I was reading the FIDE rulebook and got reminded of the game I played last summer. My rating is about 1700, my opponent's was about 1900.



The game looked something like this (I lost the notation already) and when he set up his "fortress", my opponent said "draw." (not a question) I had about 20 minutes on the clock, he had about 3. I started checking him around with my queen, but didn't really make progress. The second he got to 2 minutes, he called the arbiter. The arbiter said "play on", and watched as I continued checking him, eventually getting the queen behind the king, and forcing the rook to move. After that, I could pin and win the rook. My teammate (about 2200) was watching, he said he thought it was an obvious win, and my opponent was just making it up. I talked to the arbiter in the evening, he said that he wasn't really convinced about my technique, but my opponent messed up so I deserved the win. The arbiter did not intervene again through the game, except to tell my opponent to stop talking to me (he offered/declared "draw" about 4 times).

I'm just wondering here, what does "win by normal means" mean in Q vs R, or similar positions? My (1700) "normal means" would be checking the king around and seeing if I can make progress. I would do the same if the time control had had a 30-second increment, or if I had been the 2-minute person and he'd had an hour left on the clock. Does it mean that people with bad endgame technique cannot win games vs people with bad endgame defense technique? In particular, how much leeway should we give to the side that is defending? I guess the arbiter could see, for example, that my opponent knows the knight's pawn passive defense in R+p vs R, but other than a few other obvious cases, shouldn't the side with the extra material get a shot at playing a bunch of testing moves, even if they are not "normal means"?