Towards de title 2020. Warning: calculations!

[Schummy]

(Disclaimer: this post is not going to be liked by anybody. Those who are lovers of numbers will find a very thick and unnecessary wordy first part. Those who hate numbers will find an amount of them in the last part. Finally, those who hate long post are going to find... a long post.)



In this strange NeoWeirdness where we currently are, it could be confuse to gauge the level of likelihood each driver has in the title chase. Let's forget for a moment that if you don't drive a TotoCar you simply are not going to win the championship. Let's pretend that a number of drivers/cars have a chance, as if F1 were Indycar or MotoGP.


Has Leclerc some chance? Is Vettel totally out? What about Max, does he have a significative chance?


My point of view is through what last year I pompously called "One Race Gap". This means the transformation of the current gap from the driver to the top score driver (Lewis, of course) to an equivalent gap if only had one race to go. It is easier to evaluate the difficulty to overcome a point gap in one race than evaluate the difficulty to do it in a future number of races.


For example, Max (Verstappen, but "Max" is shorter LOL) has 52 points against 88 for LH (that is even shorter than writing "Lewis" ). That's a 36 point gap (math, math! ) with 9 races to go (COVID willing). It is not immediately clear if it is reasonably doable or perhaps very difficult. But if I say that it is equivalent to a 12 point gap with just one race to go, then our intuitive idea is clearer. Let's imagine we have one remaining race and Verstappen (I mean Max) is 12 points behind Lewis Hamilton (...), it looks rather bad for Marko's boy but it is feasible... as long as LH has a major incident (perhaps a puncture?).


So, I think we could agree is clearer to judge a 12 points gap with one race to go than a 36 points gap with 9 races to go. The 12 points gap would be the One Race Gap equivalent to the 36 points gap with 9 races to go.


Ok but, how the ORG is calculated? Sadly I cannot tell it because it is a copyrighted method of mine... LOL obviously this is a joke because the ORG is not going to make rich anybody and, above all, I described it last year in a thread, so I cannot hide it behind a copyright.


ORG is very simply computed dividing the actual gap by the square root of number of remaining races. So, Max's ORG is 36/sqrt(9) = 12. Is not it neat? (*_*)'



The following table contains the ORGs (ORG is ORG, not the initials of another very different dirty thing) for the main "contenders" (I should use here a ton of double quotes).

HAM  mer[88]
BOT  mer 30  10.0  10%
VER  rbr 36  12.0   8%
NOR  mcl 52  17.3   3%
LEC  fer 55  18.3   3%
ALB  rbr 62  20.7   2%
PER  poi 66  22.0   2%
STR  poi 68  22.7   1%
RIC  ren 68  22.7   1%
SAI  mcl 73  24.3   1%

First number is actual gap, second number is ORG, that is the equivalent gap with one race to go. Finally, third number is an estimation of the probability of overcoming that gap.


Nobody (except Hamilton himself... ) has "easy" this title. Think about Bottas having to get 10 points more than LH with just one race. Possible, yes, easy task, no.


Bottas and Verstappen (i.e. Max) are the only ones with a reasonable probability. The rest of them are in a dark and cold probabilistic purgatory (does really exist a cold purgatory??). And Vettel effectively is out, out of the title chase and out of Ferrari, of course.


I am happy to have been able to reach the end of this "long and winding" post, I thought I would never do it. If there is still someone reading this last part, please blink two times, so I can know... and check your brain with a professional. Probably you need to take hidroxicholoroquina as soon as possible.

[BertMk2]

It's an impressive effort in trying to inject some sort of hope that maybe there's still a title race. But realistically we all know that it's Hamiltons title already.

[chillibowl]

i would imagine that even if a couple of Middle East races were added to the end of the season, not much would change?

and what does 'poi' mean for the racing point drivers (no doubt something simple which i cant figure out lol).

[Greem]

Originally Posted by chillibowl

and what does 'poi' mean for the racing point drivers (no doubt something simple which i cant figure out lol).

emphasis should help

[chillibowl]

haha...i knew i shouldnt have asked!

[crmalcolm]

I appreciate the work, and it makes for some interesting reading.

One thing I am struggling with though, is how the 12 point ORG seems to give an impression that the title race is closer than the current 36 point.
It only takes a single retirement to make closing a 12 point gap relatively easy. Whereas a 36 point gap requires a consistent below-par result from Hamilton.

That said - please don't take it as a criticism, just an observation....

[Schummy]

Originally Posted by crmalcolm

I appreciate the work, and it makes for some interesting reading.

One thing I am struggling with though, is how the 12 point ORG seems to give an impression that the title race is closer than the current 36 point.
It only takes a single retirement to make closing a 12 point gap relatively easy. Whereas a 36 point gap requires a consistent below-par result from Hamilton.

That said - please don't take it as a criticism, just an observation....

A single retirement in ONE race is not very probable, how many times has Lewis retired from a race? But imagine a retirement among the 9 remaining races, it would not give MV the lead but it could get him a lot closer... with other 8 races (it doesn't matter if before or after that retirement) to try.

If fact, ORG calculation is due to the "variance property" regarding the sum of random quantities. Sometimes it looks counter-intuitive but it is real (with certain premises).

For example, a half of the determination of the final classification in a race is done in the first quarter of the race. So, with two short quarter races on Sunday we would have the same discriminatory power than with a whole race... but the two short races combined would use half of the laps of a normal race. Of course, the nature of the competition would be different, with more emphasis in short "sprints" than in middle endurance. Actually it would not be very different that the races with refueling and high performance tyres of some old seasons.

[GT6]

To be honest losing interest in this season very fast

[wnut]

Originally Posted by GT6

To be honest losing interest in this season very fast

Me too, Lulu has screwed up the complete whitewash already this season and we will just have to come back next year to see if he can get it right.

[Adam43]

Nice work.

Originally Posted by Schummy

HAM  mer[88]
BOT  mer 30  10.0  10%
VER  rbr 36  12.0   8%
NOR  mcl 52  17.3   3%
LEC  fer 55  18.3   3%
ALB  rbr 62  20.7   2%
PER  poi 66  22.0   2%
STR  poi 68  22.7   1%
RIC  ren 68  22.7   1%
SAI  mcl 73  24.3   1%

And the output feels reasonable.

[Schummy]

Originally Posted by BertMk2

It's an impressive effort in trying to inject some sort of hope that maybe there's still a title race. But realistically we all know that it's Hamiltons title already.

Let's not despair yet! We are in this so-called NeoReality so, LH could do a Perez...


Or perhaps Mercedeses (!?) could begin to fail apart like in that movie The Great Race.

[Schummy]

Originally Posted by Adam43

Nice work.

And the output feels reasonable.

Thanks. Yes, it shows a rather unexciting season , therefore it is realistic.

[Adam43]

And it is only made worse by the winner crossing the line with a puncture - mainly because it was the same winner.

[Schummy]

After the Silverstone-II race:

HAM  mer[107]
VER  rbr 30  10.6   9%
BOT  mer 34  12.0   8%
LEC  fer 62  21.9   2%
NOR  mcl 69  24.4   1%
ALB  rbr 71  25.1   1%

Really only Verstappen and Bottas are in the race against Hamilton. Even those two have it difficult because their probabilities of over scoring LH are less than 10%. Moreover, Hamilton has the "helpful help" of RedBull and Merc's clever strategists (suggestion: RB ones should be employed by Ferrari, particularly for Sebastian).

Leclerc has still a thin line of life but supposedly for a short time. Norris nominally has also a microchance. Albon... well, no, no more drivers with microchances.

Tasks for next GP:
- Verstappen and Bottas need to recover at least 2 points to not get a worse probability against LH. Leclerc needs 4 points to recover against Hamilton.
- To be worthy contenders (>10% probability) Verstappen needs to get back 4 points from LH. Bottas needs 8 points. The rest of drivers cannot be contenders even winning the race and Hamilton doing a Vettel.

In terms of Teams Championship, Ferraris is already out (<1%), while RedBull is still in the race, with a OneRaceGap of 23.7. Can RedBull gets 24 points over Mercedes in one race?It is hard to achieve, but it would be thinkable.

[Schummy]

Crude estimations of winning in next race:

HAM 44%
BOT 28%
VER 20%
LEC 6%
NOR 2%

They have been calculated by mean of a clever combination of races and qualifications results... well, not so clever at all.

According to these estimations, the following table shows the future number of wins in the remaining 8 races (Expected, Probable and Possible, "possible meaning in a 95% probability).

     %W  Expected Probable "Posible"  
HAM  44%   3.52    3 or 4    1 to 5 
BOT  28%   2.24    2 or 3    0 to 4
VER  20%   1.60    1 or 2    0 to 3 
LEC   6%   0.48      0       0 to 1 
NOR   2%   0.16      0         0

On another side, looking not to the wins but to the race classifications, Hamilton has beaten VER, VER, LEC and NOR 4-1 by now. But it means that each one of those has a 14% chance to end the season beating LH if they were in a similar level of performance from now on.

[Schummy]

After the Spanish GP and before Belgium's race, taking into account we expect to have 11 races to go.

          gap   ORG  Title
--------------------------
HAM  mer             86%  
VER  rbr  -37  11.2   7%  
BOT  mer  -43  13.0   5%  
LEC  fer  -87  26.2   0.4%
STR  mcl  -92  27.7   0.3%
ALB  rbr  -92  27.7   0.3%
--------------------------

After 6 races and with 11 races to go, LH has already approx 86% of chance to get the title. It speaks volumes about what great competition we have enjoying.

In order to maintain their slim probabilities, VER and BOT need to recover 2 points from HAM.

To improve and become somewhat like contenders, VER and BOT need to recover 6 and 12 points from Lewis.

If HAM wins and BOT gets 0 points, Valtteri will be practically out of the championship, with a probability for the title of approx 1%. It then would become a two horse "race" or rather a horse and a limping donkey race.

[Knowlesy]

You have to say, looking at the stats, Hamilton has a very real shot at the title here.

And I mean that.

[Schummy]

A shot with a bazooka, methinks

Update after the "surprising" GP Belgium (surprised for not having rain)

         gap   ORG  Prob  
HAM  mer[157]        92%  
VER  rbr  -47  14.9   4%  
BOT  mer  -50  15.8   4%  
ALB  rbr -109  34.5   0.1%
NOR  mcl -112  35.4   0.1 
LEC  fer -112  35.4   0.1%
STR  mcl -115  36.4   0.1%

I have included Albon et alter just for comedy purpose.

Ferrari's form is becoming to be tragic. Look at the position of the teams with Ferrari engines... I wonder if they have erroneously built a F2 engine. In particular, Ferrari team is looking very, very embarrassing. We all wonder what happened with last year's engine...

Poor Sainz... he is doing the infamous Alonso's path to hell. He is even driving for the same teams.

[TrapezeArtist]

Originally Posted by Schummy

Ferrari's form is becoming to be tragic. Look at the position of the teams with Ferrari engines... I wonder if they have erroneously built a F2 engine. In particular, Ferrari team is looking very, very embarrassing. We all wonder what happened with last year's engine...

Embarrassing for Ferrari, yes. It really makes me wonder what they were doing with last year's engine to give it such a margin of superiority over this year's. I don't subscribe to the idea that Formula 1 "needs" Ferrari to be competitive though. They're just another team. We have Red Bull, Renault, McLaren and Racing Point to try to put some pressure on Mercedes.

Enzo must be turning in his grave. He famously believed that motor racing was all about the engine, and the chassis was just there to stop the engine scraping on the ground.

[TrapezeArtist]

Originally Posted by Schummy

Poor Sainz... he is doing the infamous Alonso's path to hell. He is even driving for the same teams.

No doubt he's got his lawyers poring through the contract for an escape clause. And they will be telling him what a wonderful job they did of giving him a cast-iron contract.

I really do sympathise with Carlos, because I like him. However, even if he could get out of the Ferrari contract, there would be nowhere else to go.

[Schummy]

It is almost unnecessary to put the One-Race Gap to guess who has good probabilities to be Champion, but here we go:

          gap   ORG 
HAM  mer [205]      
BOT  mer  -44  16.6 
VER  rbr  -77  29.1 
NOR  mcl -140  52.9 
ALB  rbr -141  53.3 
RIC  ren -142  53.7

If there was just one race to go, Bottas would need 17 points over Hamilton to get the title. Pretty hard but doable if Lewis gets a problem (or a multiple FIA's penalty...). Simulations with current season's results shows approx a 2% probability of that happening.

Verstappen is out of any reasonable scenario for title, according to his ORG. However, he has a chance to overtake Bottas. Again, simulations throw an approx probability of achieving this of 10%.

Nobody else can likely get one of the first three final positions in the championship. From 4th downwards there is a "fierce" battle between them but, frankly, who is going to bother to remember who ended 4th or 6th?

The matrix of probabilities to overcome another driver is this:

HAM   --   98  100  100
BOT    2   --   90  100
VER    0   10   --  100
 NOR    0    0    0   --

It is a pretty boring matrix, is not it? I put Norris just to visualize that the rest of the field has nothing to do against the first three.

[Schummy]

On the other side, Constructors' Championship has a bit more interest because teams are actually concerned about their final positions because economical and prestige reasons.

So, I am including the matrix of probabilities of win against each other team.

mer   --  100  100  100  100  100  100  100
rbr    0   --  100  100  100  100  100  100
mcl    0    0   --   44   60   93   94  100
poi    0    0   56   --   64   94   96  100
ren    0    0   40   36   --   88   95  100
fer    0    0    7    6   12   --   76  100
tau    0    0    6    4    5   24   --  100
alf    0    0    0    0    0    0    0   --

I have not included Haas and Williams because they have so few points that is difficult to make any estimation with simulations. But one can imagine what positions they could get in the table...

Basically, the first two positions are done and the last three too (in any permutation between them). The real (and close) fight happens between the five middle class teams (Ferrari!!). In particular, McLaren, Pinky and Renault are completely undecided between them. Ferrari has chances but is and step behind, same with Alpha Tauri (Aldebaran?).