the **Scotch Gambit** and chose quite a dynamic setup. It’s not something I play every day, but I had enough structural awareness to find solid, flexible moves. *** ### Step 3: Look for "Psychological Turning Points" Was there a moment I *sped up* unnecessarily? Trusted my instincts too much? Oh yeah, there were loads of those moments. One of the biggest improvements you can make is in your mindset. In this game, there’s a clear example where I started to push on the kingside aggressively with ...g5 and ...Rg6. The engine might not mind it, but *why* did I push there? What was I feeling? Was it strategic or psychological? Spoiler: I was partly trying to assert control, but also in a weird way it was a defensive setup. And it worked, but I need to ask whether it was *objectively necessary* or if it just felt natural. *** ### Step 4: Go Deep On *One* Endgame or Transition You don’t need to study the whole game equally. Choose one complex phase and study it like a detective solving a case. In this game, the **transition to the endgame after 30...Ke5** is a perfect training moment, as you'll see in a moment. There’s subtle pawn tension, king activity, and tons of practical decisions to be made. *** ## Game Review: *Lizard1 (Black) in a 10+5 Rapid Game* https://lichess.org/study/sicBFNa8/nEE9zFI3#10 In this game, I chose an offbeat line against the Scotch Gambit with **5...Ng4** instead of the mainline **5...d5**. Objectively, d5 is solid, but I’ve found Scotch players are often well-prepared there. Ng4 is less common but playable, and I knew enough to navigate the structure without exact theory. Already, this is a psychological decision: I wanted to steer the game into less-charted waters. This is often a good option when you're unsure if you remember the exact moves you need to play in the main line, or if you suspect your opponent could be well-prepared in it (if he plays the first moves very quickly and confidently, for example). After White castled, I played **6...d6** quickly. My logic was that if I take it slowly with **Be7** he's gonna play **Bf4**, defending the e5 pawn and hindering my planned d6 break. I thought that I either play d6 now or never. It also seems quite natural what I was afraid of: if I somehow delayed this break in the centre, White would kick my knight with **h3**, and my whole setup would collapse. But here’s the nuance: **Be7** was apparently playable. After **Bf4**, Black has the exotic idea of **g5**, attacking the bishop, and even **h5** after **Bg3** to expand on the kingside. It’s not very intuitive, but it works. That’s why post-game analysis matters: it reveals resources you may not have even considered during the game. https://lichess.org/study/sicBFNa8/nEE9zFI3#14 Then came **7\. exd6 Bxd6**. I took with the bishop, thinking it was the only logical move: develop and recapture. But taking with the **queen** would’ve allowed me to develop the bishop on a more natural square: **e7**, which blocks the e-file from checks, and preserved castling rights. Instead, in the game I lost the right to castle. Not fatal, as at least I'm up a pawn for now and the bishop on d6 is quite active, but it kind of shaped the rest of the game. https://lichess.org/study/sicBFNa8/nEE9zFI3#18 This **9...h5** move I played was, to be honest, a purely psychological moment. Everyone else in the database plays **Ne5** in this position, simply retreating the knight and attacking the bishop. It’s natural, solid, and facilitates exchanges, which is important when Black is slightly worse and under pressure. So why did I play **h5**? Honestly, I don’t know, and I also played it quite quickly. It felt like I was trying to “hold” the knight aggressively, maybe even emotionally, and I thought this idea just pairs well with my active bishop on d6 which can support the attack if White does take the knight. Later in the game, I ended up playing **Ne5** anyway, proving that the exchange was necessary. This is a good example of instinct leading you astray. White played **Nd2**, which is a good way to develop the knight, but then instead of the strong follow-up **11.Ne4**, aiming to trade my key bishop, without which my entire concept with knight on g4 hanging wouldn't make much sense, which means that I would have to return the knight to e5 under unfavorable circumstances, he went for **Nb3** to recover the pawn. Reasonable, and still leads to a pretty comfortable game for White, but it let me stabilize for the time being. https://lichess.org/study/sicBFNa8/nEE9zFI3#25 Now, after being pretty much forced to play **11...Ne5** and a few trades, I realized how badly I needed more exchanges. White had threats like **Ng5**, and my pawn on h5 was now a liability, and this is the moment I finally realized how stupid a pawn on that square really looks, as otherwise I could've just played **h6**, easily stopping any knight infiltrations. I played **Bb4**, aiming to exchange queens, which was crucial since I couldn’t castle and had to get my rooks into the game ASAP. Even though my bishop returned to **Bd6** later, the queen trade was a success. My next goal: exchange rooks on the open **e-** and **d-files** to ease the tension, as White is still in the driver seat. https://lichess.org/study/sicBFNa8/nEE9zFI3#31 Here I played **16...g6** as an attempt to bring the king to safety and activate the rooks. I do have tom say that **f6** followed by **g5** might’ve been stronger: not shying away from grabbing space on the kingside and controlling key dark squares (f6 pawn is important for neutralizing White's dark squared bishop if it comes to d4, something I struggled with in the actual game). It didn’t feel natural as it weakens the light squares, but it was kind of a computerish and creative way to get a safe space for the king to connect the rooks. Here’s a beautiful line I found in post-analysis if I did go for f6: * If White plays **Nd4**, I take. * After **Bxd4**, I play **Be4**, preventing **Bd5**. * Then **Bc6**, **Re8**, and exchange rooks. White still holds a slight edge, but it’s manageable. After **16...g6** White did play **Nd4** (though he could’ve pressed with **Ng5**, **Bd5**, or queenside expansion like **a4+b4**) and after exchanges, his bishop sat powerfully in the center. My king was stuck, my rooks inactive. I knew i needed a creative setup to get out of such turmoil. So I played: * **Rg8** * **g5** * **Rg6** This setup supported **Bd6**, activated the rook, and gave me defensive flexibility. The other rook came to **e8** to exchange. Suddenly, I had a coherent structure and White’s activity wasn’t yielding much. https://lichess.org/study/sicBFNa8/nEE9zFI3#40 This is where I think Rapid analysis shines: you spot creative plans under pressure and learn to trust them. Then, long story short, we exchanged everything under the sun: two pairs of rooks and bishops, and after **29...Kxd6**, we entered a pawn endgame. It should’ve been drawn, but I had two advantages: * My king was closer to the center. * My kingside pawns were more advanced. https://lichess.org/study/sicBFNa8/nEE9zFI3#58 I'd also like to point out that in a strange way, the fact that I didn't castle in this game actually turned in my favor, as my king ended up being more active. And after a few reasonable moves of brining our kings closer to centre, White made a mistake with the move **33\. b3**. He needed to play **f4** or **f3** to challenge my kingside. https://lichess.org/study/sicBFNa8/nEE9zFI3#65 The way he played allowed for a thematic break: **33...f4+**, and he declined the pawn with **Kd3**, but now he was running out of tempis, as he couldn't allow my king to enter through e4: he would almost immediately lose. I kept opposition with king staying on **e5**, and started pushing queenside pawns. https://lichess.org/study/sicBFNa8/nEE9zFI3#71 But then came a critical error, this time from my side: **36...c5** instead of **c6**. In pawn endgames, tempo is everything. Pushing two squares gave him counterplay with **Kc4**, grabbing my pawn and creating a passer, something I never should've allowed. After a few more-less forced moves, we reached a queen endgame on move 45. https://lichess.org/study/sicBFNa8/nEE9zFI3#89 I was up a pawn, but queen endgames are notoriously drawish. He had enough checks to hold, up until move **61**, when he blundered with **Qe1??**, allowing a queen trade and promotion. https://lichess.org/study/sicBFNa8/nEE9zFI3#121 Had he played **Qc1**, he might’ve held the draw. But time pressure got to him, and in the end I converted this to a win. *** ## Final Takeaways The goal of post-game analysis isn’t to punish yourself or gawk at brilliancies. It’s to find: * Your decision-making patterns * Your blind spots * Your psychological rhythm * And one or two key technical lessons If you do this even once per day after a Rapid game, you will **for sure** improve, and not just in Lichess ratings, but in long-term chess strength. *** **Follow me (@Lizard1 on Lichess)** if you want more blogs like this. My next post will most likely be: *"Why I Think 0+6 Blitz Is Better Than 3+0 (And What It Taught Me About Discipline)"* Stay solid. —Lizard1