Chapter 12- The Guy Who Was Certain

Devlin Aguado is twenty two, works at a phone repair counter, has owned four motorcycles in three years, and knows more about motorcycles than anybody Theo has ever met, in the specific sense that he can talk about them for six hours without stopping and has never once been wrong about anything in his own account of events.

Theo likes him. This chapter is not a hit piece. Devlin has helped him lift a bike, lent him tools, and once drove forty minutes at eleven at night to bring him a spare fuse. He is a good person to know.

He is also, in matters of diagnosis, a catastrophe.

The sprockets

In September, Theo's chain and sprockets reached the end of their life at about thirty four thousand miles, which is entirely normal, and Devlin talked him into changing the gearing while he was at it.

The argument was reasonable and Theo still thinks it was. One tooth fewer on the front sprocket makes a small machine feel considerably livelier in the range where it is actually ridden, it costs nothing extra since the parts are being replaced anyway, and thousands of people do it.

They did the job on a Sunday in the car park, which took four hours and would have taken a shop one, and it went fine.

Eight days later the FI lamp came on. Code twenty five, gear position sensor, circuit open or shorted. And underneath it, appearing intermittently, code twenty four, rear wheel rotation sensor.

Devlin's diagnosis, delivered in under a minute

"Gear position sensor. They go. Mine went on the six hundred. It's a known thing, there's a whole thread about it. Get the sensor, it's like ninety bucks, it's four bolts."

"Did yours have these two codes together?"

"Doesn't matter. Twenty five is the gear sensor. That's what it is."

This is the pattern and it is worth naming, because it is the most common form of bad diagnosis in the amateur world and it does not look like ignorance. It looks like expertise. Devlin was not making anything up. He had genuinely had a gear position sensor fail. There genuinely is a thread about it. Code twenty five genuinely is the gear position sensor.

Every individual statement was true and the conclusion was wrong, and the reason it was wrong is that a code names a circuit and Devlin was treating it as if it named a part.

The four seconds

Theo opened the log on his phone and looked at the column Mia had added in October, which by September was already three entries long.

Chain and sprockets, gearing changed, eight days ago.

That is the whole of it. That is the entire diagnostic contribution of a spreadsheet column that took four seconds to check, on a fault that a confident friend had just told him would cost ninety dollars and four bolts.

Because here is the thing about a gearing change that nobody mentions when they recommend one. The machine works out how fast it is going by counting rotations, and it works out which gear it is in partly by comparing engine speed against road speed. Change the ratio between the engine and the back wheel and every one of those relationships moves.

The speedometer reads differently. The computer's idea of what gear it should be in stops matching what the gear sensor is telling it. And on a machine that cross-checks those two things against each other, a mismatch that is now permanently present will eventually be recorded as a fault in one or the other, depending on which one the software decides to disbelieve.

This is well known on this family of machines. It is one of the first things anybody familiar with them asks when these codes appear: have you changed the gearing. Devlin, who has owned four motorcycles, had never once been asked it, because he had never once had a code he did not immediately explain.

What Theo did, which was not exciting

He put the original front sprocket back on. It took an hour and a half, because he had done it once already and knew where everything was.

He cleared both codes in the car park, rode it for a week, and checked. Nothing returned. He then wrote a line in the log that Mia has since described as the most Theo sentence in the entire document: gearing change caused it, original sprocket refitted, confirmed clear over seven days, do not do this again on this bike.

Cost of the correct diagnosis: an hour and a half and the price of the sprocket he had already bought.

Cost of Devlin's diagnosis, had he taken it: ninety dollars for a sensor that was working perfectly, an afternoon fitting it, and a fault that would have come straight back, at which point the next suspect on the list would have been the wheel speed sensor at a similar price.

The rule that came out of it

One change at a time, and re-read after each one.

It sounds obvious written down. It is almost never done, because when a machine develops a fault the natural human impulse is to fix everything you have been meaning to fix at the same time, and then when the fault goes away you have no idea which of the four things you did was the one that mattered.

Worse, when it does not go away, you now have four new variables and a fault, and no way back to where you started.

The version Theo uses is deliberately dull. Change one thing. Clear the record. Ride it for a period long enough that the fault would have shown itself. Read it again. Write down the result whether it worked or not, and especially if it did not, because a documented failure eliminates a possibility permanently and that is worth as much as a success.

What he said to Devlin, and what Devlin said back

He told him. Not smugly, and Theo was careful about that, because he had been the one holding a spanner in a car park at eleven at night more than once and there is no version of this where you get to be superior about it.

Devlin listened to the whole thing and then said, after a while: "So the code doesn't say what's broken."

"It says which circuit stopped making sense. That's a different sentence."

"Huh." A long pause. "I bought a fuel pump once."

He had. Two hundred and forty dollars, on a machine that turned out to have a corroded connector behind the tank, which he found six weeks later while looking for something else.

He bought a harness and an adapter that winter. He is now, according to Theo, extremely annoying about baselines, which is exactly what he deserves.

What Theo Learned

Confident advice is not the same as correct advice, and the most misleading kind is the sort where every individual statement is true. A code that names a circuit is not a code that names a part, and treating it as one is how people buy components that were working.

Keep a column recording what was done...

This curated preview is provided for personal reading. Copyright remains with Nobo Astoria.