Hey Mink,
That is the right answer, for sure. (And I happen to have a big
chunk of supporting data)
There are 2 main reasons, the second of which might be controversial,
but still in my opinion correct.
First, as you say, the stars in the late games tend to have the
higher win percentages, which in turn means higher percentage
of points won, which then means greater likelihood of a runout.
And, of course, while the too-good-for-the-competition early game
stars can be promoted to middle or late games, there are no late-late
games for promoting Goiko or Lopez. So they stay right there in
the lates and dominate forever.
Second, instead of the usual Skiena-style rating where one number
is used to express the strength of a player or team, IMO the underlying
REALITY of quiniela-based jai-alai is better modeled in terms
of QUANTUM handicapping, in which we think in terms of levels
or states.
So with the conventional, somewhat static view, you might say
that Goiko / Lopez / Erik / Diego win (for example) 62% of their
points and then you would try to calculate a runout probability
based on that.
But with a 'quantum' view, you allow for a player to
be in different 'states', such as 'tired from the
double performance yesterday', 'not warmed up yet',
'struggling with an injury', 'bouncing back with
a vengeance', 'totally kick-ass' or whatever might
match up with reality and have PREDICTIVE VALUE.
With this perspective, the stars such as Goiko / Lopez /Erik /
Diego might be in the kick-ass state 20% of the time, and in THAT
state their point-winning rate might be closer to 80% or so.
The whole point of this is that if we interpret their overall
performance in terms of a mixture of states, then the occasions
when they are in their very top gear will produce a higher yield
of runouts than would a steady-state view.
Which simply means that an objective analysis of jai-alai reality
would reveal a higher rate of runouts than expected by the conventional
view.
