Logarithmic ranking system

This topic contains 27 replies, has 16 voices, and was last updated by  W-K 5 years, 9 months ago.

Viewing 13 posts - 16 through 28 (of 28 total)
  • Author
  • #186567


    I prefer things exactly how they are, it is at least fairly simple.

    But as a thought exercise, this is really cool :D



    Interesting idea. What I’m not so sure about is thinking about it as pure chance – HRT does not have a winning chance of 1:24. Also, it’s not as rewarding to finish first with this system as it is with the current system (currently the point difference between two 2nd places and a 1st and 3rd place is 4 points, with your system it would be one point)… on the other hand, it would reward a 2nd place more.
    I wonder how much thought out were all the system used – are they mathematical like in this thread or is it just something like “so the winner gets 25 points, the runner-up should get 20 points, the 3rd 15 points……. No, wait, the winner isn’t rewarded enough, I think, let’s give the runner-up 18 points instead.” etc. etc.

    @enigma: You’re not alone on that one… There’s a certain short Englishman who wants a system like that (and even has got some lobbying power).



    @maksutov, I think the trouble with going down that route is, it will start to be pretty arbitrary what is awarded and what isn’t.

    That does not count for qualifying, off course. But with Qualifying the reward is/should be giving you a better position for the race, not gaining points. If we want to reward excellent qualifying results, we could always instate a price for qualifyer of the year or something. Or even give bonus money for pole or for making Q2 and Q3.

    Thanks @alex for putting up this post to carry the discussion onwards. A very intresting effort with the logarithmic system. I think you are right in higlighting how the first thing to get it right-ish is the question of what to reward in what proportions.



    I like it. I’ve always felt that the points system is not correct – the current system is too much of a compromise between simplicity and accuracy. I’d be more than happy with VettelS’s suggestion of multiplying the points by 8.

    It might make sense to give a point for pole position & one for fastest lap – particularly to liven up the end of a boring race, but I wouldn’t want to see any more than that.



    @jajo, the current system seems to have started out with a more or less logical range of points. But it was then “corrected” to give a bigger difference between 1st and 2nd (suddenly making it only a small difference between 2nd and 3rd, to small imo) and the points between 5th and 10 are a bit dodgy as well.

    When I think about it @alex, I think it should be pretty clear that winning the race is more than getting on the podium regularly. And getting 2nd should be substantially better than 3rd. But all of those are certainly rated higher than getting in a good points paying position in 4t-6th etc.

    I think winning a race coud be as much as 2 x 2nd or 3 x 3rd then to reflect its special to end up there. Then it should be a big difference to finish in the top 10 or not, although its debatable weather finishing say 8th is much better than say getting into 10th or 11th repeatedly, but I think there should be a pretty big bonus for freak results like these.
    It does stick in our minds if a guy like Buemi suddenly finishes in the top 6 or something early in his career, far more then if he gets 2-3 times 10-12th. And it does show a driver like that can make things work for himselve.

    One thing that would make it far more complicated is, if you take in account not just finishing in a position repeatedly but consecutively (i.e. giving more points for a hattrick of wins in a row then for 3 wins during a season) to reflect how hard it is. But during a season the likelyness of finishing more races in 1st should statistically be taken per race (my opinion), as each race is an individual event.
    Indeed, one could say something for even giving LESS reward for more wins, once a driver/team gets more of them, as the likelyness of it is higher due to this driver/his car being better! (this would reflect the deminishing returns in fans view of a driver when winning more and more, each win adds less to how he is rated after the first couple of wins it just gets a statistic)
    This way we should reward say a HRT driver getting into 6-10 th almost as much as a Williams driver winning and both should get more than Alonso, Button, Hamilton or Vettel winning it several times in their current cars!
    That last one is not really to rate teams/cars in this case though. Its more going towards a system for rating individual drivers (didn’t we discuss one of those a couple of months ago)



    BasCB, I’ve added a version of your (first) idea to the spreadsheet – making first place be 2*2nd and 3*3rd (etc.).

    The points for the top positions are then 24, 12, 8, 6…

    This puts a big value on first place, but it does mean that (say) 15th and 16th are basically indistinguishable – so one disadvantage is it judges the teams on different standards, i.e. for the front-runners, winning is everything, but for the lower teams, consistency is everything.

    It also throws up some interesting changes. Hamilton is now 3rd, because of his race wins. (to go on from my comments last page, would he have received so much criticism under this points system? Perhaps people would praise his risk-taking. But he’s still behind Button).

    And Senna only comes 25th! This is because like I said, for the lower teams, this system works (arguably) against your intentions, as all the positions become a bit of a blur… not racing half the season becomes a huge black mark against him. Heidfeld also suffers.

    It certainly throws up some nice results, though. As for judging driver skill, maybe another week.



    I’m with a lot of people in thinking the points scoring system used now is pretty good. But I really enjoyed reading that and found it incredibly interesting to compare the lower half of the grid more accurately. I was surprised to see d’Ambrosio ahead of Senna and Glock, however!



    love this topic, i love messing with the point system!
    i’m also in favor of extra points for reaching q3, that would make everyone at least give it a try.
    i would also suggest a new points system based on the exponential function with base 1.2
    it would give the race winner 45 points. and i would also give the fastest lap a bonus of 3 marks.
    this way, a driver could get up to 58 points on a weekend.

    1st: 10 points –> 10th: 1 point

    24th: 1 point
    23rd: 2 points
    22nd: 3 points
    21st: 4 points
    20th: 5
    19th: 6
    18th: 7
    17th: 8
    16th: 10
    15th: 11
    14th: 12
    13th: 13
    12th: 14
    11th: 15
    10th: 17
    9th: 19
    8th: 21
    7th: 23
    6th: 26
    5th: 29
    4th: 32
    3rd: 36
    2nd: 40
    1st: 45

    DNFs still get points awarded due to official classification. DQF get no points.

    the WDC would have looked like this:
    1. vettel 958
    2. webber 796 (more consistent than button)
    3. button 759
    4. alonso 751
    5. hamilton 703
    6. massa 522
    7. rosberg 466
    8. schumacher 354
    9. sutil 334
    10. petrov 316
    11. di resta 303
    12. kobayashi 277
    13. alguersuari 262
    14. buemi 239
    15. barrichello 221
    16. perez 218
    17. heidfeld 186
    18. maldonado 174
    19. kovalainen 154
    20. trulli 130
    21. d’ambrosio 126
    22. glock 115
    23. senna 107
    24. liuzzi 82
    25. ricciardo 64
    26. karthikeyan 37
    27. de la rosa 14
    28. chandhok 5

    after doing the work, i didn’t really like the result… did you?



    Thanks for that updated table @alex!

    Interesting to see how the fiirst half of the field pretty much fits with my impression. Vettel (+RB7) clearly up front, then a relatively tight group of 4 and Massa, Rosberg, Schu … up to Di Resta close together in the midfield with all having some supprisingly good races, being pretty consitent and sometimes a fluke. Heidfeld might have been in there as well, if he had been able to do a full season.

    Then we have the group of drivers around 30 points. Behind that it really does get a bit murky with Jerome finishing better than Heikki, Jarno and Timo. Sure enough he did a solid job in a bad car, maybe we have all underestimated him then? Or it shows we are getting closer, but not there yet.

    Obviously the drivers not doing all of the season would have a horrible job getting anywhere near a good position here. To evaluate how good they were, one would have to take it in relation to the amount of races they could participate in.



    in my system above, this would be the team scores:

    1. rbr 1754
    2. mclaren 1462
    3. ferrari 1273
    4. mercedes 820
    5. force india 637
    6. renault 609
    7. sauber 509
    8. toro rosso 501
    9. williams 395
    10. lotus 284
    11. virgin 241
    12. hrt 188




    “But with Qualifying the reward is/should be giving you a better position for the race, not gaining points.”

    Of course these would be radical changes that may never be implemented but I still think it would work. The weight dependency of the points can still be split. For example, if a car fails or some trivial electronic mishap occurs during a race, all qualifying efforts go to waste. Rather than having qualifying “only” give you a better start position (which is a rather obvious component) why not split the overall total points to say 20% / 80%. Where 20% awarded for qualifying positions, 80% for race (or even 10%/90% would be sufficient). This would force drivers to do their best during both qualifying and race.

    It would certainly prevent drivers from purposely not setting a time or saving tyres or whathaveyou .. I still think qualifying points should be implemented despite of the initial award intended only for the starting position.



    @maksutov, I do not really feel its fine to award points for qualifying, but thats just my opinion.

    Thats why I would rather go for a monetary bonus for qualifying success or a seperate title for regularl expert qualifyers.



    I don’t like the idea of points for qualifying or fastest laps, because they can be used to gain points not deserved because you could set a car up to just do that.
    Also as this season has shown if a driver only needs one point to secure the title the temptation would be use one or both of these methods without caring what their race position was.

    On the subject of giving points all the way down to last place, what happens to the points when there is a multi-car pile up in turn one or two. e.g. USA 2006, seven cars out at turns 1 & 2, +1 for damage sustained in that incident, retired lap 3.

    On the points scoring system, mathematically to Keep It Simple, why not try a Fibanaci series. Where the points for the higher position equals the points awarded to the next two lower positions. From the last points position that would be 1, 2, 3, 5, 8……….
    (if I missed some ‘a’s its the k/brd)

Viewing 13 posts - 16 through 28 (of 28 total)

You must be logged in to reply to this topic.