https://www.duolingo.com/BerberuEsperanto

Esperanto from English learners 700K (from 600k in 56days)! EO←ES >30K Congratulations!!

Great achievement for both teams. The EO←EN team should reach 1M by 130y anniversary of Unua Libro (26Jul17). That's 50% over the current estimated base (now 35%).

EO←ES is the fastest growing (by %).

Gratulon. Bonŝancon!

2 years ago

15 Comments


https://www.duolingo.com/jirka92122
  • 25
  • 20
  • 11
  • 9
  • 5
  • 4
  • 26

% is not a good way to measure this since growth is most likely absolute rather than relative to the existing size. So new courses will always grow fastest percentage wise.

The fact that 100k was added in 56 days is amazing. That means it is almost 2k new every day, rather than the 1k every day it was a while ago. I wonder what caused the rate increase. Is it just getting "more viral"? That is, more people know about it now and tell their friends.

2 years ago

https://www.duolingo.com/DaveRutan
  • 25
  • 11
  • 2
  • 1047

Since the first of this year I've noticed an increase in the rate of new learners. It's stalled a few times, but that may be a reporting glitch as it corrects itself in a few days.

2 years ago

https://www.duolingo.com/BerberuEsperanto

I reported that fault a few weeks ago (no reply). When the 2nd most significant digit changes it goes back. Eg. EO<ES is now 30, from 30.6; it should be 31. It will fix itself once the 3rd digit is >=0 (probably rounding issue with formatting, as they only show 3 most significant digits)

2 years ago

https://www.duolingo.com/LaArakido
  • 25
  • 21
  • 21
  • 21
  • 15
  • 12
  • 10
  • 8
  • 1550

I've noticed the same fault/error with several other languages. And it has existed for at least a year now.

2 years ago

https://www.duolingo.com/BerberuEsperanto

There's YouTube that shows majority are not comfortable with trying to understand the *compound progress (% increase). The more people use it, the more 'free marketing'. Also (long term) humans grow by % and access to internet growth is by %.

Of course, the progress is not linear or smooth. After smoothing it's like the Gartner https://en.wikipedia.org/wiki/Hype_cycle .

Russian, Norwegian, Polish, Esperanto (a bit slower) have faster growth (from EN). Hebrew, Vietnamese, Hungarian, Greek even faster as they are 'younger'.

EO for ESpaña is the fastest. From 20 - 200/day average over the whole period since 16/9'16 (missing snapshot on 11Nov). Overall 124% yearly growth (over 130 days).

EO for EN has daily long term avg 0.93, 1.10, 1.20, 1.20K over 56, 89, 120, 130 days. So, it's growing with 81% yearly growth (used to be 84% recently but slowing a bit).

Native languages would tend to plateau as there's no point to have too many L2 speakers vs native. Irish, Hebrew, Greek would have different nature (different influence).

2 years ago

https://www.duolingo.com/jirka92122
  • 25
  • 20
  • 11
  • 9
  • 5
  • 4
  • 26

With percentage, you are expecting exponential growth. We should be far closer to linear growth. Here's why: New learners are I'd bet mostly from other sources than having been told by old learners to look at Esperanto, so I wouldn't expect the number of new learners to be in any way proportional to the number of current learners.

One of my favourite pet peeves as a mathematician is people putting exponential growth on everything. Everything is measured in percent growth, which is nonsense. Exponential growth is WAY to fast. Very few things really grow exponentially.

2 years ago

https://www.duolingo.com/BerberuEsperanto

OK, exponential growth is overused. Human population increases exponentially (since the Industrial rePollution).

In the real world, there are variations on Logistic Curve of growth, which comes closer to the Gartner Cycle. ~~ Latin, French, etc followed that curve.

The limiting resources(people that need it...)/friction (effort...) is lower with EO. The biggest limits are non-awareness, political pressure, and access to resources(growing). As critical mass builds, quality/accessibility (# internet users) improves, the people have more reasons to join the fun and make their communication more meaningful.

Other languages will reach plateau much sooner (proportionally to base speakers) vs EO. So, EOs curve will be more exponential than linear (~ population growth).

The future is bright, EO EO EO.

Edit: I missed the limit of EN speakers (not growing/exponentially). But the EO-ES (and overall EO) will follow that curve.

2 years ago

https://www.duolingo.com/jirka92122
  • 25
  • 20
  • 11
  • 9
  • 5
  • 4
  • 26

Logistic is perhaps a better model, it is still a model where growth depends purely on the current population, which I doubt is the main driver of growth. The best argument against exponential/logistic model here is that in the beginning the growth jumped immediately into pretty fast essentially linear growth rather than being slow for a long time before picking up as the population in the course slowly grew.

For world population logistic works because there is no large scale extraplanetary migration, and war and disease outbreaks are thankfully not the main driver of population change. Population of the world has natural death rate and birth rate and that only depends on the current population. So logistic is the right curve. But for example already modelling population change for a town or a city where migration in/out depends on a lot of other aspects then logistic or exponential model is often way off base. Especially for small towns. So for example, "fastest growing town" is often a town or a suburb near a large city which is growing perhaps in a completely average way. It is then an artifact of geography and arbitrariness of town borders, but it doesn't actually mean anything.

BTW, the Gartner curve is a more vague concept and in our case would more be connected more the the growth rate, but I don't think we saw this sort of "cycle" in terms of rate, as the growth rate seemed fairly steady.

I think we don't really have the date to understand what is the driver of new learners, possibly duolingo has more information here, but they do not share these kind of statistics (if they actually have it).

I think best measure is relative to the growth of other languages that were out for about the same time.

Companies love to cite percent growth for a new product because it makes things always look far better than they actually are. But exponential/logistic (unless modified) are bad models here because sales do not only come from referrals based on existing customers.

2 years ago

https://www.duolingo.com/BerberuEsperanto

Libereco, Frateco, Egaleco, Espereco !

2 years ago

https://www.duolingo.com/Luko.
  • 19
  • 17
  • 13
  • 10
  • 8
  • 8
  • 3

Aparently our growth changes each time something significant happens, I think the eo-en course started growing faster when the eo-es course was released, and both started growing faster when it became available in the apps. I think next big thing will be the appearance of a course for Russian speakers and later eo-es achiving 100k and eo-en 1M. Unfortunately eo-es course was growing at a rate of 1,1k a day a few days ago but now it's closer to 900 a day.

2 years ago

https://www.duolingo.com/jirka92122
  • 25
  • 20
  • 11
  • 9
  • 5
  • 4
  • 26

I would not give much importance to day by day numbers. Best is to take longer term averages and look for patterns there. There's also a lot of other events that affect the growth. I'm sure right after New Year's will be a small spike due to people and their new year resolutions.

2 years ago

https://www.duolingo.com/Luko.
  • 19
  • 17
  • 13
  • 10
  • 8
  • 8
  • 3

Perhaps that's why eo-es was growing 1,1k per day the first weeks of the year

2 years ago

https://www.duolingo.com/BerberuEsperanto

Probably many EO-EN moved/added EO-ES, but it progressed from 14 to 450 to ~ 1K Daily Average in last 3 months

2 years ago

https://www.duolingo.com/Duolanguaged

Probably a lot of this increase came from SCRATCH, (kid programming language) as I saw a very much seen esperanto studio.

2 years ago

https://www.duolingo.com/aleksescomu
  • 13
  • 12
  • 12
  • 11
  • 10
  • 8
  • 7
  • 6
  • 4
  • 3
  • 2

antau tri tagoj mi parolis kun samurbano, fejsbuke, kiu komencas lerni Esperanton. Li estas tre tre juna, eble 13 au malpli. Mi demandos kiel li ek-trovis esperanton :)

2 years ago
Learn Esperanto in just 5 minutes a day. For free.