Key takeaways
- A draft is an allocation problem, not a talent-evaluation problem: the best pick is the one that improves your finished roster most, not the one with the highest number next to it.
- Value is measured against the player who would otherwise fill the slot, so an average player at an empty position often beats a very good player at a position you already solved.
- Tiers carry more information than rankings, because they tell you when waiting is safe and when it is not.
- Your draft slot changes the gap between your picks, and the gap — not the slot number — is what dictates how much risk you can carry.
- No drafting approach wins every time; the pool, the format and the other managers all move, so judge a method across many drafts rather than one.
Pick thirty-four of a twelve-round draft. The top name on your board is a shooting guard, and you have already taken two of those. The next name down is worse on every number on screen and plays the one position you have not touched.
Whatever you do in the next four seconds is your draft strategy. Not the plan you wrote before the draft opened — this.
A draft is an allocation problem wearing a scouting costume
It looks like a talent evaluation exercise. It is a resource allocation exercise. You have a fixed number of picks, you do not control what is available when your turn arrives, and every selection permanently removes an option from the board for everyone including you.
That reframing does real work, because it changes what "good pick" means. A good pick is not the best player. It is the selection that improves your finished roster the most, accounting for everything you can still reasonably get later.
Almost every recognisable draft strategy is a different answer to one question: how do you decide what a pick is worth to you specifically, rather than what a player is worth in the abstract?
Replacement level: the idea the rest sits on
The number beside a player tells you how good they are. It does not tell you what drafting them does for your team. For that you need the player they replace.
Say you are choosing between two candidates, using ratings out of 100:
- An 88-rated guard, at a position where you already start an 85. Net improvement: 3.
- A 79-rated centre, at a position currently filled by a 61-rated bench body. Net improvement: 18.
The lower-rated player is worth six times more to you. Nothing about that arithmetic is subtle, and people still take the 88, because the 88 is the bigger number and the bigger number feels like the better decision.
Two honest caveats. First, ratings rarely convert to results linearly — the distance from 85 to 88 is not necessarily the same amount of "team quality" as 61 to 79, in a real sport or in a simulation. Second, the fallback is not always as bad as you think; deep positions have decent fallbacks, which is exactly why they are deep. Replacement-level thinking is a direction, not a formula.
But the direction is right, and it dissolves the phoney argument between "best player available" and "draft for need". Those are not two philosophies. They are two situations. Best available is correct when the gap at the top of the board is enormous. Need is correct when the hole you are ignoring cannot be patched later. Both are the same calculation run on different boards.
Tiers carry information that rankings throw away
A ranked list is a lie of uniform spacing. It presents the gap between the fourth and fifth names as equivalent to the gap between the twelfth and thirteenth, and it almost never is.
Tiers fix this. Group the board into bands where you would be genuinely content with any name in the band. Then the strategy falls out of the structure:
- Your tier still has five names in it. You can spend this pick elsewhere. You will very likely still get one of the five.
- Your tier is down to one name. This is the pick. Waiting costs you the whole drop to the next tier.
- Two positions both have one name left. Take the one whose next tier is further down.
Building tiers without overthinking it
Sort each position by rating and read down the column looking for the cliffs — the places where the numbers stop declining gently and fall. Draw the line at the cliff. You do not need precision here. Five or six bands per position is plenty, and the exercise takes a couple of minutes.
The point of tiers is not accuracy. It is that they force you to notice the shape of the drop-off before you are under time pressure, when noticing is easy.
Reading a run
A run is when several managers take the same position in consecutive picks. It happens because everyone is reading the same board and the same cliff, and it is the single most common way a draft plan gets destroyed.
The correct response is unglamorous: check whether your tier at that position survives the run. If four names remain and three picks separate you from your turn, you are fine — hold. If the run will empty the tier before you pick again, you either jump into the run now or you accept the next tier and plan around it. What you should not do is join a run out of anxiety when your tier is deep, because that is how you end up with three players at a position you had already solved.
Your slot changes the plan, not the outcome
In a snake draft the order reverses each round, so your position determines the gap between your picks rather than their average quality.
Take ten teams. Round one runs picks 1 to 10 in order; round two runs 11 to 20 in reverse. The manager at slot 1 selects with picks 1 and 20. The manager at slot 10 selects with picks 10 and 11.
That is a 19-pick gap versus a 1-pick gap, and it has direct consequences:
- Early slot. You get the single best player in the draft and then watch nineteen names disappear. This rewards taking the one player you rate far above the field, because that surplus is the compensation for the wait. It punishes plans that depend on a specific second player being there.
- Late slot. You never get the top name, but you take two players back to back with nobody able to interrupt. This rewards taking a matched pair from the same tier, and it lets you cover two positions in one motion.
- Middle slots. Even gaps, no extremes, and the most sensitivity to runs, because there are always picks on both sides of you.
None of these is an advantage. They are different constraints, and the mistake is running an early-slot plan from a late slot.
Four strategy families, and what each one is actually buying
Named strategies proliferate, but nearly all of them are one of these four trades.
Balanced accumulation. Take the largest available improvement at every pick, spread across positions. Buys: reliability and a high floor. Costs: you rarely end up with the single best unit in any one area.
Concentration. Spend early picks on a small number of very high-rated players and accept weak slots elsewhere. Buys: a genuine ceiling and units nobody can match. Costs: your weakest slot is very weak, and formats that weight depth punish this hard.
Scarcity-first. Take the thin positions early regardless of raw rating, and fill the deep ones from the leftovers. Buys: you never finish with an uncoverable hole. Costs: you pass on better players and can end up merely adequate everywhere.
Deliberate waiting. Identify a position where the tenth-best option is nearly as good as the third and refuse to spend an early pick there. Buys: extra early picks for scarce positions. Costs: it fails badly if you misread the depth, and you cannot know you misread it until it is too late to fix.
Every one of these is right sometimes. That is not a cop-out — it is the actual structure of the problem. The board is different every draft, the format changes what the roster is scored on, and the other managers move too.
What strategy cannot do
It cannot guarantee a result. A draft has an irreducible amount of variance in it: which players are generated or available, where the runs happen, what the managers around you do. A well-reasoned draft can produce a mediocre roster and a careless one can fall into a good board.
That is not an argument for drafting carelessly. It is an argument for judging yourself on the decision rather than the outcome, and for evaluating any approach across many drafts instead of the one you remember.
In a simulation, there is a second limit worth stating plainly. When a game returns a record for the roster you drafted, that number is the output of a model scoring your team-building. It is not a prediction about anything in the real world, and it should never be read as one.
Formats decide which strategy is even legal
The reason drafting stays interesting after fifty runs is that changing the rules changes the correct answer with the same players on the board.
Pro Basketball GM Franchise includes eight draft formats — Unlimited Draft, Quick Draft, Blind Draft, Franchise Draft, Salary Cap Draft, VS Computer, Battle Mode and a Daily Challenge — and they pull in different directions:
- Unlimited Draft gives you room to compare, which is where you learn what a board looks like.
- Quick Draft removes deliberation time, so whatever tiers you built beforehand are the only structure you get.
- Blind Draft removes information, which makes scarcity-first reasoning more valuable than fine-grained rating comparisons.
- Salary Cap Draft attaches a price to every pick, converting the ranking problem into an allocation problem where a merely good player at a low cost can beat a great one at a high cost.
- VS Computer and Battle Mode add an opponent, which introduces a resource you were not tracking: what they can still take.
The catalogue description of the game is direct about what the season rewards — balance and depth carry more weight than any single marquee pick. That is a design decision, and it is exactly the kind of thing you can only learn by drafting differently and comparing.
20-0 Football Draft & GM Mode makes the same point from the other end with its Limited Draft, which removes refreshes entirely. Take away the ability to keep looking and every pick has to stand on its own, which is the purest test of whether your tiers were real or decorative.
A practice loop that actually improves you
- Tier the board before your first pick, not during. Under time pressure you will default to the ranking.
- Name what you gave up on every selection. If you cannot say which player you passed on and why, you did not decide anything.
- Track your worst slot, not your best. The roster is judged on the whole, and the position you keep not fixing is where the loss lives.
- Run the same plan several times. One draft tells you about that board. Five tell you about the plan.
- Write one sentence afterwards about why the result came out where it did. This is the only step that converts drafting into learning, and it is the one everybody skips.
Where to read next
The two big format decisions each deserve their own treatment: snake versus auction drafting covers what changes when picks are bought instead of taken in order, and salary cap basics covers budgets as a constraint. If the preparation side is what you want to improve, building a draft board is the practical version of the tier work above, and the basketball draft simulator guide applies all of it to one game end to end.
The rest of the studio's drafting and management titles sit in the sports GM category.
Good to know
