Saturday, October 6, 2012

F1 2012 Japan Qualifying Summary Snapshots

(Part of the rationale for these posts is to explore what we can learn about a race just from the published data. Many of the "anomalies"/deviations from expected behaviour or "odd things" in the charts may be easily explained by anyone who saw the session, but can we use the data to help us as questions about what happened to a particular driver for example (even if the answer turns out to be obvious, like a driver spinning off at the start of the session). Another way of thinking about this might be to ask what additional data do we need to explain what happened in a session if all we have to go on to generate a race report is the published data?

Comments appreciated particularly relating to:
1) what sorts of things do each of these charts reveal about the session/what sorts of commentary/interpretation would you like to see/how might these charts help make sense of, reveal things that happened in, or hint at possible stories relating to the session?
2) how could the charts be improved?
3) what other charts might be useful?

Thanks ;-)

Deltas to best overall sector time by driver 1/driver 2 within a team:
 Normalised sector times and normalised speed, ordered by car number:
Qualifying session laptimes, ordered by classification - both ALO and RAI, at least, appear to have slipped up in Q3? (I find the green and blue quite hard to distinguish? Use symbols, maybe? A line chart may even be in order, using session groups, because line crosses would show where things maybe went wrong?)
How close to their ultimate laptimes did each driver get? SEN and KAR were way off, and VET didn't do so well either?
 Sector comparison across qualifying - what are the deltas to the best overall sector times?
Sector time deltas compared to best overall sector times, summed, ranked by classification:
 If we order the drivers by car number, team performances are a little clearer to see?
This chart suggests to me that sector 1 is where most time is being lost in the first sector, but I'm not convinced about those colours (does a block of the pinkly colour feel bigger than a similarly sized stacked block of the green colour?) Looking at the range of the y-axis in the chart below, which shows the deltas to the best sector time, we do see that for cars 6-8, they are losing more time in sector 1 than sector 2.

Sector time deltas, ordered by car number, split out by sector:

Sector 3 is quite close cut, with sector 2 in particular differentiating between the teams at the back of the field, and sector 1 particularly differentiating those team from the top three quarters of the field. (Just noticed that I have posted a normalised chart, showing how close teams are to the 107% threshold.)

Sector deltas by rank for those sectors (maybe also worth doing by overall classification rank?):
So for example, here the order is by overal qualifying classification - of a driver looks to be high in a column, for example, PER in sector 2 and 3, they're losing time there... If they appear to be low in a column, for example SEN in sector 1 or ROS in sector2, they're making up time.

That then suggests this next chart, immediately below, where we compare sector ranking with overall classification ranking. Using overall classification on y and sector ranking on x, if a label is above the line the driver performed better on that sector than we might expect, and if it's below the line, they performed worse...
So for example, ROS did better in sector 2 than his overall classification suggests, and whilst MSC messed up/was below the line in sector 1, he did well (above the line) in sector 3... BUT also appears to have the better of WEB in sector 2? PER is underperforming in sector 2 and 3, as was also suggested in the previous chart.

The previous charts, based as they are on best sector times, suggest that HAM was performing better based on best sector times than his classification suggest. Looking at the lap time deltas relative to each driver's ultimate laptime, we see that indeed HAM was way down on his ultimate lap time in each of the qualifying sessions; we also see that HAM recorded a better time in Q2 than he did in Q3, as did RAI:
 Session and ultimate laptimes, ordered by car number:
How are the teams losing time by sector? Does this chart help?
Mercedes and Toro Rosso seem to be losing out in sector 1? Maybe ranking teams by their average sector times would be useful? Something like this maybe?
Hmm - a boxplot isn't right... not enough data for a distribution. Maybe just the tomes for each driver in a team and a bar for the team mean? How about:
Or maybe presented by team?

No comments:

Post a Comment

There seem to be a few issues with posting comments. I think you need to preview your comment before you can submit it... Any problems, send me a message on twitter: @psychemedia