Hi Run it once, this is my first started topic on the forums in 2 years :p

I recently played a hand on an Asian app, 10/20/40 game with antes.
Utg raised about 160, hero called in straddle with J8s. About 6.5k effective.
Flop KJ8r. Utg cbet about 300, hero checkraised 3x to 880, utg clicked back to 1500. Hero called with bottom 2 vs what looked like a polarised range
Turn 3off. Check / check
(Pot 3.5k, 5k behind) River 2off. Hero??
I decided that since UTG has a lot of very weak hands and we have a lot of busted hands that had to continue facing such a small 3bet on the flop (e.g open ends, some gutshots) as well as some strong stuff (e.g 2pairs/sets) that decided not to 4bet the flop: Hero should choose a large sizing that will win us the pot a high majority of the time - this will increase the ev of our bluffs as well. I thought to use the all-in size would be best as it would take away his option of rebluffing which would be the only way he could win the pot with his flop bluffs now.

However, when I ran the sim through pio (nodelocking Villain's flop 3bet to a very constricted range of AT, AQ, QT, T9, some random spazz hands like TT, AJ, 87) it got me to a river where OOP had a 62% of the EV and yet pio decided that OOP should check 97% of the time. This was the first question I had - isn't the player with higher EV supposed to be betting, especially when OOP has hands like T9 and QT that need to bluff to win?

1. Why is the EV of betting 4000 chips higher than checking, and yet PIO chooses to check a huge majority of the time?