Bed of Procrustes
Greek mythology talks about a person called Procrustes. He had a stronghold on Mount Korydallos at Erineus. There he had a bed, in which he invited every passer-by to spend the night. He also had an obsession with making people “fit” in bed.
As it happened …
If so, Procrustes would violently stretch them so they would fit the bed; Killing them. Or
In which case, he would cut their legs to make them fit his bed. Once again, killing them.
The point of the story is … Procrustes had arbitrary and rigid standards of evaluation. (Apart from the fact that he was a murdering moron) And he would violently apply them.
Procrustean Ways
We observe such ways commonly in our society.
Ones with these measurements are beautiful. All others are not.
All others are idiots.
And one with 8 lack/ annum isn’t as good.
In the IT industry, we see the same Procrustean standards.
His colleague writing 100 lines of code is not as good.
Working 6 shows a LACK of intent.
Solving 2 is NOT.
Attending 1 meeting is less productive.
While the application differs. Ideas remain the same.
Arbitrary and rigid standards of evaluation applied violently.
Problems with Procrustean measurements
Duh !! They don’t work
From the 80/20 principle, we know that 20% of issues take the most time. And 80% get resolved quickly.
Indians have a saying … “Not all fingers are alike”. Meaning, people come in different shapes and sizes. In the same way, challenges come in different shapes and sizes. If challenges have differences, their resolutions will also have differences. In terms of complexity, invested time, level of the resolutions, etc.
So, comparing 1 high complexity issue to several low complexity ones is poor thinking. Let us consider this …
Who has put in better work ?
Procrustean standards would say the – Developer. However, we know from experience that the 20% lines of code are more valuable.
Procrustean standards don’t work in the real world !
Kill the unfit
Speaking good English is one of the IT industry’s measurement units. Let us consider this to be a procrustean measurement.
Imagine a brilliant programmer from Kerela. While all is good, he isn’t good with English language. Wouldn’t most of the industry reject ( if not castigate ) him ? It would !
Now I tell you that the kid won Google’s international open-source software contest ! and that this is real. Link here (not to say that Abhishek’s English needs work. Am sure its great.)
Procrustean standards kill the unfit ( instead of grooming them ) and therefore are bad.
Killing such talents would’ve meant depriving the world of all their talents.
The same thing happens in the IT industry. Measured on procrustean standards – talents of many techies are subdued.
Metrics become goal
Rewarded behavior is repeated behavior.
9-hour workdays are common in the IT industry. Completing one’s quota of 9 hours is also a BIGG deal. A procrustean measurement !
Since organizations look to get employees to clock their time. That is exactly what employees do. Mind the fact that the hours are hardly ever “productive”.
Work stretches to fill the time, no more. No less !
Economics tells us – If the “number of tickets” is the goal, then engineers target the “number of tickets”. Despite the fact that they don’t add value.
More so, engineers purposefully create issues – so they can be solved later and give them credit. Meaning, rewarding problem-solvers results in promoting the problems.
Explained this in another blog post : The Cobra Effect – techwiddeep.com
While so, output quality does not improve.
Word of caution
Lines of code, the number of tickets, length of documents are valid measurement units. So, Someone writing 0 lines of code in 6 months is surely in a different universe. Compared to someone writing 5000 lines of code during the same time.
However, 100 lines of code in a critical place can have a bigger effect. Compared to 1000 lines of code at less relevant places.
My 2 cents …
In the movie Moneyball, Peter Brand was played by Johan Hill. He explains – that there was a mad race among teams to buy the most expensive player. And was so, across the industry.
The idea being : Matches are won by better players. Better players are more expensive.
So buy them. The cost of a player was a procrustean metric ! (expensive equals better)
Since the target is to WIN. Instead, Brand explained, teams’ should focus on buying runs.
This change brought a paradigm shift within the industry. Also, brought massive changes in the fortunes of the team.
Something similar needs to happen within Information Technology.
The idea being : Softwares have lines of code. So, writing more lines of code equals a better engineer.
Lines of code is a procrustean metric !
Similarly, the following should NOT be goals …
The goal is to deliver software. While the functionality is delivered via code, hours, tickets, and meetings. These underlying metrics are NOT goals. Therefore, trimming and stretching engineers based on such metrics leads to poor decisions.
A good system should reward and help, techies who help ‘deliver’. Period.
Procrustes was eventually captured and killed. The same thing happens to his followers.
Easier said than done. But don’t be those guys !
