How time flies! I decided I would try to give monthly updates in February and it's already May!

There's a lot cooking at Advanced Sports Logic. Here's what I'm giving an update on today:

• Fantasy Tyrants
• The DFS Optimizer: Current Version & New Version
• The Taguchi Brain
• My e-Business Solution
• The Machine for Full Season & Dynasty Leagues

Fantasy Tyrants

In the near future ASL will launch its Fantasy Tyrants premium content. Fantasy Tyrants will give you an edge in full-season and daily fantasy baseball and football, in-season and off-season. It will do deep dives and give the best advice for starters, emerging players, strategy and so on. Keep an eye out for it, and be sure to check it out. ASL will make some articles free so you can get a sampling of the depth and quality of the Fantasy Tyrants content.

The DFS Optimizer: Current Version & New Version

The Current DFS Optimizer

The ASL team continues to improve its current DFS Optimizer. We have tightened our software quality controls and by tracking our 60-day exception count since last December and have reduced issues you might experience by 94%.

More improvements are on the way:

• More frequent projection updates
• Faster operation

The DFS Machine

But that's not all. We have heard your feedback and the feedback of top daily fantasy sports players and are working to unleash a whole new level of DFS playing experience - The DFS Machine.

The DFS Machine is designed to give you every possible control to complete in Daily Fantasy Sports. At the top level, you have access to detail-rich, top-expert strategies and then you can peal down the levels of the strategy building blocks into the fine details of how the optimizer works under-the-hood.

We are leaving no stone unturned in this new version. We have worked out the math to calculate positive and negative cross-correlation effects between players on the same and opposing team. So, you will have the power of probability on your side and can even know the probability a lineup will be above a particular fantasy point threshold, and how the other lineups in the contest will move above or below the same threshold based on ownership estimates.

Besides more powerful features, the user interface is getting a bit of a facelift. Here are some preliminary raw wireframes.

...

The Taguchi Brain

Dr. Genichi Taguchi, was a Japanese engineer and statistician that formulated and refined "the Taguchi method" from the 1950's to 80's to improve the quality of manufactured goods. His technique can be applied to large data systems with an infinite multi-dimensional solution space to find the most stable configurations that produce high quality results. The magic of the Taguchi method is to identify a very small subset of the total solution space to find the best solutions. It's like intelligence for finding a needle in a haystack.

ASL's Taguchi Brain was prototyped in 2021 and after running for eight days converged on a solution for NBA DFS. The ASL team has only picked up this project again recently. The first step in building the Taguchi Brain is building a robust historical dataset, and that is exactly what we are working on now.

Once the data set is in place, we will make ASL's prototype live again, and teach it the solution space of the full feature set of coming The DFS Machine. Once live and productized, ASL's customers will have the power of "the brain", rounding out ASL's full solution for Daily Fantasy Sports.

My e-Business Solution

ASL invested a lot to build a complete online presence system. A lot more is required than just setting up a website. To sell products online, one must have a product marketing flow, a check out and subscription management system, a newsletter system, a content management system, integration with Google Analytics, and so on.

Since some of our white label partners for our apps didn't have some of these systems on their website, we offered to help, and this developed into a service. We are now polishing out this service, to setup more quickly and completely out-of-the-box with the hopes we can enable other small businesses to get off the ground faster and easier, both, in and out of the fantasy sports space. We already manage one partner's website, but the ASL team is hard at work to be ready for our first beta-customer of the new full service solution.

The Machine for Full Season & Dynasty Leagues

I know this is the update many of you want most of all.

The team has made more progress on The Machine since my last update. Now, in additional to all the high level functions and APIs being defined, all the low level math functions have been coded. The Machine has 13 underlying very special math functions. They are so special, we don't even name the functions, we just number them.

For example, Function 9 takes massaged team win record probability data and coverts it into playoff seed probability data:

Those that remember The Machine will remember the graphs inside The Machine that showed this information. The first graph above shows the probability each team will obtain a specific win records sufficient for making the playoffs and the second graph shows the probability that each team will be a specific seed into the playoffs.

And those that remember The Machine will also remember that we show the convolution of the fantasy point probability distribution of your starting lineup for each week. Now, we not only have the convolution function, but also have added positive and negative correlation to it. If you have used The Machine before, perhaps you will appreciate this graph:

In this diagram the fuchsia and blue are two convolutions of the fantasy point probability distributions two groups of players in a lineup. The yellow is the regular convolution of those two inputs, giving a total fantasy point probability distribution for all the players together in both groups.

However, our new math enables us to add cross-correlation effect. When an offensive player is in the same lineup with a defensive player on the opposing team, that creates negative-correlation. The better the offensive player does, the worse the defensive player does and vise versa. This limits the ceiling but it also raises the floor. A dominate team in a match up will want to reduce risk of a low floor and will benefit from negative-correlation, but an underdog team benefits from positive correlation. For example, batters on a baseball team have positive correlation. When one batter gets on base, another batter gets one more at-bat, so the performance of the batters tends to rise and fall together, creating greater upside, but risking lower downside.

The light blue curve shows 100% negative correlation, making the distribution narrow. The orange curve shows 50% negative correlation, and the purple curve shows 50% positive correlation, making the distribution wider, giving it more upside.

Now this math will be baked into the new version of The Machine when it arrives. The previous version of The Machine was found to increase a user's probability of winning by 2.7x.

With this enhancement, are goal is to get to a 50% win rate in a 12-team league in which no one else is using The Machine.

We still have a long way to go before its ready, but we are making progress. The remaining steps include:

• Build the high-order functions/APIs, such as the season calculator, the different recommendation engines, etc.
• Design the user interface and make the wire frames - for those that loved the old user interface the new one will be similar and equally rich, but more modern and intuitive - (Don't worry - we aren't going to Microsoft Word on you!)
• Code the user interface
• Integrate with league providers
• Integrate with projection providers

I hope this update was helpful and interesting. Stay tuned, there is much more to come!