# Naturally, I see it differently

## Posted on September 24, 2003 at 04:41:02 PM by Craig G

All mechanical systems and formulas are going to have their weak
points and blind spots.

If your system is "win only", then any sub-population
of players who have trouble winning but consistently "go
like hell" when PS is on the line, will not be handled correctly.
Assuming they exist.

However, I have different reservations about the Skiena system.
Where you are concerned about its failure to distinguish between
a 7 point win and a 4 point one, (how often does that occur?)
I see big problems at the other end.

Suppose you attended live, and said "Wow, the wheels have
really fallen off for Megazola, he's only scored 3 points in his
last 10 games, and those were gifts!". Well, the Skiena numbers
would be awarding him the same number of points won as players
who were playing competitively but finishing OOTM.

I generated my own Skiena-style numbers and used a Factor X table
to do a test case:

5- 89 185
4s 223 176
3- 93 121
8- 64 144
1- 92 235
5s 223 155
7- 76 152
6- 84 168
4- 92 205
3- 93 121
=====================
1129 1662

So our observationally-hopeless guy appears to be winning 40.5%
of his points. (11.3 vs 16.6 ) That seems like a bigger problem.

Another concern is that the system appears to regard winning from
an 8 as a bigger accomplishment than 5, 6, or 7, whereas the reality
of quiniela jai-alai is just the reverse.

Of course, the real test of any system is how well it works against
the competition. Apparently it works very well. Since I haven't
read his book, I'll refrain from any further comments until such
time as I get a copy.