How to Prepare for a FIDE Over-the-Board Opponent
The first time I played a FIDE-rated team match, I spent the train ride trying to guess what my opponent would play from his rating alone. He turned out to be a retired maths teacher who had been playing the same 1.d4 and the same Stonewall setup for about twenty years, and every one of those games was sitting in a public database. I found that out after the game. Online opponents are easy to scout because there's a username. I assumed tournament opponents were a mystery, but if they play FIDE-rated events their games exist, flowing from tournament bulletins into public collections. For decades, mining that trail was practical only for professionals with database subscriptions. This guide is the club-level version of that workflow, built on Chess Stalker's index of 11M+ FIDE-rated games. You search by plain name, and it's free.
For the screenshots I'll use the most looked-up profile on the site: Magnus Carlsen's FIDE report. Every panel below is from his real page, and your club rival's report has the exact same structure.
Look them up by name
Open Chess Stalker, switch the search to FIDE and type their name as it appears on the pairing list (or their FIDE ID, if you have the rating card in front of you). If they've played rated tournaments, you'll typically find dozens to hundreds of their games already collected, and the site builds the same scouting report an online player gets, from their tournament record.
Names are messier than usernames, which is the one real friction of OTB scouting. The results show the federation, the title and the FIDE ID next to each name, so check those before you commit: with a common name you can easily open the wrong person's report (above, the world champion is followed by an IM and two more rated Magnuses). If the record looks suspiciously thin, try a transliterated spelling; federations don't all romanize names the same way.
Once you're on the right profile, the header alone is worth a pause: ratings in all three rhythms side by side, and the size of their game record.
The rhythm gap matters here just like online: someone rated well above their classical strength in rapid lives on intuition, and someone whose classical number is the strong one will usually be better in long maneuvering games than in tactical chaos. Choose the fight accordingly.
Read their classical repertoire
Classical repertoires are narrower and more stable than online ones. Nobody experiments with a dubious gambit when the game lasts four hours and the rating damage sticks for months, so the line they played in their last three tournaments is very probably the line you'll face. That makes OTB preparation more reliable than online preparation, not less. The Openings card sorts their record into weaknesses to exploit and strengths to avoid, and you can flip it to whichever color you'll be facing them with.
When you read the card, the percentages matter more than the names of the openings. A main weapon scoring 55% and a secondary line scoring 35% tell you exactly which move order to prefer, and the lines that barely appear in their record are the ones they've rehearsed least. Then apply the same intersection logic as in the online preparation guide: one line per color where their numbers are worst and you feel at home. From here you can also open the full Repertoire tree and walk their moves branch by branch on a board, with their stats at every step.
Let Stockfish read their whole record
The Deep Analysis card runs their classical games through Stockfish 18 and reads the errors across the whole set, which is something you physically can't do replaying games one by one. Tournament players tend to have fewer but deeper habits than online blitz players, and a habit that survived years of classical play will survive until your game too. Look at where in the game their accuracy dips, and what kind of mistake repeats.
23% of errorsHow I read a card like this for a normal opponent: the phase with the lowest number is where you want the game to still be tense, and the critical-moment average tells you when. If their blunders cluster after move 46, don't trade down to a dead position on move 35; keep pieces on and make them live through their own bad zone. The problematic-piece stat sounds like a gimmick until you see someone whose errors are 40% knight moves, and then it stops sounding like a gimmick. Below the tiles, the same card lists the exact blunders Stockfish flagged, move by move and opponent by opponent, and the error heatmap maps every mistake onto the board: Magnus's errors pile up on d2 and f3, so those are the squares where you'd force the decisions. All of this is free, by the way; deep engine analysis of a full OTB record is not a paid feature.
Scout their online accounts too
Many tournament players are active on Chess.com or Lichess under a findable username, and that second dataset is gold for one specific reason: clock behavior. Classical databases barely show how someone holds up in time pressure, while a few hundred online blitz games show exactly that. The online scouting walkthrough covers that side step by step. When the two repertoires disagree, trust the classical one for your game and treat the online one as a preview of what they might be preparing next; people test new openings online before daring to play them at a board.
Rehearse the matchup
Knowing their weak line on paper doesn't mean you can steer a live game into it; that part I had to practice. Chess Stalker can build a Twin Bot out of everything it just analyzed: a bot that opens like them and drifts into their kinds of positions, rather than playing perfect engine moves. Drill your chosen line from step 2 against it until the moves come without thinking.
Honest note on what's paid here: the Twin Bot built from an online account is free to play, while the Twin of an OTB player is a PRO feature. If you'd rather not pay, the free fallback works fine: every board on the site has Stockfish 18 built in, so open the key position from your chosen line and drill it against the engine, both sides, until the plans feel like yours.
Walk their OTB repertoire tree
The shortlist from step 2 is usually enough, but when I want the whole picture I open the Repertoire section: a tree of every move they've played with either color, carrying games and win rate at each branch, and a board next to it drawing their most-played moves as arrows. Click into a branch and the engine evaluation follows you, so you can tell how deep a habit really goes and where the numbers actually turn against them.
Note the spread: 1.e4 and 1.d4 carry most of Carlsen's 4,478 White games, both above 70%, and even his sideline experiments score well: that's what a hard target looks like. On a club player's tree, look at the tail: a first move they've tried 17 times at 53% is a different opponent from the one behind their main weapon, and one click opens that branch on the board with Stockfish running.
Go deeper if you need to
Everything above is free. PRO adds the pieces that matter most for a long tournament: endgame statistics across all 14 types with every game linked (classical games reach endgames far more often than blitz, so this table earns its keep over the board), PGN downloads of up to 5,000 games per pull if you prefer to study in your own tools, and the Twin Bot for OTB players. For a single club match, the free report is plenty.
Frequently asked questions
How do I find the games of an opponent who isn't a titled player?
FIDE-rated tournament games get recorded and published: they flow from tournament bulletins into game databases. Chess Stalker indexes 11M+ FIDE-rated over-the-board games searchable by player name, so ordinary club and tournament players can be looked up the same way grandmasters are.
What if my opponent has no games in any database?
It happens with juniors, players returning after a long break, or events that only submitted results without game scores. In that case check whether they have a Chess.com or Lichess account. Many tournament players are active online, and an online repertoire is a strong hint of the over-the-board one.
How recent do the games need to be?
Weight the last two or three seasons most heavily: repertoires drift, and a line someone abandoned years ago is not a target. Older games still matter for style, because clock habits and endgame technique change much more slowly than opening choices.
Is scouting a tournament opponent allowed?
It is standard practice at every serious level; professional players employ seconds who do exactly this for months before a match. The only rule is that all of it must happen before the game. Consulting anything during play is forbidden.
Is any of this behind a paywall?
The core is free: searching any FIDE player by name, the full report with statistics, play style, Stockfish deep analysis, opening weaknesses and the repertoire tree. PRO adds endgame statistics across all 14 types, PGN downloads and the Twin Bot for OTB players.