Scheduling for an 8 lane track using GPRM
Re: Scheduling for an 8 lane track using GPRM
Does this chart only work in multiples of 8 cars when running 8 lanes?
- Stan Pope
- Pine Head Legend
- Posts: 6856
- Joined: Sat Jul 05, 2003 7:01 pm
- Location: Morton, Illinois
- Contact:
Re: Scheduling for an 8 lane track using GPRM
No, it is good for arbitrary number of cars. The numCars/numLanes reference was an approximation.Buckeye wrote:Does this chart only work in multiples of 8 cars when running 8 lanes?
Stan
"If it's not for the boys, it's for the birds!"
"If it's not for the boys, it's for the birds!"
- Stan Pope
- Pine Head Legend
- Posts: 6856
- Joined: Sat Jul 05, 2003 7:01 pm
- Location: Morton, Illinois
- Contact:
Re: Scheduling for an 8 lane track using GPRM
Yes!!! When numCars >> numRounds * numLanes * (numLanes - 1), this approach should almost totally eliminate the desire to resequence heats! There may still be some benefit available to resequencing, but the improvement will be small enough that most folks would decide not to use it!Stan Pope wrote:Yes, starting the search using numCars/numLanes instead of 1 would usually produce better distribution during first round. After the first round, the "keep heats even" measure deteriorates because generator values get pushed farther and farther away from numCars/numLanes. Same thing happens as the number of lanes increases.Cory wrote:Stan, the phase shift ordering superiority makes me think that maybe we should change the way we look for PPN generators, e.g., perhaps all generator values should be some minimum value based on numCars and numLanes, when possible. This can help keep cars from re-racing too soon when there are lots of cars, for people who don't like re-ordering the charts.
Stan
"If it's not for the boys, it's for the birds!"
"If it's not for the boys, it's for the birds!"