An Intelligent Way to Kill Time:

15 Feb, 2010 is a place you go to play cards (aka Bridge or Contract Bridge), for those of you who aren’t familiar with the game, its basically played by four players who form two partnerships; the partners sit opposite each other at a table. The game consists of the auction (often called bidding) and play, after which the hand is scored. brings you all the good things in the popular card game most Arabia gahwa regulars are familiar with, the setting, the sounds, but most important, it brings you that hard to find fourth player that always seems to cancel last minute, it does that by allowing you to add any number of required players as AI (Artificial Intelligence machine) player called THETA. I saw one person playing with 3 THETAs instead of people, she explained: “because they’re faster than humans”.

Those of you who are familiar with similar services such as Jawaker or 6arneeb are about to leave now, but hang on, it gets better, apparently the website was built for this AI machine called THETA instead of it being the other way around, meaning the creators needed a lot of data to provide THETA in order for it to learn to play like humans, so they built the website in order to teach it!

So not only will you be participating in creating a smarter AI machine every time you and your friends play Tarneeb, but you’ll probably have fun too.

The website is straight forward, for those of you who want to jump in and play a fast game and go, you don’t have to register, in addition to it being very colorful and built with engaging graphics & characters, containing public and private chat, and a user interface supporting both English and Arabic for most of the website.

Tarneeb Game Room

We tested the site for basic functionality and ease of use, and thought we would warn you about a couple of issues, first of all Arabic, which currently, isn’t supported in the Game Room, in other words you can’t use Arabic in chat, and can only view it with English labels, although its founder assured us this issue is being currently being tested.

Another issue we had was with the meaning of some of the labels in the Game Room as you can see above, simply put, we didn’t know what they meant till we played a couple of games, because the status of a player is shown as an image with no description, not even in the Help section.

Last but not least, was weak ‘Bad Word Filtering’ in public chat, which seems to be a hard task, that with transliterated Arabic being the only means of communication (no Arabic chat support remember), in addition to the lack of filtering Bad Words in the English language in private game chat also, yet when raising the issue with the website’s management, we were told its being resolved.

In my opinion, the website is built quite professionally, basically with Flash and .NET on the front-end, and obviously with some serious potential, which led us to have a small Q/A with its Founder Mahmoud Hajo who filled us in on some things we were wondering about..

When was THETA created? What branch of Adaptive Technology is it based on?
Tarneeb co-founder Kareem Seddik and I began working on Project THETA in early 2008. The first prototype created was named THINKY, a mock-up version written in VBA and actually presented through EXCEL! We were obviously consumed by the idea of creating a very smart computer player that can master card games and learn from its mistakes. Project THETA was born!
The idea driving Project THETA extended beyond mathematics, we wanted to actually mimic human behavior in all its complexity; a very aggressive idea indeed. In order to achieve this, we had to resort to artificial intelligence and machine learning techniques. THETA stands for Adaptive Tarneeb Emulating Human Thinking (in reverse). Without going into the techy, geeky details, and uncovering the guts of the algorithms used, let me illustrate with an example: As you may know, one of the main stages of the Tarneeb game is the betting stage, during which the power suite (the Tarneeb) is determined by the highest bidder. THETA considers multiple factors when deciding how much it should bet. For example, the strength of THETA’s hand, how much THETA’s partner bet, how much each opponent bet, who is left in the betting game, THETA’s position on the table, the total score of the two teams so far, hesitation factors and more! But this is not all. After THETA homes in on the bet value, it reverts to its experience, and tries to “remember” similar situations and learn from what has happened. Based on that, THETA may decide to get very aggressive, or play conservatively this time. The ultimate goal is to make the best decision that will on the long run increase its team’s chances of winning. Now remember, THETA has infinite memory, and it uses that memory to store every move played, not only by it, but by every single human player that ever visited So far, THETA has statistical database of over half a million complete games.
When it comes to deciding which cards to play, THETA has a number of strategies that it checks every time it has to play a card. Based on its view of the world, it decides which strategy to implement. It goes without saying that playing the right card is even trickier and more complicated than betting. It is, in many cases, a matter of art not science.
So, THETA is learning with time, trying to mix it up until it finds the statistically best strategies that have the optimal chances of winning. This feedback system can in fact be viewed as a simple neural network.
To sum up, THETA is learning at a very fast pace. It has yet to unleash its real powers. Perhaps it is waiting until the site comes out of Beta. That is when THETA will take all its experience and dominate the game’s arena, so my advice to all the human players out there: BEWARE of what you do in front of THETA!
For other cool information about THETA, visit:

What’s the business model? Has a revenue stream been generated?
We have been very consumed by the notion of smart AI machines that play card games. We wanted to develop highly sophisticated machines that anyone could challenge through a simple browser. Given our passion for Tarneeb, the obvious gap in the market, and the feedback we received about Project THETA, a business plan was put together to capitalize on this opportunity. The model has then been tuned to fit the Middle East Market. The result is a solid strategy with revenues forecasted in the near future. We are still in the development stage and very excited to see the pieces of the puzzle all fall in place.

How Many users are registered?

Our website supports Guest registrations and Member registrations. We currently have about 60,000 registrations with a 1:8 member to guest ratio. The fundamental difference between the two is that members have more website features enabled. These include chatting, a private inbox, score keeping, a profile, and more freedom to interact with other users. The guest registration is a quick instant play method that allows visitors to immediately experience the game. We track these conversions closely as we add features or simplify the user interface. On a daily basis, we have hundreds of members and thousands of guests that visit us. Every day is a new record for Tarneeb, and this really gets us excited. Currently, we have surpassed the six figures in monthly visits.

When did you launch website? came to inception in March of 2009. It was a beta launch, and we are still pretty much in that mode. We had a soft launch under a different domain prior to February, but that was mainly for proof of concept.

Are you receiving any kind of funding? has been growing very organically, in all aspects. We have not spent a cent on advertizing or promoting the site, and we have not sought any external funding thus far. This aligns with our strategy and our planned capital structure. The founders have been keeping the project afloat from their piggy banks. At the right time, we will be open for that smart investment that will accelerate our road to success.

How long did the development process take? (Concept, Technology, Website)

How big was the first developing team? How big is your team today?
It is common to hear about startups that started in the basement or garage. In our case, it started on a volley ball court in Austin, Texas. The project then moved to a couch with an initial team of less than a handful. Kareem and I were very dedicated, but most importantly, it was all about having fun. All our team members had full time jobs (and for some of us we had families, music bands, or MBA programs), but we always found time for Tarneeb. Most of our development was done after midnight, or on Saturday nights (yes, we used to spend our weekends with THETA!). Today, our team has grown considerably and is continuing to do so as the project proves itself.

Do you have plans to expand?
Absolutely, this is merely the beginning… all I can say is STAY TUNED!

What is the hardest thing you faced (are facing) trying to launch the site?

We have definitely faced a fair share of challenges launching the website. As traffic started to increase, downtime naturally started becoming more expensive. Sometimes ignorance is bliss. We were obsessed with making our solution so robust and scalable that we pushed our launch date several times. Eventually, we just had to go for it. Had Microsoft decided to wait until Windows became bug free, we would have still been running on DOS J! Perhaps the hardest thing that faced our launch was to decide on when!

Any advice to starting entrepreneurs in the region?

In our opinion, the most important aspect of any start-up is the team. We cannot stress the importance of making sure you have the necessary passion, commitment, talent and trust on the team. These values will be put to test time and time again, and will ultimately determine the outcome of the business. Secondly, approach any project with a solid business plan, a sound strategy, smart goals, planned contingencies and realistic expectations. Lastly, have fun. Enjoy it — every second of it. After all, the project is your dream!

All in all, I personally think the idea of building a website to provide data that helps an AI machine learn how to play like a human (hopefully a good one too) is plain cool and smart, I hope they fix the couple of issues they have holding them back from providing a service that can compete with the other existing services in the region (especially language-wise), and inspire others.

Share and Enjoy

  • Facebook
  • Twitter
  • Delicious
  • LinkedIn
  • StumbleUpon
  • Add to favorites
  • Email
  • RSS

  1. Ahmad Hammad  |  February 15th, 2010 at 6:28 PM #

    Tarneeb is not quite bridge, its similar but bridge has much more to it.

    Nice article anyhow.

    keep it up.

  2. uberVU - social comments  |  February 16th, 2010 at 3:26 AM #

    Social comments and analytics for this post…

    This post was mentioned on Twitter by oooooozoooooone: RT @ArabCrunch: An Intelligent Way to Kill Time:

  3. Mazen Labadi  |  February 16th, 2010 at 9:18 AM #

    The site is not bad, but personally I think is much better. That’s just my opinion

  4. Gaith Saqer  |  February 17th, 2010 at 1:18 AM #

    THETA is an AI and machine learning not a bunch of javascript code so to me it is a winner 🙂

  5. göğüs estetiği  |  February 18th, 2010 at 1:02 AM #

    Great article. There is a lot of good information here, though I did want to let you know something – I am running Redhat with the latest beta of Firefox, and the look and feel of your blog is kind of bizarre for me. I can read the articles, but the navigation doesnt function so great.

Leave a Feedback

  • Become Our Fan On Facebook

    ArabCrunch on Facebook

  • Popular Posts

  • Recent Comments
    • saranraj: These crowd funding platform is an need ...
    • Farnsworth: Really?? Arab crowdfunding???? o0?? Why ...
    • Danni: Nice to know these kind of crowd funding...
    • Abdul Majeed Shoman: Hi Great article. Just want to correct...
    • Salem: That is a very strategic addition to Cyb...