"My rule was I wouldn't recruit a kid if he had grass in front of his house.
That's not my world. My world was a cracked sidewalk." —Al McGuire

Thursday, December 31, 2009

RPI is killing MU - Schedule tougher Cream Puffs!

At this point in the season, MU is so far ahead of my expectations coming in and DJO is such a better player than I could have hoped for to give us a third threat, that I can’t believe I am contemplating that we could get an NCAA bid this year. In addition, even if we sneak in as a 12-seed, I think this is the kind of team that could shock a 5-seed in the opening round.

Unfortunately this leads me to my one criticism of the program – scheduling. If we continue to play this well, we could head into Selection Sunday on the bubble, and if that is the case then the fact that we played Presbyterian, Centenary, MD Eastern Shore and Grambling should cost us the bid.

Remember that the biggest frame of reference for the selection committee is the RPI, which doesn’t care about margins of victory. There are two things that make you drop in the RPI – losing a game OR playing a team not in the Top 200 (give or take a few spots) at home regardless of the result. Pomeroy (www.kenpom.com) is a much truer measure of how good a team is because it accounts for victory margins, so beating Presbyterian by 40 or losing at West Virginia by 1 point both make you move up in the standings. But for the RPI, both make you move down.

So in short, Marquette has the worst possible schedule for the RPI. Playing in the Big East every year we play a ton of games we could lose (the first way to drop) and then we schedule six home games against teams nowhere near the Top 200 that will make us drop substantially even if we beat them 100-0.

As a result, MU is the 22nd best team in the US according to Pomeroy, but only the 81st best team in the land according to RPI, which is about 31 spots below where you typically need to be to get an at-large bid.

I know I harped on this last year, but I want to walk through the numbers once from my RPI spreadsheet in an urgent plea to stop scheduling these bottom 100 teams and just move the cupcake schedule up one notch so we are scheduling the teams that are in the top 200.

For arguments sake, let’s say the Pomeroy projects play out for everybody and MU finishes a strong 20-10.

The first thing you need to understand is that the RPI calculation only gives us credit for 0.6 wins for a home win (and likewise we only are given 0.6 losses for the games at WVU and Wisconsin, but get 1.4 losses for NC State and will get 1.4 wins for each road victory). Therefore, despite beating up on the four previously mentioned cupcakes at home, the 20-10 projection would have us going 13-4 at home, 2-1 on neutral courts, and 5-5 on the road, which once adjusted for home court means the RPI formula just considers us a 17-10 team.

If that were to play out, then the RPI calculation works like this (ignoring the Big East tourney for the moment):
MU real record 20-10
RPI record 16.8-9.6 (25% of formula)
Opponents record 550-350 (50% of formula)
Opponents’ opponents’ record 15,000-12,000 (25% of formula)
Final RPI: .5694, 68th best, about 18 spots out of tourney

Now, to make my point, let’s pretend we didn’t play Presbyterian, Centenary, MD Eastern Shore and Grambling so I drop all their projections out of the equation. Even though we give up four wins (or 2.4 wins on RPIs scale), look what happens to our RPI calculation once we get all of their losses plus the losses of all their opponents out of the equation. Now this is what the NCAA selection committee is reviewing.

MU real record 16-10
RPI record 14.4-9.6 (25% of formula)
Opponents record 513-272 (50% of formula)
Opponents’ opponents’ record 14,000-10,000 (25% of formula)
Final RPI: .5976, 44th best, easily in the tourney

Just by NOT PLAYING those four games we would move up about 24 spots in the RPI and clearly be in the tournament. Now I realize it isn’t really practical to just play 26 games. But let’s say we upgraded just a little – not to the Top 200 which I advocate, but added four more games against teams just as good as South Dakota and North Florida. At that point the end of year RPI equation – getting our four wins back but against slightly better competition, looks like this:

MU real record 20-10
RPI record 16.8-9.6 (25% of formula)
Opponents record 579-324 (50% of formula)
Opponents’ opponents’ record 14,500-11,000 (25% of formula)
Final RPI: .5877, 50th best, on the bubble but 20 wins and in Big East should pull us through.

Obviously realistically we need to schedule six creampuffs that we know we will go 6-0 against, particularly in a year like this where Buzz needs to experiment with new players.

How risky would playing Top 200 creampuffs be? Since 2005, MU has gone 27-2 when playing a home game against a team ranked between 101st and 200th in the land in Pomeroy, with the only two losses being to St. Louis after Travis Diener broke his hand and we had no point guard, and to a North Dakota State team that went on to go 20-8 and was ranked 127th.

Going 6-0 against Top 200 creampuffs instead of teams ranked from 250th to 347th will typically move us up 30 spots in the RPI each Selection Sunday, and even if we get tripped up and go 5-1 against Top 200 Creampuffs, we will be much higher than being dragged down by six wins against bottom 100 teams.

That being said I really am hopeful that we will be in the NCAA tourney with this team ahead of schedule, but for future years we will make tournaments and get much better seeds if we upgrade our Creampuff schedule even a little bit.

We aren’t the only team. Here are the 10 teams most hurt by scheduling based on comparing how good they really are by their Pomeroy ranking to how many spots lower they are in the RPI due to bad scheduling:

Team, Pomeroy rank, RPI rank, difference due to scheduling weak opponents:
UTEP 38, 155, -117
Utah State 50, 146, -96
Arizona St. 21, 114, -93
Missouri 12, 104, -92
Wright St. 40, 125, -85
Memphis St. 20, 94, -74
Ohio State 18, 89, -71
Georgia Tech 24, 92, -68
Marquette 22, 81, -59
Minnesota 7, 61, -54

1 comment:

Ryan said...

Enough...let them play the games