TL;DR

dpei's biggest pet peeve in low-to-mid elo TFT: players looking at a comp guide, seeing an endgame board, and trying to play that board right now even though half the units are stages away. The better approach is to play your strongest current board and let the wins and gold compound into the endgame naturally. Every suboptimal decision made "for the endgame" costs HP and gold in the present — and the endgame never arrives when you're dead on stage 4.

Key Takeaways

• Play your strongest board for the current stage, not the one you hope to hit in 10 rounds. The endgame comp is the destination, not the route.
• Intermediate carries matter. The mid-game unit that wins you stages 3-4 is what funds the endgame through saved HP and streak gold. Selling it early because "it's not in my final comp" usually costs the game.
• "Closest board" test for slamming items: when you hit a component, ask which unit on your current board will actually hold it soon, not two stages away.
• Resist unlock-chasing when you're already strong. If you're win-streaking on Board A, don't break the streak to unlock Trait B. You can unlock it later on neutrals without paying the streak-break cost.
• Every decision compounds through streak gold and HP. dpei quantifies it in the featured game: a few "endgame-first" mistakes would have cost him ~24 gold and ~20 HP by stage 5 — the difference between winning and dying.
• Suboptimal items still streak. Sometimes the "perfect" item isn't available on a component. A slightly worse item that wins you the round is better than a perfect item that arrives after you've dropped to 10 HP.
• The principle transfers across sets. Specific units and traits change patch to patch; "strongest current board" doesn't.

Timestamps

• 0:00 Intro — new patch, fast-9 meta framing
• 5:13 First "strongest board" call in the game
• 7:37 Strongest-board check mid-stage-3
• 16:14 Resisting the temptation to unlock a trait when current board is already winning
• 29:26 "Play for your third or whatever" — play for the stage you're in
• 40:27 Retrospective: the strongest-board path was stronger than the "higher cap" path
• 40:43 "The most important thing is not to think too far ahead"
• 45:20 The hypothetical: what -24 gold and -20 HP would have cost
• 47:09 Closing — start thinking about strongest board NOW

The mistake: treating the endgame board as the current board

The most common pattern dpei sees in low and mid elo: a player opens a comp guide, sees the recommended stage-5 composition, and tries to play that composition at stage 3. They skip buying current-stage units because "those aren't in my comp." They save items for a 5-cost that's two stages away. They refuse to unlock intermediate traits because the guide doesn't list them.

The problem is that a comp guide is a destination, not a route. The endgame board assumes you arrived with enough gold, enough HP, and enough tempo to play it. All three of those come from winning rounds in stages 3 and 4 — which you can't do if you're playing a half-built version of the endgame instead of the actual strongest board available right now.

"Strongest current board" is a constant question

At every decision point — buying a unit, leveling, spending items — the question dpei keeps returning to is "what's my strongest board for this stage?" Not "what does my comp look like at stage 5?" but "what lineup of units I currently own or can easily hit wins me this next round?"

Sometimes that means:

• Keeping a unit you'll eventually sell because it wins you rounds now
• Playing a trait combination that isn't in your "final" comp
• Using items on an intermediate carry rather than saving them for the endgame unit

The criteria is always: does this lineup win me the next round? If yes, play it. The next decision can re-evaluate.

The "closest board in the future" rule for items

Item slamming is where endgame-fixation hurts the most. A classic mistake: you hit a component that could complete a BiS item for your stage-5 carry, so you hold it — and sit on a three-slot item decision for the next four rounds while losing fights.

dpei's rule: when you slam an item, think about the closest board in the future, not the eventual endgame. Which unit on your board will actually hold this item in the next 3-5 rounds? That unit gets the item. If the answer is "my endgame carry, in two stages," the item probably shouldn't be slammed yet — and certainly not on the wrong unit to "keep slots open."

In the featured game, dpei deliberately avoided completing the BiS endgame item because the endgame carry was "two stages away." Instead he built Shojin on his current carry — the unit that would actually hold an item in the next 10 rounds. That's a "suboptimal" item from the endgame perspective, but it won the present-stage rounds that got him to the endgame at all.

Resisting the unlock trap

Every comp has trait unlocks that look attractive: a 4th piece to hit the next tier, a specific unit to enable a new trait entirely. The trap is breaking your current strongest board to chase an unlock that doesn't actually make the next fight better.

"I'm on a win streak, it doesn't make that much sense to unlock. I'll unlock later on neutrals." — dpei

The key insight: trait unlocks usually don't have to happen right now. If your current board is winning rounds and streaking, you can delay the unlock to a neutral round where there's no opportunity cost. Breaking a streak to unlock early costs streak gold, costs HP (from the fight you suddenly lose), and often doesn't even make the next fight better.

The HP and gold accounting

Every decision compounds through two currencies that are easy to ignore: HP preservation and streak gold.

In the featured game, dpei walks through a counterfactual at the end: what if he'd taken a couple of "higher cap" augments and items earlier, playing for the endgame? His estimate: -24 gold and -20 HP by mid stage-5. That's:

• 24 fewer gold for his final rolls (potentially 24 fewer shop refreshes to hit his 5-costs)
• 20 fewer HP as buffer — which in the game would have put him at 1 life for the decisive fight

Those numbers aren't abstract. They're the difference between 1st and 5th. And they compound from what look like small, forgivable mistakes early: a suboptimal item, a broken streak, a trait unlock chased too early.

"Every game is on a razor's edge whether or not you see it." — dpei

Suboptimal items still win when they keep your streak

A subtle but important point: sometimes the "correct" BiS item isn't buildable from the components you have. Players often sit on components waiting for the right conversions.

dpei's answer: slam a slightly suboptimal item if it wins you the round. Breaking your streak to optimize one item is almost always worse than holding the streak with an imperfect item. In the featured game, he called out making Ionic Spark and Sunfire instead of the "better" Morellonomicon and Void Staff — because the better versions weren't available and the suboptimal items kept the streak alive. The streak gold and HP saved outweighed the item-quality gap.

It's skill, not luck

The closing point dpei hammers is that games like this one look like highrolls from the outside. A Leblanc unlocked at the right moment, a streak that compounded to 40 gold at stage 4-5, an endgame that hit cleanly — all of it looks like the game "gave" him the win.

It didn't. The decisions leading up to the endgame are the skill. Strongest-board play, intermediate carries that win rounds, resisting the unlock trap, slamming items for the present — those compound into the HP and gold that make the endgame possible. The same board composition with worse piloting ends up 5th or 6th.

"It's not the comp itself that matters. It's the piloting of the comp." — dpei

The transferable principle

Specific comps and carries change every patch. The question "what's my strongest current board?" does not. If you're looking at a comp guide and feeling stuck because "I don't have my endgame units yet," that's the moment to close the guide and ask the strongest-board question instead. Play what wins now. The endgame arrives when you've kept enough HP and built enough gold for it to arrive.

Related

• Strongest board strategy — the foundational concept this video builds on
• Win streaking — the streak gold that compounds from strongest-board play
• Rolling strategy — when to actually roll for the endgame units vs stay on your current board
• Item strategy — applying the "closest board" rule to item slamming
• Stage by stage — what "current stage" actually means for decision priorities
• Transitions — how to actually move from an intermediate board to the endgame without breaking a streak
• TFT Economy Guide — the compounding gold/HP currencies that strongest-board play preserves