The Pomeroy Index

In his book "The Grand Prix Car" Laurence Pomeroy developed the concept of an performance index for Grand Prix cars. He awarded the 1906 Renault an index value of 100, and then, by comparing the performance of cars from successive years in comparable conditions, constructed an index for the performance of all major cars until 1953. This work was continued by LJK Setright who took the analysis up to 1966. Setright arrived at the conclusion that the 36 valve Ferarri which won the Italian GP that year had an index value of 200, making it just twice as fast as the REnault.

Incidentally, neither Pomeroy nor Setright thought drivers made enough difference to be worth considering in their analysis!

Does anybody know whether the index has ever been created for the years since 1966? If not, it is something I would like to try for myself, perhaps through this forumif there is sufficient interest.

There was an event held at Silverstone every year called the Pomeroy Trophy, where I think the results were worked out using Pom's theory.
I might be wrong but i think it was Aston Martin Owners Club or VSCC that ran the meeting. I feel sure it is still being run to this day.

As for the driver's contribution... how could an engineer ever allow for that?
Dealing in black and white could never permit emotions, varying skills, determination or excellence play a part...

------------------
Life and love are mixed with pain...

Huw,

THe Pomeroy trophy was (and is) something completely different. It uses an incredibly complicated formula, including interior dimensions and performance in certain test to determine the best all-round touring car. Trouble was, it was rarely won by the sort of cars Pomeroy wanted to encourage.

Roger
I, for one, would like to see you apply the formula to cars since 1966.
In fact I would like to do it myself, but have too many other time consuming projects on my hands. I could offer encouragement from the sidleines if that will help...
Pomeroy used to, if my memory serves me correctly, refer to this formula in reviews of the GP year in The Motor Yearbook from the late 1940s to the late 1950s.
I always have been in awe of it.

Roger
By the way, The Pomery Index aside, I just looked at your profile and your e-mail address suggests your name really is Roger Clark.
Whenever I saw your name on a posting I thought you had adopted the name of the great rally driver.
My major memory of him was a man, well lubricated with alcohol, fast asleep on the floor in the aisle of an Ilyushin airliner between Tashkent in Uzbekistan and New Delhi in India in 1993.
In middle age, he could be in that state each night, then wake up and drive like a 20 year old budding world champion the next day - on stages of that 1993 London-Sydney Marathon.
He never ceased to amaze me.
I had a disagreement with him in the mid-1970s and the piece I wrote about it, I was told later, went all the way to the Ford board room in Dagenham.
But in 1993 he had mellowed as, I have found, many ex-GP drivers do, and we got along well.
And wow! Could he drive!
Sadly, his neglect of his health appeared to catch up with him and he has since passed on - long before his time.

Barry,

Yes it is my real name. I suppose you always feel an affinity for a namesake, and i always followed roger's career even though I don't following rallying that closely. As you say, he really could drive, and was about the only home grown talent to take the battle to the Scandinavians.

I will tackle the Index, but don't expect quick results. Time is always pressing, and I work slowly!

I just hope Barry's mental picture of a drunk crumpled in a heap of Indian sqalour hasn't torn down the image too much...

------------------
Life and love are mixed with pain...

Sounds like we had more in common than a name...

Dear, oh dear!

------------------
Life and love are mixed with pain...

I am reviving this ancient thread to illustrate a project I started over the Christmas holidays.

I wanted to investigate the improvement in the performance of Grand Prix cars over the world championship era, from 1950 to date. I did this by considering the pole position speed at Monaco and Monza. I chose those circuits because they have been in use for most of the years in question (with changes, of course) and because they represent more or less the extremes of the speed spectrum. The results of the analysis are shown here.

http://www.jumbani.d...201950-2005.htm

For each circuit, the first columns show the speed of the fastest practice lap and the car which achieved it. I have then shown the percentage improvement from the previous year in two columns, the first for all years and the second omitting years when there was a significant change to the circuit. Those changes are noted. A column to the right of the circuits' analysis shows changes to the regulations which I felt would have had a significant effect on performance. I gave up on this in the 1990s when things become too difficult for me.

The mean improvement, in years when the circuits didn't change is 0.97% at Monaco, 1.21% at Monza. The difference may be a reflection of the was cars have become less suitable for the street circuit, but it may also show that technical improvements, particularly aerodynamics, have greater effect at high speeds. If the average of 1.21% had been maintained every year at Monza (ie, if the circuit had not been changed) the cars of 2005 would be lapping it at 377kph, or approximately 236 mph.

I find it interesting that there are some years when the improvement is less than one might expect. One example is 1964, when the cars were largely unchanged, but Dunlop had introduced new tyres which should have increased speeds. Another is 1978, where Andretti's speed at Monza is very little faster than Hunt's of the year before, despite the technical leap forward represented by the Lotus 79. 1982 stands out as a particularly good year in this respect.

By way of comparison, i have also created a table from Pomeroy's Grand Prix Car, showing how performance developed over the years 1906-52. In this case, speeds are represented as an index value, not by speed on any particular circuit.

http://www.jumbani.d.....01906 -53.htm

The mean annual increase in those years was 1.51% of 2.30% if you discount the years when the formula changed to reduce performance significantly (1922, 1938 and 1952). This may be due to more frequent changes in regulations in recent years, but it seems undeniable to me that progress in the first 50 years of Grand Prix racing was greater than in the second.

I would welcome any corrections or discussion on this work. Both tables are html version of Excel 2003 spreadsheets. I can make the original spreadsheets available if anyone is interested.

Fascinating exercise, Roger
I don't have the patience to do this sot of thing for myself, but really appreciate it when someone else does

Thank you Roger.

Can you apply your method to Dieppe 1907, 1908 and 1912, given that the circuit was virtually unchanged in that period, and compare it with Pomeroy's index?

Originally posted by Roger Clark
..., but it seems undeniable to me that progress in the first 50 years of Grand Prix racing was greater than in the second.
...

The law of diminishing returns. You can apply that to many things. It's also the reason that the wealthy F1 teams do best when regulation changes and the smaller fish then gradully catches up.

And good work, btw.

Originally posted by Roger Clark
Both tables are html version of Excel 2003 spreadsheets. I can make the original spreadsheets available if anyone is interested.

Yes please Roger

Did F1Racing do something similar to this recently? Or am I imagining things?

Originally posted by Kpy
Thank you Roger.

Can you apply your method to Dieppe 1907, 1908 and 1912, given that the circuit was virtually unchanged in that period, and compare it with Pomeroy's index?

For most of the period he covered, Pomeroy didn't have the benefit of officially timed practice sessions when all cars could be assumed to be driven to their limits. He therefore had to take his evidence from lap times in the races and sometimes from other events including sprints and hill climbs. Apart from this, and expressing the results as an index value, his method was the same as mine, or perhaps I should say the other way around.

Originally posted by jcbc3


The law of diminishing returns. You can apply that to many things. It's also the reason that the wealthy F1 teams do best when regulation changes and the smaller fish then gradully catches up.

And good work, btw.

I'm not sure about this. If you looked at Pomeroy's analysis when he published it you would not see any signs of diminishing returns. I can see no reason why such a law should apply at one time (ie the present) and not others. I think that technical progress is a function of tree things:

1 the resources factories are able to devote to design and development
2 the innovation of the designers and their ability to introduce new ideas
3 the stability of the regulations.

In recent years, changes to the regulations have reduced progress, as they were designed to do. In the past, changes have often stimulated progress, examples being 1934, 1954, 1958, 1961 and 1966.

Originally posted by Roger Clark

I'm not sure about this. If you looked at Pomeroy's analysis when he published it you would not see any signs of diminishing returns. I can see no reason why such a law should apply at one time (ie the present) and not others. I think that technical progress is a function of tree things:

1 the resources factories are able to devote to design and development
2 the innovation of the designers and their ability to introduce new ideas
3 the stability of the regulations.

In recent years, changes to the regulations have reduced progress, as they were designed to do. In the past, changes have often stimulated progress, examples being 1934, 1954, 1958, 1961 and 1966.

I may put my foot squarely in my mouth now.

The changes you refer to here are to larger engines so wouldn't you expect a "stimulated" progress?

Regarding your third point I would say that you are right in as much as stability gives you the ability to understand the technology better. However, I would also argue that stability of regulations also proves the law of diminishing returns.

As Ron Dennis is fond of saying, major rule changes benefits the bigger teams as they have money to throw at a certain problem. The smaller teams then gradually catches up duing the following seasons as the technology filters down and they have thrown a couple of years of development budget after the car.

As technology matures, change slows, and in Grand Prix racing as in anything else it is change which improves performance.

Consider: by 1930, a 1900 passenger motor car was unbelievably antique and primitive. In 2000 I bought a 1970 motor car which was entirely up to date technically (although its steering and handling were a bit 60ish), and its platform and engine, suitably modernised, form the basis of the current model.

In Grand Prix racing, the years of radical innovation - rear engine, monocoque tub, all-independent suspension - have long passed. I can't think of anything likely to change that setup as radically as the change to that layout in the 50s and 60s. Change now, whether hedged about by rules or not, is bound to be incremental, not radical.

What we have is endless, inventive and expensive refinement of an old concept.

In 1968, LJK Setright published "The Grand Prix Car, volume 3", a companion to Pomeroy's earlier work. In it, he continued the analysis of the Pomeroy Index up to the 1966 season. Being Setright, the cars he chose were often somewhat eccentric, but I have added his conclusion to those of Pomeroy.

http://www.jumbani.d.....formance .htm

The fastest car listed is the 1966 Brabham with a Py Index of 197.5. This represent the car's relative lap speed over a variety of circuits. We can use it to evaluate the performance of the fastest car of 1967, the Lotus 49 by comparing their lap times on circuits where they both ran. There are six such circuits, the following shows the amount by which the Lotus was faster:

Spa + 6.6%
Zandvoort +3.7%
Nurburgring +3.3%
Monza +4.0%
Watkins Glen +4.3%
Mexico City +5.6%

I am inclined to regard the Spa figure as not representative, a combination of an extraordinary performance by the Lotus driver and a less than typical performance by the Brabham in 1966. If we take the mean of the other five circuits we conclude that the Lotus was 4.2% faster than the Brabham, giving it a Py index of 205.8. Reference to Pomeroy's an Setright's results show that this is an outstanding improvement for an era.

In the Grand Prix Car, Pomeroy put forward his celebrated theory, that if all else is equal, circuit performance will vary with the sixth root of engine power. He could give no reason why this should be so, merely observing that it was so, based on his records of almost 50 years' motor racing. If we assume that the Repco 620 gave 330 bhp and the 1967 DFV 400, the sixth root law suggests that the Lotus should be 3.3% faster than the Brabham. It seems probable that most of the remaining 0.9% improvement came from the tyres which were developing rapidly at that time.

In 1968, the Lotus was again the fastest car, and we can compare it with its 1967 stablemate on eight circuits:

Jarama -0.23%
Spa -19.2%
Zandvoort +0.9%
Nurburgring -21.0%
Monza +2.2%
Watkins Glen +2.0%
Mexico City +2.2%
Kyalami +2.9%

I have included Kyalami because in both 1968 and 1969 the cars which ran there were basically those which had finished the previous season. We should ignore the figures from Spa, where the 1968 Lotus never ran properly, the Nurburgring, where it rained throughout practice and Jarama where Team Lotus were understandably thoroughly disorganised following two driver fatalities.

The 1968 season was notable for the rapid development of wings and the Lotus had high, suspension mounted wings at the last four races listed. If we take the mean improvement of these, we conclude that the car was then 2.3% faster than in 1967. Apart from the high wings, the 1968 Lotus had modified suspension, greater drivability from the engine and improved tyres. We can assume that the 0.9% improvement at Zandvoort came from these things and the low wings and spoilers that were then used. This implies that the high wings were worth 1.4%. I would have expected it to be more.

A 2.3% improvement gives the high-winged Lotus a Py index of 210.5.

Incidentally, the fastest Lotuses at Monza, Watkins Glen, Mexico City and Kyalami in 68/69 were driven by four different drivers, which must prove something, but I'm not sure what.

In Autosport of 25 November 2004 the Pomeroy index is applied to cars up to that time.

Originally posted by karlcars
In Autosport of 25 November 2004 the Pomeroy index is applied to cars up to that time.

The article was by Mark Hughes and covered the world championship era. Strangely, they had significantly different results from Setright for the same cars. My biggest gripe is that there was no discussion about how they arrived at their results, beyond saying that it was based on practice times. to me, the investigation and analysis are more interesting than the end result. I had hoped that we could have a discussion on TNF about this but it seems that there is not the interest.

Being Autosport they had to build the article around the "5 fastest cars" which they did by the number of index points improvement from the previous year. It would have been better to have done it on the percentage improvement. For what it's worth, their top five were:

1 1967 lotus
2 1966 Ferrari
3 1954 Mercedes-Benz
4 1970 Ferrari
5 1962 Lotus

As these cars all come from relatively early in their period, it is probable that doing the calculation on percentage gain would give similar results.

Also being Autosport, they had to illustrate a piece about the 1966 Ferrari 312 with a picture of Bandini at Monaco, but Haymarket have previous on that one.

It would be interesting to apply the same principles to Tasman cars at Warwick Farm and Sandown Park... and the NZ circuits, too...

You'd see a quantum leap at the Farm with the wings of 1969.

Regarding the apparent stagnation of 1964, I think that's probably explained by the trade-off of tyre width adding to drag on straights while giving superior cornering power. That would certainly apply at Monza.

One wonders, too, with reference to 236mph lap speeds at Monza, just what you'd see at Leyburn if it were still usable.

Originally posted by karlcars
In Autosport of 25 November 2004 the Pomeroy index is applied to cars up to that time.

does someone have that "up to date " pomeroy chart and could post it

I may have missed a reference in the thread but Stirling Moss and Laurence Pomeroy did a book called -Design and Behaviour of the Racing Carpublished in 1963. In the book Pomeroy and Moss discuss post war GP cars and technical aspects. I can quote more from it if reqiured.
Richard

Not sure if this is off topic, but...

Some of the Le Mans entries during the '50s would sport a roundel, similar to an RAF roundel, on the front wing, just below the windscreen; particularly British cars such as Astons and Lotuses.

I think I read somewhere that these had something to do with the Pomeroy Trophy - can anyone confirm this?

Thanks

Not sure if this is off topic, but...

Some of the Le Mans entries during the '50s would sport a roundel, similar to an RAF roundel, on the front wing, just below the windscreen; particularly British cars such as Astons and Lotuses.

I think I read somewhere that these had something to do with the Pomeroy Trophy - can anyone confirm this?

Thanks

Absolutely nothing to do with the Pomeroy Trophy. The roundel denoted that the car was competing in the Rudge-Whitworth Bienniel Cup. See here and or carry out a "Search" for Rudge Whitworth for more detail.