The points are being extended down an extra two positions, but there are going to be more teams lining up on the grid. Depending on how many teams actually make it to the first GP, it could be that it's the same proportion of starting cars that are receiving points at the finish. All the best, Andrew.
... the new one's even worse: First = 25 Second = 18 Third = 15 Fourth = 12 Fifth = 10 Sixth = 8 Seventh = 6 Eighth = 4 Ninth = 2 Tenth = 1 Biased even integers, F1's new reward system is can of worms. Only way to score an odd number of points would be to finish 1st, 3rd, or 10th, an odd numbers of times. Final world championship will inevitably wind-up, a preponderance of drivers and constructors with equal scores, having accumulated a preponderance of even integer sums in their final points tally. The more prolonged the Grand Prix season, the all more skewed will be the championship, biased even integers. Those people don't know beans about quantitative analysis. Point regimens are archaic. Wholly unnecessary. Proper way to do it, how to calc who really won the thing -- just plug 'n chug the ordinal values (e.g., 1st, 2nd, 3rd, 4th place...), as though you're crunching the harmonic mean. It's perfect. Works every time. Regards, ~ Samuel
Samuel, Do you mean you could give a point relative to the position? 1st gets 1 point, 2nd gets 2 points, etc... Then, as in golf, the person with the lowest score wins? All the best, Andrew.
It's really simple. We've been doing this decades, evaluating average qualifying position, to ferret out who the fast guys really are. They should have been scoring the world championship like this, decades ago. Goes like this: y = #GPs/[(1/first race) + (1/second race) + (1/third race)...] Doing this RPN, just enter the number of races, then divide it, by the summation of the inverse of each ordinal value. Use the inverse key on your calculator. Take the set of ordinal values over 5 races, say, a 1st place finish for the first GP, a 2nd place in the second race, then 1st, then 3rd, and then 19th, for example. Right? You plug and chug, to ascertain the harmonic mean, like this: y = 5 races / [(1/1th) + (1/2nd) + (1/1st) + (1/3rd) + 1/19 ... ] y = 1.732 See what I mean? That is, average finish would be better than a second place average. The driver with the highest ordinal ranking (e.g., the lowest value), should be the one who wins the championship. Beauty of the harmonic mean, limited distortion. Harmonic mean is resilience to statistical outliers. Contrast, just one 19th place finish, calculating rudimentary midpoint theorem, arithmetic average finish in each Grand Prix... (1+2+1+3+19)/5 = 5.2 (?) ... is a gross distortion (e.g., right skewed). To say the driver averaged 5th place would be an aberration. Midpoint theorem is a poor proxy, for the true mean (e.g., the truth). Example given, this driver merits somewhat better than a 2nd place finishing average. Not somewhat worst, than 5th. Not only does harmonic mean peg the true champion, without fail, every single time? It ferrets out the fastest driver, every single time. Pinnacle of motorsport, with a NASCAR style points-based scoring system from the stone age -- people think guys in F1 are smart. If they really were, then I think they would have figured out something like this, 50 years ago. Acid test of any points based reward system is evaluated, by how close it rank orders, relative to the harmonic mean. As it is, new point system they've proposed -- it's a train wreck. Thing about train wrecks -- point of impact is always hundreds of meters downrange, the point of derailment. Regards, ~ Samuel
Thanks for the explanation Samuel - it's a good idea for accurately determining who has the best performance throughout a season. Perhaps the only problem with it is that it does take some mathematical skills to be able to follow it. The average sportsfan can easily appreciate winning a set number of points for a certain placement in the race, but it's more difficult for them to comprehend mid-race how various positions being held by title contenders are affecting the championship. The TV networks would have to be on the ball to give a current calculation for the championship standing for each driver if the race finished at that moment. While very accurate, it could be difficult to communicate and confusing for the newly initiated. All the best, Andrew.
Trick is, devising a points-based system which approximates the harmonic mean. Points based scoring systems are inherently defective, already. All the worse, the latest proposal -- there needs to be more odd number value integers in the reward system. Latest proposal, heavily biased even integers, is prone to (1) exacerbate natural breaks between similar drivers, and (2) agglomerate statistically dissimilar performance, on coincidental even par values (e.g., muddies the water). Of all F1's reward systems, this one worked best: 1st = 9 points 2nd = 6 3rd = 4 4th = 3 5th = 2 6th = 1 Denote the 1:1 parity, even to odd integers. A nice, neat, simple, eloquent linear decay, for a points-based scoring system, that one is nearly impossible to improve upon. Were I encharged the task of improving it, to reward points to 10th place? I'd multiply the old 9-6-4-3-2-1 by three. My two cents, I've always felt points should be doubled for any Grand Prix declared a wet race. Regards, - Samuel
... a contradiction looms. See for yourself, if and when you find time: http://en.wikipedia.org/wiki/2010_Formula_One_season#Drivers After three Grand Prix, Felipe should not be leading Fernando, if Nico trails Jenson. Yet another contradiction, Robert's beaten Lewis, twice. Like Nico, who's beaten Jenson two out of three GP, Robert trails Lewis in his championship point tally. F1's new point system's discrepancies loom. If Felipe indeed leads Fernando, having beaten him two out of three GPs, then so too should Robert -- Lewis, and Nico -- Jenson. Worse, statistical reality, Red Bull leads McLaren after three GP. New points system artificially inflates in importance, McLaren, in the world constructors' championship. best, ~ Samuel
Personally, I like the new system. I like the bigger gap from 1st to 2nd and it makes more sense to award points through 10th rather than 8th...
These are racers, not mathematicians. (Everybody hated "probability and sadistics" class. ) The old 6-4-3-2-1 system was based on six races a year. Two points were a significant component of the total. Two points between first and second was significant in a 24 point total. Not so much so, when drivers get over 100 points in a season. With 19 races in a year, points inflation was just responding to the number of races, more than the relative merit of the positions. What's odd is that FIA imposed these new points over WRC too --- and the last rally had only 11 entries in WRC (and "Jr") classes ... and only nine finished. (One point went to a local club rally driver.) (Using the same points across all FIA series is what makes points for "pole" and "fast lap" indigestable to FIA.)
