Top level play Pavlicek data

[PhilKing]

There is one huge anomaly - balanced 20-21 counts.

 

They appear to perform massively better in first seat than second, yet the 18-19 and 22-23 ranges perform about the same. Anyway I'll chew it over. Maybe it's just a horrible range to define over first seat preempts. I'm prepared to listen to arguments that 424 and 382 hands is not close to enough hands to draw a firm conclusion, but a difference of 82 points a hand seems like a lot to attribute to variance.

[cherdano]

There is one huge anomaly - balanced 20-21 counts.

 

They appear to perform massively better in first seat than second, yet the 18-19 and 22-23 ranges perform about the same. Anyway I'll chew it over. Maybe it's just a horrible range to define over first seat preempts.

Well, the other difference is, of course, that most of us have an efficient way to show a 20-21 balanced hand in first seat, and not a great way to show 18-19 balanced (especially in competition).

[benlessard]

1- a random balanced hand is negative in 1st and 2nd seat is expected, having a balanced hand is weaker both in offense and in defense (suits break evenly)

 

2- Having only 1 pts higher is more significant than i previously tought. From -28 (random) to +3 (10-11) to +100 (12-14) for 31 pts and a 97 pts are big gaps in expected score only for half a pts and a few pts on hands that are a close to average. The 97 gap is a little bit surprising I guess the the odds of making game your side combined with the lower odds of a game against you make a significant difference with every pts.

 

short on time right now ill add other stuff later.