Additional Analysis

Does Draft Position Matter?

We wanted a more confident answer than either league alone could give, so we looked at draft position across 272 team-seasons spanning two different fantasy leagues — Whiskey League and KAW (Keep Austin Weird), a separate 12-team league some of our own managers also play in. Each league alone only had 128 or 144 team-seasons to work with; pooling the two gives a considerably clearer read on whether where you draft from actually predicts how your season goes.

Playoff Gap

16.7 pts

Early vs. Late playoff rate

Middle's Title Share

58.3%

of all 24 titles, from just 34.6% of team-seasons

Pick #1 vs. Pick #12

75% vs 38%

playoff rate — almost double

Early / Middle / Late, Combined

Same bucket definitions used on each league's own page — first four picks, middle four, last four (12-team-equivalent) — summed across both leagues.

Bucket	Playoffs	Finals	Title Rate	Last Place	Win %
Early n=90	62.2%	18.9%	9.4% (8.5)	8.9%	51.4%
Middle n=94	56.4%	19.2%	14.9% (14)	6.4%	51.9%
Late n=88	45.5%	14.8%	1.7% (1.5)	11.4%	46.6%

Early and Middle are close on playoff rate (62.2% vs. 56.4%) and win percentage (51.4% vs. 51.9%) — the real gap is between those two and Late, which drops to 45.5% playoffs and a 46.6% win rate. This replicates the shape each league showed on its own: WL alone (56.8% / 57.1% / 40.5% playoffs by bucket) and KAW alone both had Late as the clear outlier, just with more noise individually.

Championships Don't Follow the Same Curve

Playoff odds decline steadily from Early to Late. Titles don't — Middle actually leads.

Title Rate by Bucket

Early

9.4%

Middle

14.9%

Late

1.7%

Middle has produced 14 of the combined 24 titles (58.3%) despite being only 34.6% of team-seasons. One read: the very top picks provide a playoff floor, but Middle picks may combine strong-enough talent with better snake-draft positioning for a true title run. That's a plausible story, not a proven one — the data shows the pattern, not the mechanism.

Pick-by-Pick, Combined

All 12 slots, 24 observations each (fewer at slots 4 and 9, where 10-team seasons don't map onto an exact 12-team-equivalent position).

75%16.7% title raten=24

66.7%12.5% title raten=24

54.1%2.1% title raten=24

50%6.3% title raten=16

58.3%12.5% title raten=24

62.5%10.4% title raten=24

50%20.8% title raten=24

54.1%14.6% title raten=24

62.5%6.2% title raten=16

50%2.1% title raten=24

37.5%0% title raten=24

Pick #7 has the highest combined title rate of any slot (20.8%, 5 of 24 titles) — worth noting, but with only 24 observations per slot, a single title swings a slot's rate by roughly 4 points. Treat any one slot's number as a data point, not a verdict; the bucket-level pattern above rests on a much sturdier sample.

A Replication, Not Just an Aggregate

Draft position's correlation with season outcomes, computed independently in each league — same direction in both, but only Whiskey League clears conventional statistical significance on its own.

League	vs. Finish	vs. Win %	vs. Scoring
Whiskey League n=128	r=0.227 p=0.01	r=-0.217 p=0.014	r=-0.18 p=0.042
KAW n=144	r=0.116 p=0.165	r=-0.094 p=0.262	r=-0.057 p=0.499

Positive r on “vs. Finish” means a later draft slot correlates with a worse (higher-numbered) finish; negative r on win % and scoring means later slots trend toward fewer wins and fewer points. Both leagues agree on direction for all three. KAW's smaller, noisier league doesn't reach significance on its own (p > .05 throughout) — but a same-direction result in an independent dataset is exactly what you'd want to see if Whiskey League's effect is real rather than a fluke of one league's history.

Methodology & Caveats

This page pools each league's already-published Early/Middle/Late and per-slot summary tables, weighted by team-seasons — it is not a single regression run against a combined raw dataset. That's a meaningfully different (and weaker) form of evidence than a proper pooled model.

A rigorous next step would be a regression on the raw per-team-season rows from both leagues together, controlling for league and season, and explicitly testing whether the penalty for drafting late ramps up gradually across all twelve slots or kicks in specifically in the final third. The bucket data above is consistent with the second story — Early and Middle track each other closely, and Late is where things fall off — but that hasn't been formally tested yet.

One definitional note: each league's own “Ends vs. Middle” table (not shown here) uses a different grouping — Whiskey League's Ends is a clean Early+Late split (n=86), KAW's Ends is a narrower group (n=52). Those two tables should not be merged directly; this page only combines the Early/Middle/Late three-way split, where both leagues use the same definition.

Descriptive, not causal — same as every other draft-position stat on this site.