Formula 1 will use new television graphics power by artificial intelligence to show data such as the ‘overtake probability’ during races next season.
Speaking at the Amazon Web Services Re:Invent 2018 conference, F1’s managing director of motorsport Ross Brawn revealed three new graphics developed for the 2019 season making use of AWS’s machine learning technology.One of the three will show an ‘overtaking probability’ when two drivers are fighting for position. “Wheel-to-wheel racing is the essence and critical aspect of the sport,” said Brawn.
“Now with machine learning and using live data [and] historical data, we can make predictions about what’s going to happen. So the graphic shows what we expect is going to happen in this event.
“What’s great about this is that the teams don’t have all this data. We as Formula 1 know the data from both cars and we can make this comparison. That’s never been done before.”
F1 appointed AWS as its ‘official cloud and machine learning provider’ in June. Brawn showed another example of its technology being used to give insight into a driver whose tyres have overheated. He said the technology will allow fans to see data which has previously only been available to teams.
“We can look at the history of the tyres and how they’ve worked and where he is in the race and machine learning can help us apply a proper analysis of the situation,” he explained.
“We can bring that information to the fans and help them understand whether the guy’s in trouble or whether he can manage the situation. These are insights the teams always had. We’re going to bring them out to the fans and show them what’s happened.”
The technology will also be used to give deeper real-time insight into race strategies. “We’re going to take all the data and give the fans an insight into why they stopped and when they stopped. Did the team and driver make the right call? The info box [will] give the fans that insight we can build using machine learning.”
Advert | Become a RaceFans supporter and
The same technology will be used to inform future F1 decisions, such as track design, race format and grid layout, Brawn added.
“Further down the road what’s really exciting is we’re going to investigate the influence of the tracks and the racing formats on the quality of the racing.
“Can we create tracks that achieve better racing and better overtaking? Can we build models to allow us to do that? Can we change the format of racing to make it more exciting and less predictable?
“For example, what happens if we change the formation of the starting grid, so instead of being spread out it’s bunched up? We believe that using machine learning, AWS is enabling us to do these things.”
Video: Ross Brawn at Amazon Web Services Re:Invent 2018
Brawn had more to say about F1’s work with AWS and showed examples of how the new 2019 television graphics will look:
Advert | Become a RaceFans supporter and
2019 F1 season
- Crying in the Melbourne car park at 2019 grand prix was my career low – Ocon
- McLaren Racing reports reduced £71 million loss in 2019
- Kvyat: Hockenheim podium last year was “my biggest achievement” so far
- How the FIA’s new encrypted fuel flow meter targets Ferrari’s suspected ‘aliasing’ trick
- “He smashed my office door”: 23 must-see moments from ‘Drive to Survive’ season two
Sham (@sham)
30th November 2018, 7:10
Seriously? With F1 not being predictable enough, more predictions forced on us.
Personally, I want to see the gaps and be excited because we don’t know what will happen.
The whole problem is that most of the time we know what will happen.
Mashiat (@mashiat)
30th November 2018, 7:52
@sham Do you think this is going to change that? This is just a model that makes a prediction, and will basically give us information that we as viewers can gauge for ourselves. Just because the model says that someone is likely to overtake the car in front, doesn’t mean it’s going to happen, and vice versa. Also, we aren’t fully aware of the limitations of this model. For example, is it going to take into account the difference in power unit horsepower, and component wear between two cars fighting? As well as how much electrical energy and fuel is available to each driver? Or the level of risk a driver is willing to take? Context plays a huge role in determining an overtake, and this model is likely based on theory and past statistics. This doesn’t make F1 “more predictable”.
Mark Zastrow (@markzastrow)
30th November 2018, 8:23
@mashiat I agree completely. Knowing the batting average of a cricketer or baseball player doesn’t detract from the suspense of what will happen, rather it provides context that adds to the drama before it unfolds and helps us put the outcome into perspective once it has happened.
I too will be curious to know the limitations of the model and what exactly it has been trained on. It seems to me that the simplest way to go about this is to base it purely on the timing and spatial gaps and velocities of the cars, without taking into account component life or driver style. I think the point is not to try to predict whether a particular overtake on screen between a particular pair of cars and drivers will or will not happen, but whether a generic car or driver would do so, which would allow us to better judge how all the other variables may have contributed to the outcome.
bosyber (@bosyber)
30th November 2018, 9:14
Well put @mashiat & @markzastrow, I think it is an interesting attempt to give the viewer more context; it won’t influence the actual predictability of the sport (at worst, our naieve lack of insight in what the teams already know is stripped away – I can’t think of that as bad), just our view of it.
I do hope that it isn’t too intruisive so it can easily be ignored by those who are not interested/just want to watch the on-track fight for themselves with minimum clutter to block view of that.
StefMeister (@stefmeister)
30th November 2018, 13:38
@bosyber
During the presentation they showed previews of the graphics there planning.
https://i.redd.it/zsr1nujw88121.png
https://i.redd.it/kpjxiiiw88121.png
https://i.redd.it/5xaf0ghw88121.png
bosyber (@bosyber)
30th November 2018, 14:52
Yep @stefmeister, and I have to say, I can choose to ignore those. Didn’t realise before that the ‘cornering speed’ thingies I saw in recent races are also part of this effort until watching the video, but makes sense.
Aldoid
30th November 2018, 14:56
It seemed like a ridiculous idea at first, but after seeing those screen grabs I realize I’ll barely even notice because I didn’t really pay much attention to any of those. I still miss the simple graphic with the g-meter…
BasCB (@bascb)
30th November 2018, 17:01
Yeah, i too agree that to me it will be an extra thing to look forward to. And to compare how accurate their predictions turn out to be (i.e. Hamilton coming up on say Stroll in his FI with older tyres and DRS would be almost 100%, Vettel having a go at Verstappen at a corner in say Suzuka would be closer to say 20% chance.)
In a sport where so much is managed from the engineering side, I love the idea of making that kind of analytical perspective visible for viewers – we can only hope it makes good riddance of the commenator getting all excited for a DRS pass on new tyres, or hyping up the chances of a Ricciardo trying to chase down Gasly for 6th in the last 6 laps.
eljueta (@eljueta)
30th November 2018, 7:15
It’s great because now we can know when to nap and when to stay awake.
ColdFly (@)
30th November 2018, 8:36
Great idea. @eljueta
I’ll develop an app linking the AWS feed to the alarm on my mobile phone ;)
bosyber (@bosyber)
30th November 2018, 9:24
To help those that are looking for such efficiency, the part with Brawn starts at about 1:38:52 in the above video @coldfly @eljueta
Ronald (@mosquito)
30th November 2018, 7:16
I hope the data is going to be provided by someone else than the guys that give us the probability of rain….
ColdFly (@)
30th November 2018, 8:37
It’s not the guys, but the dice they use ;)
Phylyp (@phylyp)
30th November 2018, 8:57
@coldfly – LOL :)
Joao (@johnmilk)
30th November 2018, 14:49
Monaco GP 2019:
*Graphics appear on the screen*
Probability of overtake: Error, cannot divide by zero
Nugo (@nugolo)
30th November 2018, 19:13
Great :D I expect a lot of NAs
georgeboole (@)
30th November 2018, 20:47
If its the same guys showing a lastname driver then i expect some Venn diagrams too
JohnH (@johnrkh)
30th November 2018, 7:22
If it’s anything like the download time indicator on windows it will be as useful as an ashtray on a motorbike.
Todfod (@todfod)
30th November 2018, 9:05
LMAO
Gwan
30th November 2018, 7:26
There are do many cool ways to use data, why add something to make (watching) the racing less exciting and more predictable?
spoutnik (@spoutnik)
1st December 2018, 10:20
Launch an r-factor A.I. race in parallel so that you can choose to watch the most exciting race of the two.
WimB
30th November 2018, 7:31
Trust the powers that be to answer a question noone was asking. [Sigh]
Pedro Andrade
30th November 2018, 8:17
My thoughts exactly, I have never seen anyone requesting anything similar. And it is a dumb idea.
Fer no.65 (@fer-no65)
30th November 2018, 13:45
Incredible isn’t it? I suspect they are going to show overly optimistic values too. There will always be a possibility of overtaking, they won’t say it’s 0% because thats boring and unexciting which is something they want to avoid.
I don’t see why this has to be shown on the telly. It’s like they dumb down the sport because it’s not relevant information.
Also, I suspect this has to do with the new gambling scheme they’ll be part of next year.
BasCB (@bascb)
30th November 2018, 19:50
– It sure is. Seems that while it took a while, they are exactly giving me what I wished for in the interview about what I would love to see more of when visiting the Austrian GP last year @fer-no65 – yes, they are giving us more insight in what is happening behind te scenes!
Mashiat (@mashiat)
30th November 2018, 7:45
F1’s equivalent of Expected Goals (xG) in football. I wonder if people are going to start analyzing who is “overperforming” and “underperforming” in terms of Expected Overtakes (xO) in a season.
Nitzo (@webtel)
30th November 2018, 7:47
Cant wait to see this fail.
My prediction: Overtaking probability is going to be consistently less than 33% (will be shown on TV :-P) and eventually they will do away with it come Monaco !!
Andy (@andykb)
30th November 2018, 7:48
I’ve always held Brawn in high regard but I think that he’s starting to lose the plot. Liberty have already added too much needless tat to the on screen graphics, this is definitely not needed and I can’t imagine why anyone thinks that it’s a good idea.
GongTong (@gongtong)
30th November 2018, 13:33
Are y’all still watching on 15″ tellies or something? I don’t have any problem with the extra graphics. Even the ones I don’t find useful. I have a foreign missus though so I’ve adapted to subs over everything…
GongTong (@gongtong)
30th November 2018, 13:35
Edit: it’s not because of her English. Her hearing is bad on one side.
Formal apology issued. With scepticism.
Yaru
30th November 2018, 15:19
I’m not sure about the upcoming overtaking ones but the current graphics are great.
matt90
30th November 2018, 17:00
I particularly like when somebody is on the edge of the drop zone and is identified as ‘at risk’ in a large graphic, while being the only driver with no time displayed, thus giving no reference point for what time those in the drop zone actually have to achieve.
Mark Zastrow (@markzastrow)
30th November 2018, 7:51
I love this kind of stuff.
What I didn’t appreciate before watching the actual demos of the overtake probability (1:46:55 in the video) is that the onscreen figures update in real time, synced to the video feed.
So as one car closes in on another we can see the probability of an overtake going up. This would presumably allow us to put a number on, for instance, Ricciardo’s late-braking prowess, based on how much lower-percentage his moves tend to be when he enters the braking zone.
And the same during pit stops—it runs continuously during the entire stop, so we can see where the cars gain or lose time over the course of the exchange, and whether it’s the stop itself or the driver that made the difference. For instance, Mark Webber said Vettel was always terrible at pit entry—I wonder if we’ll be able to see that onscreen in the data next year.
I can see a case for limiting the use of these graphics to replays, in order to preserve suspense. But I think it’s fantastic that we’ll now have access to this kind of data.
BasCB (@bascb)
30th November 2018, 17:11
Now that you write it down like that @markzastrow, I can already see the betting ads – With live updated rates and even a “cash out now” option when you gamble a bit and the driver actually looks like they might make it when it looked unlikely to go ahead before that – this thought came up combining it with the annoucement earlier this year about FOM working with betting companies to “enhance” things.
Actually guys @fer-no65, maybe these new partners were the ones who DID ask for this kind of graphics. – Although even if not, I am sure Liberty have the foresight to see how this kind of live updated things can attract more attention from an audience too.
I know I always loved the way how at Indycar you got far more predictions and stats on screen. Just thing of what @coldfly mentioned above about linking it with his alarm. what if you could suscribe to a streaming channel giving you “likely goals, setballs, overtakes for your favourite sports” and automatically switches to footage of something like that live happening. Just imagine the possibilities for the future.
MCBosch
30th November 2018, 7:57
Less graphics on screen please. And reroll the F1 live timing app.
Yaru
30th November 2018, 15:20
Current graphics are great, hope they continue this style.
Mashiat (@mashiat)
30th November 2018, 7:57
Something that I would actually like to see is a “live gap” between two drivers when one of them has made a pitstop. For example, if Driver A is two seconds ahead of Driver B, and Driver B goes for the undercut and pits. I would like to see a theoretical gap between the two based on an estimated pitstop time, just to see how much Driver B is gaining on Driver A on the out-lap (or subsequent laps). This would have been useful in China this year for example, where I think a lot of people were surprised Bottas managed to jump Vettel in the pits, and this would allow us to follow exactly how he did it.
Mark Zastrow (@markzastrow)
30th November 2018, 8:09
Yes, this is exactly what they are doing at 1:47:33 in the video.
Mark Zastrow (@markzastrow)
30th November 2018, 8:11
I keep forgetting I can add links in these comments. :)
https://youtu.be/ZOIkOnW640A?t=6453
Mashiat (@mashiat)
30th November 2018, 8:33
@markzastrow Thank you. I really haven’t been able to set aside 165 minutes to watch the whole video. Since the undercut is such a prominent part of F1 now, I feel like any additional information they can provide would be invaluable.
Mark Zastrow (@markzastrow)
30th November 2018, 8:37
Agreed. And FWIW I just skipped ahead to Ross Brawn’s segment—thankfully, it’s only 10 minutes or so! ;-)
bosyber (@bosyber)
30th November 2018, 9:25
@mashiat, the Brawn bit start at about 1h39m into the video.
StefMeister (@stefmeister)
30th November 2018, 13:45
@mashiat
Which is exactly why I don’t want to see this data, I want to be surprised because that to me is what makes the sport exciting.
This is the sort of data I really don’t want to see because if we know where somebody is going to end up after a pit stop then it takes away a big part of the tension/excitement you get by not knowing if a particular strategy has worked or not until the car rejoins the track.
Cheryl
30th November 2018, 8:06
Less graphics and more racing please.
Yaru
30th November 2018, 15:21
You can go watch it live. :)
Meanwhile I’ll be happy to sit on my TV and watch the awesome graphics overlays.
John H (@john-h)
30th November 2018, 8:32
So shallow. Allow the viewer to work this out for him/herself. That’s part of the fun. This just takes something away.
Speaking as someone that works with AI in industry, this sounds like using technology for the sake of it, the worst kind of application.
Todfod (@todfod)
30th November 2018, 9:13
@john-h
Agree completely. When you look at a driver attempting an alternate tyre/pit stop strategy and falling down the order, it’s interesting to see the gaps, his current pace and make your own estimation. The mental calculations followed by the uncertainty of whether it will work or not, adds to the excitement of the race.
Can you imagine how exciting the Chinese GP would have been this year if Ricciardo pitted for new softs and as soon as he exited the pits, there was a graphic saying 80% chance of win?!? It would completely ruin the anticipation and excitement of whether he could pull it off or not.
I thought they were trying to make F1 less predictable and more exciting. Then they come up with this ridiculously daft idea of simulating the order of finishing in races and broadcasting it to users. This idea is as bad as Bernie’s sprinkler idea. I thought this kind of nonsense was supposed to stop once Liberty takes over.
John H (@john-h)
30th November 2018, 9:32
Indeed @todfod. Suddenly those sprinklers don’t sound too bad :)
John H (@john-h)
30th November 2018, 9:36
@todfod By the way, sorry to go off topic but just to say I thought of you the other day when Alonso retired from F1. I still remember having some quality debates with you during the 2008 season (maybe 2007, not sure) and beyond (me being a Hamilton supporter). Nice to see how the two drivers respected each other through the past 10 years, shame we never got a repeat with Alonso in a Ferrari for 2017/18, would have been fun.
Todfod (@todfod)
30th November 2018, 9:59
@john-h
Yeah.. It was probably in 2008. That’s when I discovered this site.
Man.. A real shame we didn’t get to see them really go toe to toe in championship winning cars after that 2007 season. I think both Hamilton and Alonso were the fiercest of rivals, but still respected each other’s abilities..and surprisingly I think both Hamilton and Alonsos fans feel the same way.
Was a great moment to see them do a formation lap in Abu dhabi follows by their interviews. Great moment honestly.
bosyber (@bosyber)
30th November 2018, 9:41
I see what you are saying, but, well, a lot of this is info that the teams have had for ages, so in effect, you are saying you want to remain happy to be ignorant, aren’t you @todfod? All it does (when it works) is show us more accurately what the team thinking is.
Because I do sometimes like to just watch what is happening without any distraction, I do think it needs to remain something that I can blend out when watching And to be sure, I have mostly ignored those ‘cornering speed comparisons’ they have had in the last few races – didn’t know that was from this data usage before. But, I love that it is there, since it only allows me to better know why teams do things; I do wonder how long it will be until teams start taking that stuff from the screen and put it into a feedback loop though!
Todfod (@todfod)
30th November 2018, 9:54
@bosyber
They definitely have this data. No doubt. I just rather this kind of data not be shared with users. As you said, I rather stay ignorant and not know.
It’s like starting a movie and with te start credits they tell you how the movie ends.
I just do not see the point in this exercise. If F1 implements this it will be a prime example of how technology can be used to ruin a sport.
Robbie (@robbie)
30th November 2018, 14:35
@todfod I didn’t get the impression they were going to predict winners during the race. I don’t think Brawn has implied that has he? And of course I know you may be saying this tongue in cheek, but sprinklers would actually dramatically affect the races and imho the integrity of the sport, whereas what Brawn is talking about would not change anything about the actual action on the track.
Todfod (@todfod)
30th November 2018, 14:42
@robbie
The way I see it, if you’re giving probability of overtake you’re giving the viewer a rough idea of where he’ll finish. We just don’t need that information. We rather see a driver chasing another car and get closer and closer instead of a bar graph with probability on it. It just doesn’t add any value to racing whatsoever.
While this might not affect the integrity of the sport… It definitely hampers the integrity of watching a race.
Robbie (@robbie)
1st December 2018, 0:59
@todfod Fair comment. Haven’t entirely wrapped my head around what it will be like, but I sure expect from 2021 that there won’t be the predictability of drs, and the action on the track will be more enthralling, such that probabilities are something closer to 50% ie. could go either way so the graph tells us to watch it play out and see what happens anyway. Or, if the odds lean heavily one way, let’s see this time be the start of the reversal of that trend by that driver for whom the odds are the least for success. If we have a clearer picture of the condition of the tire that Brundle has been commenting on, on that particular driver’s car, let’s watch and see if the driver pits or adjusts and makes it last longer. The commentators already hint at certain likelihoods for us already, and we get clues from some radio comm, and we figure stuff out on our own as well.
I see what you mean though about overkill of info, but I think I have to experience it to see if I find it harmful to the integrity of my experience on F1 weekend. If I find it no real distraction, and this kind of supercomputer technology enthralls more young people to watch, then that would be fine by me. Since I’m struggling to envision all that it might and can be, I’ll reserve judgement.
I don’t know what Brawn means when he says change the formation of the starting grid and make it less spread out and more bunched up, nor how AWS can help with that. Is he literally saying input into the system the lines closer together and have AWS figure out the odds of cars starting more closely to each other causing more chaos, or not? More excitement?
Panagiotis Papatheodorou (@panagiotism-papatheodorou)
30th November 2018, 9:47
Spot on mate.
hahostolze (@hahostolze)
30th November 2018, 8:33
I really, really hope they show this graphic with the probability to the drivers in the car. Imagine letting Danny Ric know he has a 30% chance? Epic. Not every driver, mind…
Also, how do you calculate probability when it is dependent on the quality of the driver?
Mashiat (@mashiat)
30th November 2018, 8:40
@hahostolze As I mentioned in my comment above, it is somewhat similar in principle to Expected Goals in football. It showcases the probability of scoring for the average footballer. People who are able to overperform xG can be classified as clinical finishers, and it is the same in F1. If Ricciardo is able to keep overtaking when the model predicts a not-very-high chance of an overtake, it will be a testament to his ability to outperform what you would expect most drivers to be able to do. Equally, there may be certain drivers like Bottas who perhaps slightly “underperformed”, as they sometimes don’t make the pass when they would be expected to (Bahrain 2018), and rarely defy the odds to make an unlikely overtake.
hahostolze (@hahostolze)
30th November 2018, 8:45
@mashiat I noticed that, and I know that football xG models are now so complex they can indeed take into account how clinical a certain player is, but I struggle to see how that works in F1 with a much wider range of variables. Will a driver be more ‘clinical’ at overtaking if he has a few engine penalties in, say, a Ferrari, and DRS-slices his way through the field, than a driver who barely has to overtake and every overtake he does is laborious because it’s always, say, a Merc he has to overtake. I don’t thinkt his will work.
Mark Zastrow (@markzastrow)
30th November 2018, 8:58
@hahostolze I think that, as @mashiat says, the point of the model is that it represents a generic baseline so we can see which drivers are actually pulling off lower-percentages moves. If you trained the model to adjust Ricciardo’s chances up, then the stat wouldn’t mean as much. The word “prediction” is really a bit of a misnomer because the point is not really to tell us what will happen, but what has happened in the past so we can then make sense of how unlikely what we just saw really was.
Although, if they could train the model on driver-specific datasets so we could compare a particular driver’s odds against the generic model—or their individual rivals—that would be very, very interesting, too!
bosyber (@bosyber)
30th November 2018, 9:43
With the amount of data they have @markzastrow, I wouldn’t be surprised if they would attempt the latter (I’d want both predictions then: one showing ‘in general X would be the likelihood of it happening; for this driver it is Y).
jmlabareda
30th November 2018, 9:46
I have seen the algorithm… it’s very sophisticated
if (DRS works in this circuit)
probability of overtake=100%
else
probability of overtake=0%
rpiian (@rpiian)
30th November 2018, 11:43
Hah! Good one!
Todfod (@todfod)
30th November 2018, 12:52
It’s a little more complicated than that.
If the car in front = Ferrari or Mercedes, then probability of overtake with DRS = 50%.
If car in front = Renault or Honda, then probability of overtake = 100%
Jere (@jerejj)
30th November 2018, 9:48
Interesting albeit a bit unnecessary.
W-K (@w-k)
30th November 2018, 11:27
If they are going to use AI then what we need is an AI Commentator, that commentates and also knows when to shut up and let us listen to the cars and the squealing tyres.
Bernie
30th November 2018, 11:56
If they can let us know before the race, maybe we can make a better decision as to whether its worth watching in the first place!
Ben (@scuderia29)
30th November 2018, 12:06
So now the graphics will tell us the chance of something happening before it happens, talk about killing the excitement. There are so many problems in formula 1 at the minute, how much time and money are they investing on solving problems that don’t exist with solutions nobody wants?
Depailler
30th November 2018, 12:12
The problem is not enough overtaking manoeuvres happen because of various aerodynamic, financial and track issues already known by all.
This gimmick does not address these issues in any way.
Imran Aowte (@tifosiimran)
30th November 2018, 12:18
What’s next? Spin probability for Seb? Engine blowing up probability for Renault and Honda cars?
Tristan (@skipgamer)
30th November 2018, 12:37
One of Americas biggest tech company sees some value of in-broadcast advertising in F1. This is huge and can be seen as a good move in an important direction.
People are overreacting about the whole onscreen graphic thing as if it’s going to make a huge difference. Does anyone think Rolex are actually timing anything?
socksolid (@socksolid)
30th November 2018, 12:39
I don’t think this is going to work. People have very poor understanding of probability and what it means. If we have 80% probability the pass might still not happen and people will interpret that as misinformation and failure of the system. An 80% rain chance for example means that in the area somewhere during the day there is 80% chance of rain. But in f1 when we see that number it is interpreted as 80% chance to rain on track during the televised event. And I think f1 does that misrepresentation on purpose because higher chance of rain means maybe more people will watch it.
When you combine the poor understanding of statistics and probability and f1’s desire to misrepresent the numbers to make it look like it is better than it is… You may get very anticlimatic results when the excitement factor fails constantly because it is overestimated and misrepresented consistently.
rpaco
30th November 2018, 12:53
I cannot see that working as they cannot even stop the subtitles blocking off the top of the leader board. I realise that they just do not care, but it is such a simple thing to adjust that I hope they all go deaf an have to use subtitles then selves, then finally we may be able to see the top two.
Just three chars to the right please!
Adam (@rocketpanda)
30th November 2018, 13:06
Bit of a silly idea really. An answer to a problem nobody had.
Kinda see it being abused too, as in we’ll get stats of which drivers are better at overtaking than others and stuff like that, more reason to bash one driver over another, more pressure for the drivers themselves. Seems unneccessary scrutiny, to be honest.
As others have said too it has the danger of making a sport that already suffers with predictability becoming even more predictable, and being predictable is pretty much the death of sport. Who wants to see a battle when you already know the outcome? It’s not a lot of fun. Show, don’t tell?
Francorchamps (@francorchamps17)
30th November 2018, 13:08
I don’t see the point of this statistic. An overtake either happens or doesn’t happen, so the probability will always be 50%. “big data”, “machine learning” or “computer science” won’t change this fact.
krxx
30th November 2018, 14:58
Without passing any judgement on introducing these gimmicks, just bc an event has only 2 possible outcomes, doesn’t make their probability to occur even. For example, when you play with a lottery, you either ‘win’ or ‘lose’, but your chance of winning is nowhere near 50%.
SonicSP
30th November 2018, 15:24
*facepalm* Umm no, that’s not how probabilities work……
Garns (@)
30th November 2018, 13:13
I don’t like it- I said I was happy for Liberty to takeover as long as they didn’t ‘Americanise’ the sport- F1 has enough stats and timesheets to offer- this seems like the big hitter lining up in baseball and you get all his stats and graphics as he lines up to hit.
Its bad enough the graphics on the halo, no more please Liberty.
Yaru
30th November 2018, 15:30
Surprised to hear that as I thought having the tachometer over the halo was a pretty inspired idea.
I mean if the halo is gonna be there, might as well make use of it for something visually.
Keith Collantine (@keithcollantine)
30th November 2018, 13:14
I wonder if Brawn’s trying to be a bit clever here, and use this to demonstrate how DRS (and high tyre degradation)has made overtaking excessively easy in many scenarios, and add evidence to the arguments he’s already stated for getting rid of it.
I don’t need convincing F1 would be better off without DRS, but others seem to, and this could be a way to do that.
Tristan (@skipgamer)
30th November 2018, 13:50
Good point! Could show 100% for many of the easier overtakes with DRS and 0% for those where it doesn’t have enough of an impact. The argument could then be made that the goal is to get both scenarios closer to between 30-70%.
That being said I don’t think the data really needs to be or should be shared with the public if that’s the case.
One thing I’m hoping this means is that broadcasters will be more easily able to spot fights in the midfield where an overtaking battle is likely to occur before the fact, so we don’t just get a quick replay afterwards as we do quite often at the moment.
Robbie (@robbie)
30th November 2018, 13:55
@keithcollantine Oh great, now you’ve made me worry that some still need convincing, lol. Seriously though, you may have a point, but I truly believe the 2021 cars will not need drs nor be designed with it in mind. After all, the ground up restoration of the cars is the perfect opportunity to do away with it. Put another way, I will be massively disappointed if drs is still on the 2021 cars as imho it’s removal is a no-brainer once they get away from the current iteration of cars that are designed for it.
StefMeister (@stefmeister)
30th November 2018, 13:34
I think there’s too much clutter on the screen as it is so i’m not a fan of adding more.
Additionally as I said in a comment a few days ago I’d rather they give us less information rather than more because I am starting to feel that the amount of information we are getting is adding to the predictability & removing some of the tension & therefore excitement we used to get from not knowing some of this stuff.
I don’t want to be told how likely/unlikely an overtake is via a silly looking graphics full of numbers, I want to watch the cars & work that out for myself. I don’t want to be shown the likely gap between cars after a pit stop because half the fun you get from that sort of thing is not knowing exactly where a car will come out. I don’t want to see exactly how the car & tyres are performing with another graphic full of numbers, I can see this already from the timing data & how a car is behaving.
If we know what tyres everyone is on, If we know pretty much when everyone is pitting, If we are shown how likely or unlikely an overtake is, How a car is performing, What the likely gaps will be after a stop then why are we then surprised that as viewer’s things seem more predictable than they used to be?
I want to focus on the cars & the track action rather than been distracted by clutter that is predicting things i’d rather figure out for myself by watching the cars. I don’t want to be told/shown how a race is likely going to play out, I want to figure it out for myself & be actually surprised by it when it happens because that is a part of what makes watching races enjoyable to me.
Nunof
30th November 2018, 14:46
Well, this was expected after an US-based takeover of the sport: TV screens real state 80% covered by beffuddled stats, 15% with flashy advertisement, and 5% with the actual event🇺🇸
Adub Smallblock
30th November 2018, 15:00
Hey, I can plan my pee breaks!
Adub Smallblock
30th November 2018, 15:01
Can’t wait to see the probabilities shown for Stroll!
Paul F (@)
30th November 2018, 15:24
How about the likelihood of Verstappen crashing?
Maciek (@maciek)
30th November 2018, 17:31
Clearly this doesn’t go far enough: let’s have the graphic getting bigger as the probability increases with a big red seethru 100% flashing during the maneuver and a siren going off with digital glitter and confetti when an overtake happens – c’mon F1, thing big!
JWT
30th November 2018, 18:20
This is a god-awful idea made by people who have absolutely no clue about sport from a spectator’s angle. The whole premise behind sport as an enjoyable event to watch is it’s raw unpredictability.
We wait eagerly for a car to come out of the pit exit when an undercut is attempted, we hold our breath when a Ricciardo or a Verstappen is lining up an overtake. This tension is what makes sport enjoyable. And this enjoyment is the only reason sport exists as it does, but otherwise it is completely arbitrary.
We don’t want statistics and numbers thrown in our face to tell us how likely something that may or may not happen will happen. It kills the tension completely. These new graphics are another example of tiresomely vapid marketing talk done only to please the corporate at Amazon et al.
Another step taken in the wrong direction.
Kenny Schachat (@partofthepuzzle)
30th November 2018, 19:28
How could Liberty be so out of touch with what fans want to see and how they want to experience F1?
I have to assume that the majority of viewers find a greater part of the excitement of overtaking to be in the *drama* of it’s unpredictability, right? Why would they want any distraction to that drama, esp. some wonky “AI” stats? NO, no NO!
This is an example of doing things just because they can. It also seems like it’s am attempt layer interest on something that’s become less inherently interesting. If it’s something that the viewer can turn off, then fine.
I thin the real story is likely more along these lines:
“Oh well, our 2019 aero regs aren’t going to help overtaking after all…Hey Amazon is willing to pay us a ton of money to plaster their “Powered by AWS” all over the screen. Plus, we get to free world wide exposure at the conference.”
Wish I had an AI emoji that was powered by AWS for “Shaking my head”.
Kenny Schachat (@partofthepuzzle)
30th November 2018, 19:41
BTW, the NFL has been playing around with these AWS probability graphics for several months. They’ll show a video replay on a long pass play with these cluttered AWS stats about the probability of the catch, etc. Fortunately, they aren’t being used much during the games and not at all in real time. These have been limited to selected replays and mostly used in commercials for AWS, etc. They provide zero insight into the play: they’re just wonky window dressing.
Sure, probability stats can be valuable in decision making but the great part of the excitement and drama of sports lies in it’s unpredictability and those wonderful moments when the play overcomes the odds or when an someone does something extraordinary. OTOH, odds and predictions are essential to betting, so maybe that’s what the F1-AWS overtaking stats are really all about: emphasizing real time betting on the races. Uh, oh.
Nunof
30th November 2018, 21:15
For those who want to watch just Brawn’s keynote(1h39m17s) on the video above
https://www.youtube.com/watch?v=ZOIkOnW640A7&t=5957
stefano (@alfa145)
1st December 2018, 0:22
Yeah thank you, some predictability is definately what was needed. Finally, right? Why bother racing, just do a simulation now. You certainly don’t lack computing power.