A projection is only as trustworthy as the data behind it, and a player is only as valuable as the games he actually plays. These rules deal with the two ways that trust breaks down: a player whose recent availability is shaky, and a player the projection sources cannot see clearly at all. They are the guardrails that keep a thin or missing signal from masquerading as a confident one.
Fires when a player appeared in fewer than 12 games last season.
Availability is a skill, and recent injury is one of the better predictors of future injury. A player who missed a meaningful chunk of last season — fewer than 12 games — carries elevated risk into the next one, whether from a recurring issue or a body that has shown it breaks down. AutoTiers applies a 12% penalty to price that in. It pairs with the prior-year games discount, which separately keeps those missed games from dragging the player's raw production line down unfairly — the discount adjusts the projection for missing data, while this rule adjusts for the risk the injury implies. And note the counterweight two seasons out: the Year After the Year After boost rewards players who have since returned to full health.
Fires when a player appeared in fewer than 8 games last season.
This flag marks the most serious availability concerns without changing any score — it stacks on top of the injury-history penalty as a visible warning rather than a second haircut. A player who managed fewer than eight games is in a different category of risk, and the flag makes sure that shows up on your board even if his talent alone would place him high. It is there to make you pause and check the situation, not to make the call for you.
Fires when a player has no current-season projection from any source.
This is the bluntest rule in the set, and it is about confidence, not talent. When none of the projection sources carries a current-season number for a player, AutoTiers is essentially flying blind — the player may be a deep-roster afterthought, a name that failed to match across data sources, or someone whose role is genuinely unknown. Rather than invent a projection or rank the player as if the data were solid, the rule halves the score to push low-confidence names down the board where they belong. If you know something the data does not, you can always override it — but the default is to trust a confident signal over a missing one.