Subject: Re: Nationals Predictions (Scientific!)

From: price@carla.lbl.gov (Phillip N Price)

Date: 1995/11/02

Newsgroups: rec.sport.disc

Some more Nationals predictions, courtesy of my simulations

based on Jim Parinella's input parameters:

First, who's gonna win it all?

NY : 0.41

DoG : 0.30

DH : 0.26

PCS : 0.02

Other: 0.01

The fractions for PCS and Other are uncertain due to the small

number of simulations---500 isn't really enough to give a good

estimate of the odds of an unlikely event. Actually I think PCS

should be about 0.03.

Assuming JP's odds estimates are reasonable, then I've made a

lousy bet. I got 10:1 that somebody other than NY, DoG, or

DH would win...a fair bet, if you believe JP's odds, is more like

30:1.

Other possibly-interesting quantities:

Call the "DoG" pool the pool containing DoG, DH, SD, Chain, Ring, Lemon

and the "NY" pool the pool containing NY, Z, PCS, Seattle, Cornell, Miami

By "tie" I mean teams have the same record in pool play, so the

ranking is decided based on head-to-head (in 2-way tie) or based

on within-group record or point differential (3-way tie).

Odds of 3-way tie for 1st in DoG pool: 0.05

2-way 1st DoG : 0.25

3-way tie for 2nd in DoG pool: 0.12

2-way 2nd DoG : 0.16

Odds of 3-way tie for 1st in NY pool : 0.03

2-way 1st NY : 0.14

3-way tie for 2nd in NY pool : 0.09

2-way 2nd NY : 0.21

Odds of a 3-way tie for 1st in either pool : 0.08

2-way tie for 1st in either pool : 0.36

3-way tie for 2nd in either pool : 0.20

2-way tie for 2nd in either pool : 0.34

Unfortunately, JP is out of touch (in transit to

Nationals) and I didn't ask for his permission to post

his odds matrix...I don't think he'd mind, but I don't

want someone from, say, Seattle, to say "That little

weasel only thought we had an X% chance of beating DoG?"

and do something to alter the outcome of a game or

whatever.

I will re-do the calculations with my OWN odds matrix,

which I will post tomorrow or Friday.

--Phil Price