developing an eagle eye for pawn forks that will never exist

the best pawn forks exist in alternate dimensions, along lines that your calculations tell you will therefore be avoided, allowing moves that seem risky on the surface. pawn forks are so powerful that anyone who can see them coming wont let them happen to begin with. 

but if you cant already see future pawn forks, and your opponent deftly avoids them, its easy to never make moves that take advantage of the threat. and that means not getting experience seeing the ones that never happen.

i am bad at seeing future pawn forks. i am bad at seeing them as a calculated threat to be avoided, rather than as an immediate threat to be handled.

so i want to dig into this example, to help my pawn fork vision.

i came upon this position today, where the calculations for my best next move should include seeing two different potential pawn forks.

the rook moving in was a mistake, but i didnt see it at the time. i thought, if i dont move my knight out of the way, the rook just takes, right? 

i cant protect it with anything. the f pawn is pinned, and the queen isnt safe on h4. i dont have a good counter threat or distraction, right?

i thought my best option was to retreat my knight and hope to avoid starting the skirmish in the center too soon. if i could get away losing only a pawn, that seemed like my best option. there is only one immediately safe square to go to.

unfortunately, that move subjects me to a very foreseeable pawn fork:

this fork gives me no choice but to engage in a losing battle on e5.

looking at this game afterward, trying to see whether this was just a losing position to start with, i was surprised to see that there had been a way to save my knight!

bishop b4? why? 

it threatens the rook, but the rook just takes my knight, right? why is the computer saying white would deflect with c3?

if you guessed pawn fork, you have successfully understood todays theme.

once the bishop is out of the way, the d pawn is uncovered and can move two entire spaces to get to e5 in time.

yes, isnt it obvious now:

or maybe not so obvious. if a pawn were on d5 wouldnt the bishop just take it?

its hard to see 3 moves ahead to where the queen could recapture, since in the starting position theres two pieces between it and d5. its a pile of 3 pieces in a row that all need to get untangled for this to work. but if i had better pawn fork vision i would have been able to visualize this:

if bishop takes, queen takes, and the rook has to run away.

both my bishops are mobile and can threaten the rook around the board until it has to go back home, winning me tempo in the meantime to take the e4 pawn and develop my other bishop. it comes out to an even trade on points, but im still up a pawn, im ahead on development, and my bishop pair is more powerful than whites knight pair on the board thats left.

if white gets into this fork, the best move saves the rook from the fork in a surprising way. this alternate line is included in the analysis below, but black still comes out ahead in the exchange.

so thats why the rook wont take my knight if i move that bishop to b4 (or out of the way to e7). white instead would have avoided the battle and taken just the pawn, which is the best outcome i could hope for in this position against a savvy opponent.

i needed to see ahead to a fork that would never happen, to know it would never happen. and thats hard to practice.

of course, we both only got to this position in the first place by making mistakes, so anything is possible.

heres the annotated position: