Just thinking if there's an objective metric on how to quickly evaluate Quad 1/2/3/4 wins and losses to give a sense of the strength of a resume for NCAA seeding.
Let me try this and see if you buy it.
Quad 1: Win = +4, Loss = -1
Quad 2: Win = +2, Loss = -2
Quad 3: Win = +1, Loss = -4
Quad 4: Win = 0, Loss = -6
I'll run through the NET Top 20 and see how it maps to BracketMatrix (not fully updated so it won't be perfect...looking at you, UNC)
2. Purdue (40 points) - #1 S-Curve
1. Houston (32 points) - #2 S-Curve
4. UConn (32 points) - #3 S-Curve
11. Wisconsin (28 points) - #6 S-Curve
13. Kansas (22 points) - #8 S-Curve
14. Marquette (22 points) - #9 S-Curve
3. Arizona (21 points) - #7 S-Curve
9. UNC (21 points) - #4 S-Curve
17. Dayton (20 points) - #16 S-Curve
5. Tenn (18 points) - #5 S-Curve
18. Duke (17 points) - #15 S-Curve
20. Utah St. (17 points) - #22 S-Curve
15. Creighton (16 points) - #10 S-Curve
12. Illinois (15 points) - #18 S-Curve
16. New Mexico (15 points) - #27 S-Curve
10. Iowa St. (14 points) - #13 S-Curve
7. Alabama (13 points) - #11 S-Curve
8. Auburn (13 points) - #17 S-Curve
19. Baylor (13 points) - #14 S-Curve
6. BYU (12 points) - #19 S-Curve
I don't have Kentucky (21 NET, #12 S-Curve, 12 points) or San Diego State (23 NET, #20 S-Curve, 16 points) on here but otherwise the Top 20 NET mostly map to Top-5 seeds, just not in the right order.
This metric is probably a better correlation than the NET rankings (especially for the very best teams) but obviously still in need of some tweaks.
The main point is that teams racking up Quad 1 wins will rocket up the S-curve compared to ones that don't, even if they do well with the predictive metrics like KenPom (and sort of NET.)
Let me try this and see if you buy it.
Quad 1: Win = +4, Loss = -1
Quad 2: Win = +2, Loss = -2
Quad 3: Win = +1, Loss = -4
Quad 4: Win = 0, Loss = -6
I'll run through the NET Top 20 and see how it maps to BracketMatrix (not fully updated so it won't be perfect...looking at you, UNC)
2. Purdue (40 points) - #1 S-Curve
1. Houston (32 points) - #2 S-Curve
4. UConn (32 points) - #3 S-Curve
11. Wisconsin (28 points) - #6 S-Curve
13. Kansas (22 points) - #8 S-Curve
14. Marquette (22 points) - #9 S-Curve
3. Arizona (21 points) - #7 S-Curve
9. UNC (21 points) - #4 S-Curve
17. Dayton (20 points) - #16 S-Curve
5. Tenn (18 points) - #5 S-Curve
18. Duke (17 points) - #15 S-Curve
20. Utah St. (17 points) - #22 S-Curve
15. Creighton (16 points) - #10 S-Curve
12. Illinois (15 points) - #18 S-Curve
16. New Mexico (15 points) - #27 S-Curve
10. Iowa St. (14 points) - #13 S-Curve
7. Alabama (13 points) - #11 S-Curve
8. Auburn (13 points) - #17 S-Curve
19. Baylor (13 points) - #14 S-Curve
6. BYU (12 points) - #19 S-Curve
I don't have Kentucky (21 NET, #12 S-Curve, 12 points) or San Diego State (23 NET, #20 S-Curve, 16 points) on here but otherwise the Top 20 NET mostly map to Top-5 seeds, just not in the right order.
This metric is probably a better correlation than the NET rankings (especially for the very best teams) but obviously still in need of some tweaks.
The main point is that teams racking up Quad 1 wins will rocket up the S-curve compared to ones that don't, even if they do well with the predictive metrics like KenPom (and sort of NET.)
