WSOB 2010 Statistics and PBA Pattern Observations

Ask Mo Pinel and the bowling industry's best your general questions, and get straight answers.

Moderators: Mo Pinel, purduepaul, MathIsTruth, ballspinner

Forum rules
Ask Mo Pinel and the bowling industry's best your questions, and get straight answers.
This forum is moderated exclusively by Mo & Friends.
Locked
User avatar
purduepaul
Member
Member
Posts: 219
Joined: June 26th, 2009, 3:13 am
THS Average: 215
Positive Axis Point: 4 5/8 over by 5/8" up
Speed: 19
Rev Rate: 300
Axis Tilt: 13
Axis Rotation: 65
Preferred Company: Radical Bowling

WSOB 2010 Statistics and PBA Pattern Observations

Post by purduepaul »

I have always been an advocate that patterns should not only be classified by side to side ratios, but also length and volume. Our simplistic view of patterns currently is not enough. So using some of the skills I have learned, I will give you an analysis of the pattern scoring at the WSOB in Vegas. Now some assumptions have been made (players all have the same opportunity to score, patterns were the exact same...etc.) Here are the observations.

First, the set included 229 bowlers who bowled 12 games on every pattern (That is important). Since the sample size is 12 games by each, it is over the minimum of 10 so our error is within specification. Here is the data.

Average on Pattern:
Shark: 196.51
Scorpion: 203.41
Chameleon: 200.06
Viper: 201.33
Cheetah: 215.93

As you can see we have a good range in averages, lets now run a paired t test in order to see if we can statistically say that the patterns are different. We will now start at the low and go to high to compare.
Paired T for Chameleon - Shark

N Mean StDev SE Mean
Chameleon 229 200.06 14.34 0.95
Shark 229 196.51 17.90 1.18
Difference 229 3.550 11.562 0.764
95% CI for mean difference: (2.044, 5.055)

If the confidence interval includes 0 in the range, then we can not say the patterns are different, for this one we can say that on average there is a 3.55 pin difference.

N Mean StDev SE Mean
Chameleon 229 200.057 14.344 0.948
Viper 229 201.331 13.484 0.891
Difference 229 -1.274 10.582 0.699


95% CI for mean difference: (-2.652, 0.103)
T-Test of mean difference = 0 (vs not = 0): T-Value = -1.82 P-Value = 0.070

Here is an example of us not being able to say there is a statistical difference between Chameleon and Viper. This is due to the 95% CI including 0.

N Mean StDev SE Mean
Scorpion 229 203.41 15.45 1.02
Viper 229 201.33 13.48 0.89
Difference 229 2.077 11.876 0.785


95% CI for mean difference: (0.530, 3.623)
T-Test of mean difference = 0 (vs not = 0): T-Value = 2.65 P-Value = 0.009

Again there is a statistical difference between these two patterns.

Paired T for Scorpion - Cheetah

N Mean St Dev SE Mean
Scorpion 229 203.41 15.45 1.02
Cheetah 229 215.93 15.91 1.05
Difference 229 -12.527 13.480 0.891


95% CI for mean difference: (-14.282, -10.772)
T-Test of mean difference = 0 (vs not = 0): T-Value = -14.06 P-Value = 0.000

This is the one where the biggest difference in any two patterns is and the cheetah will statistically 12.5 pins higher than Scorpion.

By this analysis, it can be said that maybe the pattern experts out there should look and actually investigate what contributes to pattern scoring and what does not contribute. Since obviously side to side ratio is not the only thing!

Paul
Paul Ridenour
former Sr Research Engineer at USBC
Radical Bowling Technologies Staffer
Locked