Thursday, January 20, 2011

Much Ado About Points - NASCAR's Proposal for Non-Change

UPDATED: The chart in this post has been updated to reflect NASCAR's official announcement (1st will get 47 minimum by way of 43 + 3 bonus + 1 for leading a lap). Oddly enough the actual new system makes even less change than the original rumored one.

A lot is being argued these days about NASCAR’s possible new points system clearly making yet another attempt to manufacture closer championships, and while I’ve seen ideas and theories I haven’t really seen some specific bits of data that many are overlooking when doing comparisons. This came to a head for me when USAToday’s Nate Ryan traded thoughts with me on Twitter about why NASCAR can’t have consistent close championships while IndyCar has gone down to the final race 7 out of the last 10 seasons.

That coupled with the many different writers out there throwing out their opinions on what NASCAR should do just pushed me to take a good hard look at statistics and historical factors amongst many series and their structures.

So like the crazy person I am, I decided to spend the last few days becoming the motorsports version of SETI, swimming through all the point structures, historical finishing data and variations I could get my hands on in an effort to find some kind of rhythm.

The first thing I already knew that I re-affirmed… The American LeMans Series has a crappy website with bad links and no useful information to this research…thank you Wikipedia. Pertaining to point systems however:

It’s All About the Differentials Baby.

A lot of people do notice the disparity from “rewarding leaders enough” in NASCAR vs. other series, but the problem goes much further than that (as you’ll see below). In order to accurately compare all the many point systems out there, you need a common denominator, rather than comparing 185 pts. for first place to 50 points for 1st etc. The common denominator across all series is first place getting the most points.

If you consider 1st place in a race as 100% of a possible goal, and any place below that as a smaller percentage of the total possible goal, then this is how the point differentials break out amongst some main series. (I excluded ALMS because they use 3 different systems depending on race length)

What you will notice first is that NASCAR’s current plan and proposed new plan are basically siblings, with a very slight variation. This means at the heart of it, they are changing nothing. The biggest issue of course is too much rewarding mediocrity. NASCAR’s proposed new system still awards 50% of the points possible for 22nd place. Tell me which you think is worthy of better standing, someone who finishes 22nd twice in two races, or someone who dominates and wins a race, but then succumbs to an engine failure in the 2nd race.

For comparison’s sake, in IndyCar, you must finish 7th or better to receive at least 50% of first place's points, in Formula One you need 4th place for 48% or better. This is clearly the biggest issue for NASCAR; Jimmy Johnson doesn’t run away because he wins the majority of races, he runs away by stacking up points in finishing 6th-22nd place: 2010 (9 times), 2009 (13), 2008 (15), 2007 (11), 2006 (18).

If they gave fewer points for doing sub-par finishing, Jimmie may have still won the championships, but he definitely would not have pulled away with so many mid-level finishes.

The second issue for NASCAR (and also IndyCar/Grand-Am) is that they don’t zero out at any level. NASCAR’s new system would come as close as possible to accomplishing this with 1 point for last, but at the current moment, you could do pace laps, pull straight into the pits, go home and eat a sandwich and fall asleep before the race is over… and you’d still get 20% of the possible points as a guy out there racing to the edge for every lap.

Interestingly, this was the first
picture that came up when
I typed 'nascar' into Google
I’d also think this can likely be a cause of the large amount of NASCAR wrecks; because crashing out of a race really doesn’t hurt you all that much in the long run. Grand-Am and IndyCar for all intent and purposes end their point distribution at 13th and 26th place; the remaining breakouts are simply for the Indy 500 and 24 Hours of Daytona with the larger fields. As a result, their systems also award 20% and a whopping 37% in Grand-Am for going home taking a bath and watching the end of the race from the couch.

The problem with not zeroing out points is this; you tell me who had the worst season:

Driver A) 6 starts, 3 top 10 finishes; qualified 2nd for a race, led 15 laps at the Indy 500.

Driver B) 5 starts, finished 14th or better in 4 of them including 18 laps led and a 6th place finish.

Driver C) Qualified at or very near the bottom in all but 1 race of 16, finishing last of all running cars for every race.

It doesn’t take a genius to know both Paul Tracy and Mike Conway had better years than Milka Duno in 2010, yet she finished 74 and 93 points ahead of them simply because she showed up to more races, even though she sucked horribly at all of them.

Zeroing out is the one thing Formula One most definitely gets right; though I’m not advocating IndyCar, NASCAR or Grand-Am need to zero out the points at the halfway point of the field like F1. But doing it at 75% of the field would make a lot of sense. That translates to needing 32nd or better in NASCAR to get points, IndyCar=20th or better, Grand-Am=10th or better. And obviously the 75% can vary per the amount of cars qualified. 

Other bits of mediocrity that NASCAR can stop giving points for is simply leading any lap. Leading the most laps is an accomplishment, staying out during a caution to take the lead for a lap, only to later get back on the lead lap due to a “lucky dog” is not something to be rewarded; but that’s been a result of NASCAR trying to allow people to all keep up with Jimmie Johnson. The only problem is Chad Knaus and Jimmie are smart guys, so they make sure to get those points too.

One thing NASCAR has less immediate control of, and possibly the greater reason for IndyCar’s consistency is track diversity.

Hendrick motorsports isn’t just dominating because they have the answer to so many questions, it’s because the Sprint Cup series doesn’t require them to take that many different tests. Most tracks are close enough in characteristics that the team has less setups to nail, and many tracks are run multiple times in the same season furthering the issue. Nail one base setup and you’ve got 60-70% of the tracks. That’s where IndyCar has the goldmine:

Short flat ovals, short banked oval, large banked ovals, flat Indy, bumpy street courses, skinny street courses, airport courses, hilly terrain courses; racing in the rain as well.

There isn’t a setup that will work on any two of those, nor a driver who is the best at them all; and it’s partially why IndyCar’s contenders stay closer.

IndyCar Season Most oval pts. Most road/street pts.
2010 Dario Franchitti - 268 Will Power - 412
2009 Scott Dixon - 406 Dario Franchitti - 270
2008 Scott Dixon - 482 Helio Castroneves - 233
2007 Dario Franchitti - 463 Scott Dixon - 217
2006 Dan Wheldon - 415 Scott Dixon - 125
2005 Dan Wheldon - 536 Tony Kanaan - 133
Since IndyCar brought road/street courses back, no driver has ever been the point’s leader in both road and oval courses. This means, its kept drivers honest and allowed others to catch up to the leader by doing more of their forte…  If you added more road courses to Sprint Cup, I guarantee you Juan Montoya would immediately start challenging Johnson.

Lastly, can everyone stop slapping one series’ point system onto another? Many writers LOVE to do this, but different rewards elicit different actions. If shots within the free-throw area were only worth 1-point in the NBA we’d see a lot less dunks, if they shortened marathons to 16,000 meters runners wouldn't save so much energy, and if NASCAR had Formula One’s point system, they’d stop settling for Top 20 finishes; so let’s stop assuming similar results with different setups.


The Speedgeek said...

OhmydearsweetJesus. This is possibly my favorite blog post that I've read in the three years I've been reading racing blogs. Have you been living in my brain or something for the last two weeks, Wedge?

Yes, yes and yes. The biggest key to points systems, in my mind, is rewarding winners proportionately over 2nd, 5th, 10th, 20th and last places. IndyCar and F1 do a decent job of this, generally (excepting IndyCar's awarding the same number of too many points for a big chunk of the bottom of the field), and it's something that NASCAR and GrandAm (huh, is that related?) do not. Midfield results in NASCAR and GrandAm award far too many points, and therefore do not encourage drivers to, you know, try to win races.

The next biggest thing, in my opinion, is to allow drivers to be in contention through differing means, so long as there isn't one clear dominant champion. You got into this with your Tracy/Conweezy/Milka bit a little, except it works on the other end of the grid as well. If a guy wins a bundle of races, but blows up just as much (Kyle Busch-style), I like when a guy can score just as many points by winning a little, scoring a bunch of top-5s and having basically no bad days. The IndyCar system does this well, too, but like I said, actually does too much to reward 18th and back.

Last, 100% right about why NASCAR has the same 2-3 teams and the same 4-5 drivers dominating 5-6 year chunks of the sport: they visit about 65 mile-and-a-half ovals every year, about 21 2-mile ovals and like 5 weirdo tracks (Indy, Pocono, the road courses, Dover, maybe Phoenix, Bristol, Martinsville). OK, I'm overstating, but if you can rack up setups that allow you to finish top-5 at every 2- and 1.5-mile oval, you're going to wind up top-3 in the Cup standings. Oh, and as long as you can plonk around in 5th through 10th and keep your nose clean during the Chase, maybe win a race or two there, you're about guaranteed to win the Cup.

OK, really, that's it. Until I come up with some more comments. Really, really great job, Wedge.

Anonymous said...

This is an awesome post! Well done on the brilliant work here!

The graph is excellent as it shows the relative points paying position differences perfectly - and long confirms suspicions I have had about various series.

F1 points have always been elitist and about winning while NASCAR has always been about consistency and turning up - which is why cars seemingly half the race down will continue. I'm not surprised they don't want to change anything because as you say, it allows drivers with big names and not as much speed the chance to make the chase when they shouldn't deserve it, in fairness.

Also Indycar seems like a fairer system which doesn't surprise me too much. I think other series would do well to adopt their model.

Final comment: you mentioned that people do comparisons on series based on different points systems. This is done a lot in F1 as the points systems have changed a few times over the years, but oddly the results end up being very similar. The major difference is when you change from a 'dropped scores' system to an 'every race counts' system, but otherwise most F1 seasons would have had the same champion anyway ;)


Pat W said...

This is brilliant stuff. IndyCar really gets the right points balance, though it should stop giving out prizes just for turning up. That's the part I most like about the F1 system.

Having a completely linear points progression seems very odd to me. I like that most series reward the top four or five places with increasingly more points over the next lowest position.

Good work.

CurlingRacer said...

Great post! I agree with you on the IndyCar side. Stop giving points out after #20. That works now. As the series gets stronger, move it up to #15.

Paulo said...

Just a minor correction: as 4th place in F1 nets you 12 points, that's not quite 50% of 1st place points (25). In effect, you need to be 3rd or better to get more than half - and that is what I really like about the F1 points system. You need to work hard just to score points, never mind stay in front of the leader.

As for this points proposal, I see getting tweaked again and again and again. Not only because the NASCAR bosses have a big tendency to tweak (see: Chase rules), but because this system will turn out to be very flawed very quickly.

Allen Wedge said...

You are exactly correct Paulo. 4th is 48% rounded, but I figured for the sake of argument 48% was closer to 50% than 60% is, which is what 3rd gets, so that why I went with it as the measuring stick, just need better wording.

We totally agree, and the argument still holds, award points for doing something worthy of merit, not just showing up.