Key takeaways
- A draft simulator is a constrained optimisation puzzle: a fixed number of picks, a pool you do not control, and a scoring model you can only infer from results.
- The highest-rated player on the board is frequently the wrong pick, because value is measured against the player you would otherwise start, not against the field.
- Positional scarcity decides which picks you can safely delay — take the scarce slot early and the deep one late.
- A simulated season record is the output of a model, not a prediction about anything real; treat it as a score for your roster construction.
- Draft formats are the difficulty setting of the genre, because changing the constraint changes what a good pick even means.
Pick six of a salary cap draft, and the best player left on the board plays a position you have already filled twice. The next name down is worse on every visible number and covers a hole you have not touched. Take the higher rating and you have spent a pick on a backup. Take the lower one and you have, on paper, made your team worse.
That four-second decision is the game. The 82 games that follow are just the scoreboard reading back what you already decided.
What the genre is actually about
Strip the jerseys off and a draft simulator is a constrained optimisation puzzle. You get a fixed number of picks. You do not control the pool. There is a scoring function that converts your finished roster into a result, and you can only ever infer its shape from the results it gives you.
That last part is what makes it worth playing more than once. If the model were visible you would solve it in an afternoon. Because it is not, the loop is: form a theory, draft to it, read the record, revise the theory. That is a genuinely different activity from playing a sports game, and it is why people who have no interest in tapping through a match will happily spend twenty minutes on a draft board.
It is also why the genre lives comfortably on a phone. A draft is a sequence of discrete decisions with a clean stopping point. There is nothing to twitch, nothing to time, nothing you lose by being interrupted.
Three ideas that separate good drafters from lucky ones
Most bad drafts are not bad luck. They are one of these three mistakes, repeated.
Value is relative to your own bench, not to the field
The number next to a player's name tells you how good they are. It does not tell you what drafting them is worth to you.
If you already have a strong player at a position, the second one you draft there does not add their full rating to your team. They add the difference between themselves and the player they displace, and if they displace nobody, they add close to nothing. Meanwhile the position you left empty is not costing you a small amount — it is costing you the entire gap between a real starter and whatever the roster puts there instead.
This is the single most useful reframe in the whole genre. Stop asking "who is the best player left" and start asking "which pick moves my worst slot the furthest".
Positional scarcity decides your order, not your preferences
Some positions are deep in a given pool and some are thin. If you can reliably find a competent option at a position in the late rounds, you should not spend an early pick there, even if the early option is genuinely better. If a position dries up quickly, you take it before it does — and you accept a small loss in raw quality to avoid a large loss later.
Scarcity is a property of the board, not of the sport, and it changes from run to run. Reading it is a skill. The tell is usually the shape of the drop-off: a position where the third-best and the tenth-best options look similar is one you can wait on. A position where quality falls off a cliff after two names is one you cannot.
Every pick is an opportunity cost
There is no such thing as a free pick. Taking a player is simultaneously deciding not to take every other player, and in a format with a budget it is also deciding not to spend that money elsewhere.
This sounds obvious and it is routinely ignored, because rating numbers are visible and opportunity costs are not. The discipline is to make yourself name the alternative before you confirm a selection. If you cannot say what you gave up, you did not make a decision — you clicked the top of a list.
Formats are the difficulty setting
Basketball Draft GM Franchise ships eight draft formats: Unlimited Draft, Quick Draft, Blind Draft, Franchise Draft, Salary Cap Draft, VS Computer, Battle Mode and a Daily Challenge. The important thing about that list is not its length. It is that a roster which dominates one format can fall apart in the next, because each one changes the constraint you are optimising under.
Consider what a budget does to drafting. Without one, the board is a ranking problem — take the best available thing that helps you. Add a cap and it becomes an allocation problem, where a merely good player at a low price can be a better pick than a great player at a high one. The same board, the same players, a completely different correct answer.
Competitive formats change it again. Drafting against an opponent introduces a resource you were not tracking before: what they can still take. Now a pick can be defensive — you take a player earlier than their value justifies because losing them to the other side costs you twice.
If you are new to the genre, the honest advice is to spend your first several runs in an unconstrained format learning what the board looks like, and only then add constraints. Learning the pool and learning the rules at the same time means you cannot tell which one beat you.
Reading the season result without over-reading it
This part matters, so it is worth being blunt about it.
When the season simulation returns a record, that record is the output of a model. It is a function of the roster you built, evaluated against whatever the game weighs — quality, positional coverage, balance, and the depth you did or did not leave yourself. It is a score for your drafting. It is not a prediction, it does not describe anything that happened in the real world, and it has no bearing on any actual game, team or player anywhere.
Two practical consequences follow.
First, judge yourself on the process, not on one result. A single season is one sample of a model you cannot see. If you want to know whether a drafting approach is better, run it several times and compare the distribution, not the best run you got out of it.
Second, use the record diagnostically. A disappointing record after a draft you were proud of is information, and it is usually pointing at exactly one of two things: an uncovered position you talked yourself out of caring about, or a roster stacked in one area at the expense of everything else. The catalogue description of the game is explicit that balance and depth carry more weight than a single marquee pick, and the simulation is where that shows up.
GM Mode is where a good draft gets wasted
Drafting well and then abandoning the roster is the most common failure in franchise play. Draft day sets your ceiling; the front office decides how much of it you reach.
The GM side of the game covers contracts, trades and rotations. Each of those is a different kind of decision from drafting:
- Rotations are a distribution problem. You have a fixed amount of playing time and a roster that is not uniformly good. Concentrating minutes on your best players raises your average but thins you out; spreading them protects you against a weak spot but dilutes your strengths.
- Trades are the only tool that can fix a structural mistake. If the draft left you long at one position and short at another, that is precisely the imbalance a trade exists to correct — you are converting surplus you cannot use into coverage you need.
- Contracts are drafting with a time dimension. The question is no longer only what a player is worth, but what they are worth for how long, and what committing to them stops you doing later.
Franchise play across multiple seasons compounds all of this. A team built to peak immediately and a team built to stay strong are different teams, and choosing between them before the draft rather than during it is most of the skill.
A routine that reliably improves results
If you want a concrete method rather than principles:
- Scan the whole board before the first pick. Find where each position falls off. That map, not your favourite player, decides your order.
- Name your alternative out loud on every pick. If you cannot, you are not choosing.
- Fill your worst slot, not your best one. The team is only as good as the position you neglected.
- Leave one pick's worth of flexibility if the format allows it. Boards develop; a plan with no slack cannot respond.
- After the season, write down one sentence about why the record came out where it did. This is the whole learning loop, and skipping it is why people plateau.
Which slot counts as your worst is a question about coverage rather than ratings, and it is the reason five high-rated players at overlapping positions lose to four and a specialist. How to build a balanced basketball roster works through that in full.
Where to go from here
The drafting principles above port cleanly across sports, but the shape of the problem does not. Football has the deepest roster and the widest spread of positional value, which is covered in the football draft simulator guide. Baseball is the most statistically explicit, and the baseball draft guide goes into what those numbers mean before you draft on them. Hockey is the tightest puzzle of all — six players and a coach, where a single weak slot has nowhere to hide, as the hockey draft guide explains.
If the front office is not what appeals and you would rather live inside one player's career than run the organisation, that is a genuinely different game; the basketball career sim guide covers it. The rest of the studio's management and drafting titles sit in the sports GM category.
Good to know
