Art of the Gimmick Rally
by John Meader
Sports Car, February 1962
Of the wiles and stratagems of the gimmick art there is no end. The desire persists here and there to set up rallies of increasing "difficulty", preferably with newer and better tricks. Of course, these should not be used openly. It is a point of honor that a cloak of reasonableness should cover the gimmick. An occasional protest can then be ruled unjustified with a straight face.
This is a matter of some delicacy. The possibility of a protest is a calculated risk which may have to be taken, but a flood of protests must be scrupulously avoided. If a gimmick is too crude it is likely to be a source of embarrassment.
Modern research in the gimmick field has developed some technical advances which promise to turn the art into a science. It is now possible to classify most gimmicks. This facilitates their selection and adaptation in such a manner as to eliminate a revealing thread of repetition, and it makes easier the study of a gimmick's age and history. A really new one may be good for a whole season!
It is a distinct aid to ingenuity to pinpoint the underlying theme of a gimmick, so that new varieties can be devised on a mass production basis with minimum risk of detection. Several new and interesting species of gimmicks have recently been discovered and described. It is even possible to formulate a general theory which, if it can be verified experimentally, will constitute a scientific breakthrough of major importance.
The following procedure was adopted in a recent investigation organized with great care.
First, a thorough search of the literature was undertaken. (See Bibliography. In addition to the works listed, a monograph by Potter is in preparation.)
New data were gathered in quantity by utilizing the services of J. Jittery Fairlamb, the long-distance off-course rally champion, who posed as an inexperienced rallyist and subjected himself and his navigator, Wilful X. McNutt to 27 more or less important rallies during the 1960-61 seasons. These sorties wore out two automobiles, but calibrated replacements were made available.
After 8,811 route instructions, 2,857 speed changes, 803 checkpoints, 4,222 miles off course, 55 traffic violations in 18 states and 10 spin-outs, the data cards were punched and thrown into the I.B.M. sorters and tabulators. Out of the analysis came the conclusions summarized below.
The number of gimmicks, always small, was found to be even smaller in relation to the scores achieved. In other words, a gimmick was likely to be good for considerable mileage, accounting on the average for 423 penalty points per contestant. This was very encouraging, although it seemed attributable in some measure to a bias in the scoring system which placed a heavy value on stray travel.
The "ppcg" index (penalty points per contestant per gimmick) showed a slight tendency to decline over the two year period. This decrease was traced partly to a diminishing frequency of gimmicks, and partly to a growing awareness of defensive resources; i.e., effects of the so-called learning curve, offset somewhat by a significant increase in the novelty and subtlety of the gimmicks encountered. According to the theory of the learning curve, each doubling of the number of rallies run tends to lower the "ppcg" index by 20 percent. At this rate winning scores may be predicted to occur to any competitor taken at random--the best way to taken them-after 16 years of intensive campaigning, provided the gimmick-incidence rate remains constant. Disclosure of this long-term threat to the value of gimmicks as a means of increasing rally scores stimulated all of the workers on this project.
A possibly discordant note was sounded when some of the evidence pointed to occasional negative influences. The gimmick record itself was good, but the ratio of protests to gimmicks was found to be increasing, indicating (of course) a need for better gimmicks. There seemed to be a minority opinion that scores could be increased without resort to gimmicks--as by counting ten points per second instead of one--but this view was dismissed as trivial. In scattered localities agitators have begun to organize cells dedicated to the heretic teachings of von Neumann, et al. Williams contends that the value of a game can be calculated, and that a negative value for Blue (the player in his picturesque notation, where the dealer is Red) indicates an unequal contest, theoretically good for Red until he runs out of Blues.
The research reported above was carried out largely on a statistical basis. Some think it failed to reach to the fundamentals, although it has provoked some furious thought. Other studies, following inductive methods, lead to additional conclusions, some of which may be valid.
Outstanding among these is the discovery that there is not only a creative purpose in the gimmick, but also a conventional understanding that originators are not permitted to take undue credit publicly for their scientific zeal and accomplishments. This code of professional ethics demands concealment of the gimmick. Local organizers may occasionally speak of the situations they have devised as a "challenge", a "searching test", an "interesting experiment", a "tension-builder", or even a "psychological overload". And they may take a justifiable, quiet pride among themselves in their DNF ratios, but expressions like tie-breaker" are frowned upon. A certain protocol must be observed. Standard practice is to describe all rallies in the advance notices as straight TSD or equivalent, and stop there.
Some 92.8 percent of all gimmicks, it has been found, may be classified as of single-, dual or triple-threat type.
The single-threat gimmick-for example, the wrong-way sign which cannot be read until the car has overrun the turn or speed-change point-is passing rapidly into disuse, along with such variants as the "typo" (typographical error), the "approximated approximate mileage", and the "indefinite identification". Formerly effective, these strategies have be come widely known and have encouraged the growth of defense mechanisms. Most rallyists will no longer go past a possible turn until they have, so to say, cased the junction thoroughly-they gather and confer at a doubtful mileage, and they have found that tailgating pays off handsomely in a really wild rally.
The two-threat gimmick, of more recent origin, enjoys a considerable current vogue. No doubt it grew out of a simple or single-threat gimmick dressed up for a special occasion--for instance, the use of a wrong-way sign at night, or an incorrect mileage just before a checkpoint. These situations are elementary, easily recognized, and they risk a certain amount of grumbling if not forthright protest. But there are many newer variants, such as making separate instructions out of simultaneous actions; that is, requiring a second maneuver before a first is completed (the "dangling instruction"), or requiring the use of data there was no apparent need to note (the "overlapping instruction" or "cold deck"). One advantage of these is that they can be made technically correct, virtually protest-proof. This type of gimmick, in up-to-date guise, seems likely to enjoy a long and useful life, although it does tend to produce a gradual diminution of entries.
The triple-threat or multiple gimmick, so far a minor variety, is increasing in favor and frequency. In principle it allows wider scope for ingenuity, has a higher "ppcg" index, and is available in practically unlimited quantity. An obvious illustration is the wrong-way sign (simple gimmick), used at night (double gimmick), with a trap checkpoint close at hand (the triple-threat). On occasion this situation has been planted in a high-speed section of the course, but further refinement is seldom necessary and emphasis should be avoided. Properly done, this type of gimmick leaves the rallyist not quite sure what actually happened to him, i.e., mentally handcuffed (cf. Houdini, op. cit.). A few contestants may see through the plot and recognize it as intentional; but this is usually on the way home from the rally, too late to do anything about it.
Other species of gimmick are not so readily classified, but their occurrence, although rare, should be reported for the benefit of the workers in this field. A few example of these are given below.
Set up an open checkpoint easily visible just beyond an intersection where a turn has to be made. Cars are lured straight in, forgetting to make the turn. When it first appeared this gimmick scored a very high "ppcg". After two years three cars are still off course, last seen heading for Canada. The performance coefficient dropped sharply in later outings, but new variations have been invented; e.g., an obscure turn just before a large, red STOP sign, where the driver will be looking for a chance to get through the cross-traffic, and is likely to miss the turn. (Add a dash of speed-change to taste, in order to keep the navigator's head down at this point!)
Take two crossroads a mile or more apart, the first with a dirt road left. Have the general instructions prohibit turns onto dirt. Now instruct: "Right at crossroads." Every one will go right at the first one and get lost. That wasn't a crossroad, see? Please refer to the general instructions!
Find a convenient suburb with three streets named Glorious Avenue, Glorious Road and Glorious Street; or just plain Glorious, or a reasonable facsimile thereof. Prescribe a turn onto Glorious Street, the last of the sequence, so as to give all cars fair and equal chances to go wrong. For best results mix this device with a previous typo where a misspelled sign marked a correct turn.
For a change of pace get the students thoroughly irritated with twenty miles of straight road at slow speed, holding up civilian traffic; then call for a swift sandwich of two speed changes and a turn within fifty feet. This one really throws 'em!
The pseudo-gimmick can be used occasionally to improve the effect of the real shocker. Bury somewhere in the general instructions a rule reducing all speeds 50% for one mile at each tunnel, then select a route free of tunnels. The smart players, who often study the general instructions, will wear themselves out looking for tunnels and overlooking some thing else. This is a selective gimmick, fairly artistic because it has no effect on beginners. Or, try "Right at Zilch's mailbox", which will annoy an experienced rallyist, bringing on a pronounced state of nerves. You know perfectly well that mailbox landmarks are not allowed in respectable competition, but the actual situation was a simple right at T. No one could be misled. Be sure to stick to that story. Or, perhaps you have already thought of ordering: "Change average speed to 33.22 mph (from 32.32) at 'SLOW CHILDREN' sign". That works!
The fuzzy-logic idea was probably discovered by accident when an ambiguity in the general instructions was overlooked by a rallymaster later horrified to see what happened to his flock. But it has possibilities. For instance, tell the entrants to follow a numbered route when told to do so. They will think they understand. (Don't say if and only if you are told to do so; that might be perfectly clear.) Add by way of further explanation: "The road you are on will be indicated by stop signs, etc." (Don't even hint that route and road have different meanings in your itinerary.) Now emphasize the rule in a special bulletin (but don't explain the explanation). This setup is good for a whole rally. It piles up the points something scandalous, 1,800 at a crack.
The multiple-source-of-information artifice is not surefire, but it is awfully good. After laying out the course divide the route instructions into four piles, selected fairly by tossing a coin. One pile goes into the general instructions, safe from scrutiny; the second onto the route sheets as usual. So far everything is more or less sanitary. The third pile is made into footnotes for the time slips handed out at controls, and the fourth is worked in as little strips stapled over the footnotes. There are excellent chances that some of the information will be lost, never seen, blown away, stuffed into the wrong pocket, or thrown out when cleaning up the cockpit at the luncheon stop. This method is absolutely impartial.
The value of any gimmick's "ppcg" is enhanced when it is used in the run to the odometer checkpoint. This is a bit drastic. Although not unheard of, it is generally considered off limits. A milder version, more highly esteemed, is to be just a little careless about the instructions covering transit zones.
On the basis of these discoveries, a general theory of gimmicks has been formulated-all gimmicks are designer to improve the score of the rallymaster against the contestants. If the intention is lacking it is not a genuine gimmick, even though it may have the same effects. This theory, first suggested by McNutt, has gradually acquired a warranted assert-ability, but experimental verification is lacking. Gimmicks are proverbially disowned by their originators, so that direct attribution and measurement becomes somewhat conjectural.
Now Fairlamb has proposed an ingenious scientific test --organize and advertise a series of gimmick rallies as such, with the rallymaster eligible for prizes and national points. Then count the number of entries, the number coming back for a second go being weighted by a factor of two, etc., and score the organizers on the ratio of DNF's to total weighted entries. As a tie-breaker, bonus points could be awarded for accidents on an ascending scale. This is a daring concept, but one calculated to produce a scientific advance of the first order.
Meanwhile, for some sidelights on the present state of the gimmick art we sought out McNutt himself. You went into this racket cold?, we asked. "Nope, we tried some local rallies first, but they was always enough prizes so you had to win something, like for instance a pint of windshield washer. Then we heard these natural rallies was the McCoy, the girls dressed better, and they ran on watches simonized with radio."
They had found the National rallies better organized, then? "Heck, yes, these guys really know how to take you! We made six starts the first year, and the best we done was sixteenth; from last that is. Then we dropped to fifteenth, but they had changed the point system. We met a lotta nice guys, and we were catching on."
You had some difficulty at first, perhaps? "Yeah, we couldn't tell what we done wrong. So we would write in for a low number and follow a hotshot. You hang onto his tail to the first check; that costs you maybe 50 points, but it's cheap at the price. Now you keep him in sight all the way. Only he goes off and you're lost. Must be he gets trapped by a gimmick. What else?"