Gravitywerx consistency experiment
An oxymoron if there ever was one! The greatest experience in die-cast racing, Gravitywerx, has been criticized for not providing a platform that a competitor could prepare for...for not providing some kind of absolutism for the purpose of "fairness" and consistency. Single file start? 10 cars at once? Cars finishing backwards and upside down? No electronic scoring?
I decided to run a duplicate round 5A race with the same lineup which was (from pole): 27,29,4,22,6,13,21,16,9,1. The first race (the one that counted) results were: 13,9,1,22,16,29,21,27,6,4. The second race results were:22,4,16,6,9,13,21,27,29,1.
In the first race, the points leading # 1 car ran a great race, finishing 3rd and retaining the points lead. In the second race, it crashed 3 times and finished last! It crashed as many times as the rest of the field combined! Unlike real racing, where the weather may be different, or the temperment of the drivers, there were no environmental differences between these 2 races.
What it does show is that the # 1 car is not as stable as the other cars at the top of the points chase, and it's hold on # 1 may be fleeting. 5 more rounds to find out!
Discussion
In the interest of furthering science, I ran a 3rd duplicate 5A race. The finish order in this one was closer to the second with the # 22 car winning again (that # 22 car is HOT in races that don't count!). The # 1 car crashed several times again but had a decent finish by staying on it's wheels in the last laps while others didn't. # 22 even crashed once. The finish order was: 22,13,27,9,1,29,4,16,6,21. Lap-by-lap leader was 22,22,13,13,22,22,16,16,16,22,22. Race took one extra lap to satisfy backup win requirement.
Voila! I finally fligger'd out why the results are so varied. It's the multiple runs! A slight difference in a single run spills into the lineup for the next one. Kind of like the whispered message gets jumbled as it passes between people. The more people, the greater the deviation. I'm going to take some stills where the cars first break from single file and compare them to see where the trouble starts.
I ran 5 identical 5 A starting lineups to compare the amount of position-swapping for one lap. There are a possible 50 swaps that can happen between 10 cars. In the 1st test there were 22 swaps between 7 cars. Sounds like a lot. But when you consider that the lead car crashed and dropped to 9th, that is 8 right there. And 9 cars moved up a spot. That is 9 more. But some didn't move up because they were passed. That is why only 7 cars moved. Some moved up or back several spots. the # 9 car moved up 4 spots. # 21 moved back 3.
Also what needs to be considered is that the single-file lineup has the lowest-points cars in the front and the winners in the back. Both ends are going to be headed toward the middle of the lineup right from the start. In every test, one of the 2 front cars dropped out of the top 3 spots. The pole sitter lost 14 spots between the 5 tests! 2nd in line lost 9, while the # 9 gained 9 spots.
The other test lap results were: 14 swaps between 7 cars, 14/5, 14/9 and 10/6.
I used the SAVE function when I was almost done in the last thread and it disappeared! It's never as good the second time! Anyway, I failed to answer the question 'why the wildly differing results, particularly with the # 1 car completely self destructing?'
It would be the same answer as to why the # 1 car led for 9 laps in round 1 only to crash on the last one.The track it was taking was causing it to hit the wall at a bad angle or maybe a bad spot on the track. It might have gotten a really good run out of a turn causing it to hit the next one too fast. It's hard to tell, and I'm not the one to dwell on it. I actually like it when a good car has a bad day. If the bad days are frequent, it wasn't really a good car to begin with.
Couple of other observations: The # 22 and # 9 have yet to crash in 50 trips down this cruel track. That would make them favorites for the championship. Staying on your wheels is worth more in this game than straightaway speed.