Ken Pom Bracket Strategy

[xxplosive]

I have been fiddling around with the brackets and haven't even come close to winning one in like 8 years when I took it all. I put together a bracket using the ken-pom power rankings and as you can assume all number one seeds advanced. I like Ken-Pom system but does really account for injuries and what not?

I mean it gives a few upsets, but its all 1,2,3's in the elite 8.

Anyone here have success using Ken-Pom for bracket filling out?

[Chandler]

of course if you follow that format all 1's, 2's, 3's will go to the elite 8

its power rankings over a long term (season wise anyways.

if the games were best of 7 series hypoethically. the top seeds would almost always progress.

but that is the Beauty of one and done Tourneys, and emotional college kids ect.

ANYTHING can happen. and short of a 1 seed getting beat by a 16 seed. We've seen just about everything happen over the years in the big dance.

I would use KenPom as a tool, but look at alot of other factors in making your picks too. especially how teams are playing the last half of the season. free throw %'s for close games between evenly matched teams.

good gaurds are very important. good assist/turnover ratios.

for upsets you might wanna pick, find teams with alot of good or at least streaky-good 3 point shooters. not just one good shooter, but multiple ones.

take into account, location and fanbase when making picks on close games.

good coaches, get good games from their teams in March.

for me, the most important thing. is to pick teams that I really enjoy watching and like to root for.

nothing sucks worse than having to root against a fun underdog that is going deep in the tourney. because it will ruin your bracket.

have fun with it

[pavyracer]

Teams play 2 games in 3 days on the road this week. It's very easy for an upset to happen. Models are good if you have data for 30 games with similar conditions. But when teams fly in unfamiliar territory and stay in a hotel for 5 days on the road how to you factor it in your model?

[MartinBlank]

2008 was the only year that I can recall all the number 1 seeds advancing to the Final Four.

An overwhelming majority of years, you are going to have upsets.

[Chandler]

in my model, a road game is a road game and a neutral court is a neutral court.

after your model spits out the #'s you can adjust for a neutral court that is a big more "homey" Like Duke first round for example.

if my model says neutral court Duke -20. while I'm thinking it is more like -22 or -23

bad example as I don't normally bet games over 11 point spreads

[xxplosive]

I was looking at a site recently called bracket science.com He does an array of different strategies for brackets. One of his strategies is just using Ken Pom's in the past 2 out of the last 3 years in ranked in the 97th percentile in ESPN bracket challenge. I believe that is pretty good but this no where near a good sample size, and the one year it didn't do well it was in the 57th percentile.

[chemicalbrother]

I was looking at a site recently called bracket science.com He does an array of different strategies for brackets. One of his strategies is just using Ken Pom's in the past 2 out of the last 3 years in ranked in the 97th percentile in ESPN bracket challenge. I believe that is pretty good but this no where near a good sample size, and the one year it didn't do well it was in the 57th percentile.

soooo, you wouldn't snap-take a 97th percentile right now?

[xxplosive]

Yes and no,
Look at the sample size I am sure someone could do that well with far different strategies? If you go with this strategy you are in a sense leaving out one of the greatest things in the tourney emotion. Do you really feel comfortable putting BYU in your Elite 8?