Draft Science · Snake drafts

Should you punt from pick one? No. Committing early costs titles.

The short answer

Deciding on a punt before the draft gives away value you don't win back: stay flexible, and let the draft show you which category to give up.

  • Seven of the nine punts we committed to before pick one clearly won fewer titles than staying flexible. Over 100 seasons, the flexible draft won about 34 titles. Punting free throws won about 26, and punting blocks about 22. Field goal percentage cost a little, and steals cost nothing we could measure. (An average team wins about 7. Our computer opponents are easy, so watch the gap, not the total.)
  • The worst punts take a second category down with them. Punt blocks and you lose rebounds too. Punt threes and points and free throws go with it. Steals, turnovers and field goal percentage took the least down with them, and they were the cheapest.
  • Punting blocks was costly in every check we threw at it. Stronger opponents, wrong projections, one centre, no streaming: blocks cost 9 to 12 titles per 100 seasons each time. How much the other punts cost moved with the league, which is one more reason to let the draft decide.

The case for punting

Punting is one of the most popular plans in head-to-head category leagues. You give up one category on purpose, free throws say, and spend every pick on the other eight. You lose free throws every week, but you win the rest so often that you take most weeks 5–4 or 6–3. It sounds like free value, and plenty of good managers swear by it.

The catch nobody measures

A punt costs you that category every single week, playoffs included, against the best teams in the league. A 5–4 win against a weak team is easy. Against a strong one, you start almost every week a category down and need five of the remaining eight from someone who is good at most of them.

So does the trade actually win more championships? Most punt advice is argued from feel and from the one season it worked. We wanted a count.

Our hypothesis: committing early gives up value

Once you've decided to punt free throws, every good free-throw shooter on the board is worth less to you, so you pass on them. And you reach for players who fit the plan, even when a better player is still there. Each of those picks gives up a little value. The question is whether the punt wins all of it back.

The alternative is how Pikola drafts by default: stay flexible. That doesn't mean building a balanced team. Pikola is free to give up as many as two categories, or three in a one-win league, where the week goes to whoever wins more categories. It just doesn't choose them before pick one. It weighs every pick against your roster and what's still on the board, and lets a category go when the draft makes it cheap.

How we tested it

Picture two managers with the same pick in the same 14-team snake league. Before pick one, the first tells the draft “I'm punting free throws” and drafts to that plan all the way. The second stays flexible: no punt chosen in advance, free to punt later if that's where the value is. Same league, same players, same luck. Then we play the season out, many times over.

Both managers are Pikola, drafting on its own. The only difference between them is the punt, so any gap in titles comes from the punt and not from one drafter being better. We ran it for every one of the nine categories, one punt at a time, 112 drafts each, and played every draft out over 40 simulated seasons.

  • 9categories, each punted from pick one
  • 4,480simulated seasons for each punt
  • 7 of 9clearly cost titles; steals broke even

Seven of nine early punts clearly cost titles. Blocks, rebounds and threes cost the most.

Exhibit 1Titles lost per 100 seasons, compared with no early punt

Zero is the same drafter with no punt chosen in advance. A longer bar is a bigger loss. A hollow bar is so close to zero that it could just be luck. Two league formats: Categories, where every category counts as its own win, and One Win, where the week goes to whoever wins more categories.

CategoriesOne WinHollow: could be luck

Read the bars as titles per 100 seasons. In a standard categories league, the flexible draft won about 34 titles. Punting steals cost none of them. Turnovers cost about two, and so did field goal percentage, small enough that it could be luck. Assists cost about five, and points and free throws six or seven each. Rebounds and threes cost about ten, and blocks about twelve.

The surprise is at the bottom. Punting blocks, the heart of the build that skips big men, cost more than any other punt: about one and a half times what punting free throws, the classic punt, cost. A blocks punt gave up about a third of the titles the flexible team would have won.

So what: a punt chosen before pick one costs you titles, and which category you pick matters a lot. Steals is close to free. Blocks costs the most.

Why: the expensive punts take a second category with them

Every punt did what it promised: the punted category collapsed and a few others went up. The difference was what else went down with it. The chart shows punting blocks, the most expensive of the nine.

Exhibit 2Punting blocks, category by category

How often each category was won in the regular season, categories league: no early punt against blocks punted from pick one.

No early puntPunt blocks

Blocks fell from 72% of weeks to 15%, as planned. But rebounds fell from 60% to 32% as well, most likely because the players who block shots are the players who rebound. Turnovers slipped too, from 51% to 45%. The punt bought back assists, points, threes and free throws, but the flexible team was already winning most of those weeks. Add up all nine and the blocks punt won about half a category fewer in an average week than the flexible team.

The same thing happened across the board. Punting rebounds dragged blocks down with it. Punting threes took points and free throws, and punting free throws took threes. The cheap punts cost little overall: steals, turnovers and field goal percentage each gave up about a tenth of a category a week or less, and blocks, rebounds and threes four or five times that. Rank the nine punts by how many categories they won in an average week and you get almost the same order as the titles chart.

So what: a punt is only as cheap as what it drags down with it. Bigs bring rebounds and blocks together, and shooters bring threes, points and free throws together. Give up one and others go with it.

Steals and field goal percentage were cheap this year, for a reason

The two cheapest punts tell you something about this year's draft. We looked at the players a 14-team league drafts, by this year's projections and Yahoo's draft order.

Field goal percentage was already the flexible team's natural punt. A drafter taking value this year ends up long on shooters: the flexible team won threes in 76% of weeks. And this year's shooters hurt field goal percentage. Of the 30 best three-point shooters a 14-team league drafts, 28 are below average for it. So the flexible team was already losing field goal percentage, winning it in only 42% of weeks. Punting it on purpose made official what the draft was already doing.

Steals come from everywhere. Of the 30 best steals players drafted, 22 are above average for assists, but 15 are also above average for blocks and 15 for rebounds. Steals aren't tied to one kind of player, so you collect them whoever you draft: even punted, steals were still won in 38% of weeks, down from 58%. Blocks are the opposite. Of the 30 best shot-blockers, 25 are above average for rebounds and only 3 for free throws, which is why punting blocks takes rebounds with it.

So what: the cheapest punt is the category your draft is already giving up, and that depends on the year's players and how your league drafts. That's a reason to watch for it as the draft goes, not to plan it in advance.

Punters lose their titles in the playoffs, like any weaker team

Most punts barely hurt a team's chances of making the playoffs: in a categories league, the flexible team got in 98% of the time and the punters 91% to 98%. Most of the titles went after that. For every punt that cost titles, at least three in every four of the lost titles were lost in the playoffs.

That isn't special to punting, though. We also ran a weaker drafter that never punts: it takes the best player by Pikola's plain player values, with no planning for matchups. It won about 6 fewer titles per 100 seasons than the flexible team, and 89% of those went in the playoffs too. Knockout weeks against the best teams magnify any weakness, and a punt is one.

So what: making the playoffs isn't the test of a punt. Winning there is, and a punt makes that harder the same way any weaker roster does.

One-win leagues didn't change the answer

Punters often say the strategy is made for one-win leagues, where you only need five categories of nine to take the week. In our runs the format made little difference. Every punt cost about the same in both, and the order barely moved. The biggest gap was field goal percentage, which cost about two titles more per 100 seasons in one-win leagues, small enough that it could be luck.

So what: don't let a one-win format talk you into a punt.

Punting blocks stayed costly in every check. The rest moved with the league.

A study like this invites fair objections, so we tested the main ones. Each check reran the most interesting punts against the same flexible drafter, with one thing changed: stronger opponents, who all draft by Pikola's own values; seasons played on projections that are wrong by 15%, so the drafter can't simply trust its numbers; one centre instead of two; and no streaming.

Exhibit 3Titles lost per 100 seasons, compared with no early punt, five ways

Categories leagues. Each row changes one thing from the study as run. A dot marks a check we didn't run for that punt.

CheckNo early punt, titlesBlocksReboundsThreesFree throw %Steals
The study as runOpponents draft by ADP, need and Pikola's values33.6−11.7−9.6−10.1−7.2−0.1 (could be luck)
Stronger opponentsEvery other team drafts by Pikola's own values34.7−9.9−11.9−5.9−6.2−3.1
Wrong projectionsSeasons played on projections wrong by 15%32.5−10.3−13.7−9.1−6.1−0.2 (could be luck)
One centrePG, SG, G, SF, PF, F, C and three Util32.2−9.1−6.3−10.9··
No streamingNobody streams, so a punter can't stream its punted category34.7−11.2−7.3−7.7−3.0 (could be luck)·

Read it like this: punting blocks cost between 9 and 12 titles per 100 seasons whatever we changed. Rebounds and threes stayed expensive too, but how expensive moved. With one centre on the roster, punting rebounds cost about 6. Against stronger opponents, punting threes cost about 6 and punting steals about 3.

Free throws moved the most. In our seasons every team streams the best player available overall, not the best one for its punt, so a free-throw punter kept streaming good free-throw shooters. With no streaming at all, punting free throws cost about 3, small enough that it could be luck. A real punter streams for the build, so the main run probably overstates what punting free throws costs.

So what: punting blocks is a bad plan everywhere we looked. For the other punts the price depends on your league, which you only find out as the draft goes. One more reason not to lock one in.

What to do on draft day

  • Don't commit to a punt before pick one. Take the best player for your team and let the draft show you where you're weak. That's what Pikola works out for you, pick by pick.
  • Let a punt find you rather than planning one. If your roster drifts away from a category by the middle rounds, leaning into that is a different decision. When we let an earlier version of Pikola do that in a separate test, it neither helped nor hurt.
  • If you are set on punting from the start, pick one that takes nothing with it. Steals, turnovers and field goal percentage cost the least, at up to about four titles per 100 seasons.
  • Never plan to punt blocks, and think hard before rebounds or threes. Each takes another category down with it. Blocks was costly in every check; rebounds and threes were costly in most.
The full numbers

Regular-season results first: every punt also won fewer weeks, and these differences are clearer than the title ones because a season has many weeks and only one title. All but one clear the luck bar; steals in a one-win league doesn't. The punts rank in nearly the same order by weeks as by titles.

Exhibit 4Weekly win rate, compared with no early punt

Percentage points, regular season. The flexible draft won 60.3% of category results in a categories league and 73.1% of weeks in a one-win league.

CategoriesOne WinHollow: could be luck

Every row, both formats. “vs no early punt” is the punt minus the flexible draft on the same 112 drafts, in percentage points (titles or weeks per 100), with its standard error. Playoffs % is the share of seasons that reached the playoffs, and title % once in the playoffs the share of those that ended in a title. Categories won a week is how many of the nine an average regular-season week brought.

FormatBuildWeekly win %vs no early puntTitle %vs no early puntPlayoffs %Title % once in the playoffsCategories won a week
CategoriesNo early punt60.333.698.434.25.44
CategoriesPunt STL59.7−0.59 ± 0.1333.5−0.11 ± 1.1398.034.25.37
CategoriesPunt FG%59.3−0.99 ± 0.1532.0−1.67 ± 0.9797.332.95.34
CategoriesPunt TO59.3−0.97 ± 0.1431.4−2.28 ± 0.9697.532.25.33
CategoriesPunt AST58.0−2.28 ± 0.1828.5−5.13 ± 1.1696.929.45.22
CategoriesPunt PTS57.8−2.50 ± 0.1827.1−6.50 ± 1.1695.328.55.21
CategoriesPunt FT%57.4−2.86 ± 0.2026.4−7.23 ± 1.1594.727.95.17
CategoriesPunt REB56.2−4.13 ± 0.1924.0−9.62 ± 1.1192.526.05.05
CategoriesPunt 3PM55.2−5.07 ± 0.1723.5−10.13 ± 1.0492.225.54.97
CategoriesPunt BLK55.6−4.71 ± 0.1921.9−11.70 ± 1.1291.324.05.00
One WinNo early punt73.133.897.734.65.43
One WinPunt STL72.6−0.50 ± 0.2732.6−1.21 ± 1.0597.233.55.38
One WinPunt FG%71.4−1.73 ± 0.3430.1−3.64 ± 1.1395.931.45.32
One WinPunt TO71.2−1.89 ± 0.3430.7−3.10 ± 1.0396.031.95.33
One WinPunt AST70.3−2.84 ± 0.4028.6−5.20 ± 1.0995.729.95.22
One WinPunt PTS67.5−5.66 ± 0.4025.4−8.35 ± 1.0792.227.65.17
One WinPunt FT%68.7−4.46 ± 0.4026.0−7.77 ± 1.0493.827.75.17
One WinPunt REB65.7−7.40 ± 0.3922.7−11.05 ± 1.0790.725.05.03
One WinPunt 3PM65.4−7.75 ± 0.4022.7−11.12 ± 0.9490.025.24.94
One WinPunt BLK64.1−9.03 ± 0.4620.7−13.04 ± 1.1088.223.54.96

How it was run: Pikola's engine as it was on 9 October 2026, on the projections of 8 October. Each of the 112 drafts per punt was paired with a draft that chose no punt in advance, and each draft was played out over 40 seasons; the standard error is taken across the 112 pairs, since the seasons from one draft share a roster. The other 13 teams drafted by average draft position with a little noise, by draft position with positional needs, and by Pikola's own player values. Rosters: PG, SG, G, SF, PF, F, two C, two Util and three bench, with two spots streamed each week. Playoffs: the top eight of 14, three one-week rounds, each won by the majority of categories, a tie going to the higher seed. A named punt sets that category's weight to zero from the first pick and changes nothing else. Both sides concede as the app's Setup allows: Pikola may give up as many as two categories as the draft goes, three in one-win leagues, and a named punt uses one of them. An earlier run the same day let a named punt come on top of two (so the punter could give up three to the flexible team's two); it ranked the punts the same way, with the big-man punts costing up to about three titles more. The category shares come from two halves of the run, each reported in whole percents, so they are good to about a point. The figures on this year's players are the first 182 by Yahoo's draft order on the same projections. The checks in Exhibit 3 used the same engine, projections and concessions, in categories leagues: stronger opponents (every other team drafting by Pikola's own values with a little noise) and wrong projections (every player's per-game line off by a log-normal draw of 15%, one draw per set of opponents, 16 sets on 7 draft slots) on 112 drafts each; one centre (PG, SG, G, SF, PF, F, C and three Util) and no streaming on 56 each; the weaker drafter (Pikola's neutral player values, no matchup planning) on the main run's 112. Its standard errors judged by set of opponents instead of by draft, the cautious way since the seasons and the field share a seed, are in the methods record. This replaces the first version of this study (4 October 2026), which tested six punts, free throws plus turnovers among them, on the engine of 2 October, when naming a threes punt also halved the weight on points.

Staying flexible, with no punt chosen in advance, is how Pikola drafts by default. Pikola is a Chrome extension that advises live in Yahoo fantasy basketball drafts. Try it free in a mock draft · How Pikola is tested · More Draft Science