CSV to contest entry
Five clear steps from a raw salary file to a legal, defensible lineup.
Export the DraftKings salaries CSV for the slate you are playing and drop it into the builder. One file contains the entire slate, so names, positions, salaries and game information are read together and become available for filtering.
Every player gets a weekly projection, season average, floor (WCS) and ceiling (BCS), plus DSPP — the dollars spent per projected point. Read the columns together rather than treating one number as a complete recommendation.
Plausibility Rank blends sample size, agreement between the average and salary-implied projection, injury status and whether the salary resembles a real starter. Low-priced backups are intentionally discounted instead of rewarded for unrealistic value.
Fill the fixed slots — QB, two RB, three WR, TE, FLEX and a team defense — by hand, or press Auto-build for a cap-compliant starting point. You can remove, replace and compare selections before settling on the final nine.
Salary used and salary remaining stay on screen while you work. Go over $50,000 and the board turns red with an OVER CAP warning until the lineup is corrected. A lineup is not ready until every required slot is filled legally.
Auto-build can surface a mathematically efficient roster, but lineup construction still benefits from context. Review late injury news, expected workload, game environment and your preferred contest strategy. Lock players you want to keep, replace choices you disagree with, and confirm the cap before entering elsewhere.
Suppose a running back is priced at $6,800 with a weekly projection of 21.9 points. Divide salary by projection and you get a DSPP of about $310 per projected point, which is strong for that price tier. Flip it around and the same player returns roughly 3.2 points for every $1,000 of cap space. Those two numbers say the same thing in different languages, and neither one tells you whether the workload behind the projection is real.
That is the job of the remaining columns. If his season average is 26 points, the projection carries a variance of roughly minus 17 percent, comfortably inside the normal 12 to 28 percent discount, which suggests the model is simply being conservative rather than flagging a problem. If the variance were minus 52 percent, something else is going on: a tiny sample, a role change, an injury designation, or a price that no longer matches the job.
Finally, compare his WCS and BCS. A floor of 11 and a ceiling of 33 is a tournament profile; a floor of 17 and a ceiling of 25 is a cash-game profile. Same projection, entirely different reasons to roster him.
Prioritise floor and plausibility. You want players whose worst realistic outcome still keeps the lineup alive, which usually means confirmed starters, stable target shares and defenses in favourable matchups rather than boom-or-bust dart throws.
Prioritise ceiling and ownership leverage. A high BCS relative to salary matters more than a comfortable floor, because finishing near the projection rarely wins a field of thousands. Accept a weaker WCS in exchange for genuine upside.
One lineup should be the version you can defend. If you enter many, use the blend sort to vary the preference order between builds so your entries differ in structure rather than in one swapped receiver.
The first is chasing minimum-priced players purely because the value math looks spectacular. A $4,000 backup with a projection borrowed from a small sample will always look efficient on paper; plausibility exists specifically to strip that illusion away. The second is treating the salary cap as a score to maximise — spending all $50,000 is not a goal, and leaving a few hundred dollars unused is fine if the alternative is a worse player.
The third is ignoring the roster rules until the end. Because the FLEX can hold a running back, receiver or tight end, a lineup can look nearly complete and still be impossible to finish under the cap. Watching used and remaining salary as you fill each slot prevents that dead end. The fourth is forgetting to re-upload after news breaks; a lineup built on last night's pricing can be quietly illegal or simply outdated by kickoff.