Why I Draft Ten Forwards

I pulled the Drafters board on 2 September 2026 and counted what a finished roster looks like. Across the whole ranked pool (224 players) the average team comes out here:

I went into this expecting to find the room wasting picks. It isn't. That split is close to exactly right, and the reasoning that gets you there is worth having even though the destination turns out to be where you already were.

Count slots, not players

Eighteen players, eight starting spots. Ten of your roster does nothing in any given week, and which ten changes constantly.

The useful number is how many of a position you own per slot you can actually start. Do that division:

Two, two and a half, three. The room carries progressively deeper coverage the narrower the position gets. That depth is the correct response to variance, not waste, and most drafters build it by feel without ever writing it down.

Why the narrow positions need more coverage

Best ball never punishes you for starting the wrong player: it takes your best scores automatically. So the only thing that can hurt you is not having a good score available to take.

With five forward slots, that almost never happens. You start five of ten, which means the five worst weeks on your roster simply do not count.

With one goalie slot, it happens constantly. Your starter's team plays two games some weeks and four in others. He might get pulled, or rest, or sit behind a hot backup for a stretch. If you own one goalie, a light week is close to no score at all, in a slot you are obliged to fill. Owning three means the odds of all three having a quiet week at once are small.

A slot you can only fill once needs redundancy in a way a slot you fill five times does not. Three goalies for one slot looks extravagant until you notice that the third goalie isn't competing with your starter; he is insurance against the specific weeks when your starter barely plays.

The fifth defenceman is the same logic at lower intensity. Two slots, a position where week-to-week scoring is genuinely erratic: a defenceman's good week is usually a power-play goal and two assists, and those arrive in clusters. Five for two gives you enough draws to catch the clusters.

Where the room agrees with itself

Nobody takes a goalie early. In the first thirty-six picks the board takes twenty-five forwards, ten defencemen and zero goalies. The first goalie off the board goes at 40.3. The room has collectively decided that no goalie is worth a premium pick, and then the same room ends up carrying three of them. That looks contradictory and isn't. Goalies are cheap and necessary, which is exactly the profile of a position you buy late and buy repeatedly.

The third goalie is nearly free. Eighty of the 224 ranked players sit between picks 145 and 180, which is where the board stops meaningfully distinguishing between players. The third goalie comes out of that pile, and the forward he costs you is one the board could not tell apart from forty others.

So why ten forwards and not eleven

Because the eleventh forward would come out of the goalie or the fifth defenceman, and both of those are doing real work.

Forwards outscore everyone in a points-only format, they fill the flex, and taking more of the highest-scoring position feels strictly good. But your sixth-best forward in a given week is already not starting. Adding an eleventh means the marginal player you gained competes for a slot that five other forwards are already fighting over, while the slot you stripped had no depth behind it at all.

Ten is just what falls out when you give every startable slot roughly double coverage and hand the extra to the positions that cannot absorb a bad week.

What would make me wrong

• If goalies flex more than I think. A goalie can occupy a flex slot in this format, but it is rare in practice. If your scoring settings make a goalie a routinely correct flex play, the goalie slot count goes from one to two, the coverage maths shifts, and three goalies becomes light rather than sufficient.
• If your league scores defencemen closer to forwards. These ratios come from points-only pools where forwards outscore the blue line comfortably. A format rewarding blocked shots and hits makes the fifth and sixth defencemen far more valuable and pulls a forward off the roster.
• If your pool is managed weekly rather than best ball. Everything above depends on scores being taken automatically. The moment you have to guess which goalie plays Tuesday, depth stops being free insurance and starts being a weekly decision you can get wrong.
• If the room's split is convention rather than reasoning. I have shown that 10/5/3 matches what the lineup rules imply. I have not shown that anybody arrived at it deliberately. It is possible the market is right by habit, and habit is a weaker guarantee, since it would not adjust if the format changed.

The short version

I went looking for an inefficiency in how the room builds an eighteen-man best ball roster and did not find one. Ten forwards, five defencemen, three goalies is what the average team drafts, and once you divide by startable slots rather than counting players, it is close to what the format asks for.

The rule underneath it is the part worth keeping: cover every slot about twice, and give the extra coverage to the positions you can only start once. That is why three goalies for one slot is not waste and an eleventh forward is not a bargain.