A segment-by-segment accounting of how three utilities should be spent along Banana on Inferno — buy time at the slab, save the molotov for the barrel pit, and lend the flash to the sandbag trade instead of buying yourself a safe zone.
Everyone who plays Inferno shares one illusion: that Banana is something you rush. In reality, it is something you pay for in installments. The slab, the barrel stack, the sandbags, the pit under second floor — each segment has its own timer and its own utility budget. The smoke you spend at the slab does nothing about the angle from second floor. The molotov you save at the barrels gets charged back double at the site. This piece is not about how to play Banana. It is a ledger: three utilities, distributed across four segments.
Segment one: slab to barrels — utility buys time, not space
Most teams throw their first smoke at the mid-Banana line off the slab, aiming to push the CT rifleman back from the sandbags. But the real value of that smoke is not the block — it is forcing the CT to reposition. If he sits behind the smoke and waits it out, you get roughly a 15-second window of undisturbed advance. If he rotates away or falls back early, you get information: B is one man lighter.
The problem starts with the second utility. Plenty of teams add a molotov in front of the barrels, reasoning that they are "clearing the cubby." But the cubby at the barrel stack normally holds one player, and a CT flushed from it will naturally settle behind the sandbags — a more stable second angle. Spend that fire too early and you have effectively helped the opponent complete a free formation collapse.
My recommendation: spend only one smoke on the slab segment, and hold the molotov until the barrel stack is actually occupied. The trigger is not "I think someone is there" but "I heard the tail of a footstep die out on the inner side of the barrels." Sound reading is cheaper than utility here.
Segment two: sandbags and under second floor — the third utility should be lent, not bought
Sandbags are the most expensive stretch of Banana. It is exposed to three angles at once: the front of the sandbags, the window under second floor, and the trade position from the site. Hard-throwing a smoke to block the window usually covers exactly one angle, and you still have to trade bodies for the other two.
The better play is to lend the third utility to a teammate rather than buy yourself a safe zone. Concretely: the first man throws the flash in front of the sandbags, the second man uses the residual flash to look straight at the second-floor window, the third man covers the sandbag front. Total utility spend is unchanged, but every utility now has a named receiver. Utility does not eliminate risk; it distributes risk to the people who can handle it.
The utility line into the pit under second floor is the clearest example. If the smoke onto second floor is thrown by someone outside the site, its landing time desyncs from the entry timing. If the first man throws it himself, he is exposed for half a second before it lands. That half second is the CT's window. So the second-floor smoke should come from the second man, while the first man only holds the sandbag side along the smoke edge. What this line trains is not the throw — it is who holds throwing rights, and at what moment.
Closing: settle Banana by segment, not by round
Most reviews treat Banana as a single unit, and the conclusion is always "we didn't have enough utility." But if you keep the books by segment, three utilities are actually enough: one smoke to buy slab time, one molotov to clear the cubby once the barrels are occupied, one flash lent to the sandbag trade. What is always short is not the count — it is the absence of a clean handoff of responsibility between segments.
Next time you play Banana, try stating it before the round starts: who throws the first, who takes the second position, who owns the decision on the third behind the sandbags. Once the ledger is clear, Banana stops being a corridor you rush on luck and becomes a lane you can settle in installments.