Sergash: Kaminer was 3 years older than Botvinnik, being born in 1908. So at the time of this game, he was 15 or 16 years old, while Botvinnik was 12 or 13 years old. Kaminer was a "Soviet study composer". Thirteen years after the actual game, he gave Botvinnik a notebook containing all his 75 studies. Because Kaminer apparently had problems with the Soviet regime, Botvinnik waited until after Stalin's death to reveal the existence of the notebook. Kaminer's studies were published in 1981, in Rafael Moiseevich Kofman 's Izbrannye etjudy (1981).
Apparently this game was part of a training match opposing two young prospects? Let's have a look at it!
<7.Bf4?!> Apparently played to increase control over the e5 square, but it loses White's small advantage and might even lead to a ⩱... Normal and popular moves are 7.Be3 (known since the game Reti vs J Krejcik, 1914) and the more recent 7.Bg5 (Juraj Nikolac - Ivo Bajec, 7th Yugoslavia Championship in Sarajevo 1951, round 6, draw), both maintaining a ⩲.
<7...b6?!> creating weaknesses. Botvinnik's plan is working: Kaminer gets out of his opening plan! <7...e5! 8.dxe5 dxe5 9.Be3=> taking the pawn on e5 would mean to lose the one on e4... <Jaroslav Riha - Joseph Crha, Czechoslovakia Championship 1961 in Brno, round 8, draw>.
<9.0-0?!> It is better to remove the bishop from f4, so it would not be exposed to an attack by the black pawn coming to e5: <9.Bg5! ⩲> to be able to answer 9...e5 with 10.d5. On any other black move, White could castle. Stockfish 8 - 64 bits POPCNT.
<9...Qe8?!> Better <9...e5! 10.Be3 exd4 11.Bxd4=> Oleg Maazykin (2008) - Viktor Perekrestov (1898), Azovskie Zori Open in Eisk (Russia) 2011, round 6, 1-0.
<10.Qd2?! e5!=> White was given another chance: <10.Be3! ⩲> Again, in order to answer 10...e5 with 11.d5. Stockfish 8 - 64 bits POPCNT.
<13.Bh6?! Rd8! 14.Bxg7 Kxg7 ⩱> Following his plan, but <13.Bg5!=/ ⩲> Stockfish 8 - 64 bits POPCNT.
<15.Nd5?!> The best was to simply unpin the bishop: <15.Qe3!=/ ⩲> Stockfish 8 - 64 bits POPCNT.
<15...Nxd5?! 16.cxd5 c6!=/ ⩱> Zaminer captures with the wrong piece! <15...Bxd5! 16.cxd5 c6▢ 17.dxc6 Qxc6 18.Qe2 ∓> Stockfish 8 - 64 bits POPCNT.
<17.Rac1? ∓> Botvinnik misses the simple <17.Bc4▢=/ ⩱> Stockfish 8 - 64 bits POPCNT.
<17...Qd7? 18.dxc6 Bxc6=> Are we missing something here? Is the game score accurate? <17...cxd5▢ 18.exd5 Bxd5 ∓> Stockfish 8 - 64 bits POPCNT.
<19...Rc8?> This is a blunder! <19...Qd6 20.Rfd1 Bb7=> Stockfish 8 - 64 bits POPCNT.
<20.Ba6? Bb7▢ 21.Qxe5+ f6▢ 22.Qb5 Qxb5▢ 23.Bxb5 Bxe4=> Why not immediately <20.Qxe5+ f6 21.Qg3▢ ±> Stockfish 8 - 64 bits POPCNT.
This ends the 1st part, covering the opening and early middle game. Overall, Kaminer played somewhat better chess than Botvinnik, but the position was completely equal after move 18. and then a blunder could have given White a clear advantage...