Evaluation of benefits from mixed PWD racing methods
Stan DCmr (StanDCmr@AOL.COM)
Tue, 30 Dec 1997 12:36:21 -0500
Purpose:
Assess alternatives for awarding 4 racing trophies and qualifying
4 racers to district races among 28 cars using a 4 lane track.
Summary:
When larger numbers of cars are involved, substantial time savings,
accuracy, and/or flexibility can be accomplished by combining the
strengths of different racing methods.
Accuracy measures used:
4-Trophy - Percentage of time that car among the 4
objectively fastest receives correct trophy.
Top-4 - Percentage of time that car among the 4
objectively fastest placed in top 4.
4nX - Percentage of time that car among the 4
objectively fastest placed in top X.
Alternative 1: Use only a 13 Round Stearns Chart
13 Round Stearns Accuracy (91 heats):
4-Trophy: 65.28% Top-4: 89.68%
Error Rates:
4-Trophy: 34.72% Top-4: 10.32%
Alternative 2: Use a 10 round Stearns Chart to select 13 finalists;
Race the finalists using a Perfect 13-4 (1) Chart.
10 Round Stearns (70 heats):
13-Trophy: (ignored) Top-13: (ignored) 4n13: 100.00%
P13-4 (1) car final (13 heats):
4-Trophy: 74.50% Top-4: 94.28% 4n4: (ignored)
Composite Accuracy (83 heats):
4-Trophy: 75.50% Top-4: 94.28%
Error Rates:
4-Trophy: 24.50% Top-4: 5.72%
Alternative 3: Use a 10 round Stearns Chart to select 7 finalists;
Race the finalists using a Perfect 7-3 (1) Chart.
10 Round Stearns (70 heats):
7-Trophy: (ignored) Top-7: (ignored) 4n7: 99.83%
P7-3 (1) car final (7 heats):
4-Trophy: 80.90% Top-4: 96.10% 4n4: (ignored)
Composite Accuracy (77 heats):
4-Trophy: 80.76% Top-4: 95.94%
Error Rates:
4-Trophy: 19.24% Top-4: 4.06%
Alternative 4: Use a 10 round Stearns Chart to select 7 finalists;
Race the finalists using a Complementary-Perfect 7-3 (2) Chart.
10 Round Stearns (70 heats):
7-Trophy: (ignored) Top-7: (ignored) 4n7: 99.83%
P7-3 (2) car final (14 heats):
4-Trophy: 85.30% Top-4: 97.50% 4n4: (ignored)
Composite Accuracy (84 heats):
4-Trophy: 85.15% Top-4: 96.60%
Error Rates:
4-Trophy: 14.85% Top-4: 3.40%
Conclusions, comparing Alternative 4 to Alternative 1:
Combining a longer moderate precision method for screening
with a short high precision method for trophy assignment
reduces the trophy assignment error by 57% and reduces the
representation errors by 67% with no significant difference
in racing time.
In more concrete terms...
reduce the frequency of awarding a Scout the wrong trophy
(or NO TROPHY AT ALL) from more than 1 time in 3 to about
1 time in 7; or
reduce the frequency of not sending a deserving Scout to
the district race from about 1 time in 10 to about 1 time
in 30.
Sources:
Stearns: 28 car chart (Stearns and Peronto) as published in
"http://members.aol.com/houraweek/pack339/339pine2.html"
but scored 4-3-2-1.
P13-4 and P7-3 charts (Young and Pope) as published in
"http://members.aol.com/StanDCmr/genframe.html"
Individual method evaluations from pre-production version of a
program by Cory Young (cyoung@dcn.att.com) using the following
parameters:
Mean St.Dev
----- -----
Car Speed 10.0"
Lane Speed 1.0"
Lub Breakin .5" .2"
Lub Degradation .5" .2"
Threshold 2 .1"
Head to Head Tie-Break
Random element .5"
Trials 1000
Cory advises of program availability at "Software" in
http://www.geocities.com/Yosemite/Rapids/7146/
Details of study available from StanDCmr@aol.com or (soon) at
http://members.aol.com/standcmr/pwmeval.html
Analysis by Stan Pope and Cory Young.
Terry Howerton Sakima Group, Inc. SCOUTER Magazine Kansas City | 
