New Community Challenge – share your fun Anaplan builds and collect a badge!

The only really fun non-work model I've built is a March Madness predictor! It combines my two true loves - college basketball and Anaplan 😂

The idea originated as an Excel sheet many years ago, before I used Anaplan. Every year I added on to the Excel sheet. It never got too burdensome, but I had the bright idea this past year to leverage Anaplan instead of Excel to make it easier to iterate upon as well as build in more accuracy analysis/easier data uploads.

Not to get too into the details, but the basic functionality of the model is:

1. Load in team-based data: This includes mostly statistical data. I utilize a couple different sources which evaluate all teams in college basketball (as of this last year, it was 362 teams) on a variety of stats, from simple statistics such as 3 Point % to advanced stats such as True Shooting %. I load all of these into custom-built data flats and also load in any conference changes as I have calculations based on whether teams are Power 5 (4 now), if they are a tournament team, if they are a conference champion, etc.
2. Inputs: There are a couple of manual inputs that I still put in. The most time-consuming one is my Last 10 Game adjustment - essentially I look at each tournament team's previous 10 games and assign them a number of points based off of total wins, games vs. ranked opponents, conference champions, etc. I haven't found a great data source to pull this from, but let me know if you have any ideas 😅 I also may manually adjust a team up or down a point or two, based off of off-the-court things such as coach changes, recent player injuries, scandals, etc. I try to keep these rare and only use them when I feel strongly that something will affect a team that is not already included in the stats
3. Calculations: The main purpose of the model. I use a number of ways to essentially assign each team a "point" value for each of the stats I utilize (there are 15 in total that I currently use). I start with a Multiplier, which I'll adjust year-to-year based off of Accuracy Analysis that I perform within the model after the fact. If one stat seemed to be a better predictor for which teams would do well, then I increase its multiplier. If a particular stat seemed to have a poor correlation with team success, then I decrease the multiplier. I then multiply by a point value determined by a tiering system - so putting that Level 3 Model Builder training to good use 💪 Essentially I rank all the teams, then determine a Top X % to determine various tiers, then give a certain amount of points for teams that fall into each tier. I utilize RANK, MAX, MIN, AVERAGE, and also have created my own formula to replicate the AVEDEV formula I utilized in my Excel model to determine deviations from the mean. This is where I do the most tweaking year-to-year, to try to figure out which statistics are meaningful and to try to get them to assign a properly significant amount of "points" to the teams.
4. Output: The final output is just a score, adding up all of the points which have been assigned to each team. It's pretty simple - the team with the most points is who I will be picking to win my bracket. I still need to build out the actual bracket functionality so that it can automatically populate the entire thing for me - instead, I just look at whichever two teams are playing and look up their point value and write in the higher value of the 2 on my bracket. It doesn't take too long, but there's some opportunity to increase efficiency!
5. Accuracy Analysis: The work isn't done when the bracket is fully written in 😉 Once all of the games have been played, I look back to see how I performed. I still need to flesh out this section, but what I'm essentially doing is ranking a team's performance by which round they made it to vs. what the model predicted. A team gets a "grade" for the variance, or lack thereof, between the two. From there I end up with a final accuracy %. The hope is to increase this accuracy % year-over-year. The model still isn't where I'd like it to be, but did improve last year. I think if I can get up to somewhere around 75% then I will be happy - and rich enough to retire early! 😜

Uploading transactional rows file in Anaplan

It seems a very straightforward request, but I found it quite challenging to arrive at the right solution to solve all the below requirements/constraints:

-          The file has no clear primary key and the file structure could not be changed.

-          The primary key was possible only if considered also some long text columns (like Transaction Description)

-          After data was uploaded in Anaplan, every row was needed to be manually mapped to a Transaction Type.

-          The transactions file can contain hundreds of rows every day and the file can contain multiple days.

-          The file always contains, for one Date(day), the latest data that needs to be considered for a particular Company (deletion criteria: Company-Date combination)

-          One Day(Date) transactions can be uploaded multiple times for the same Company, but only the latest uploaded rows must be considered.

-          The previously mapped rows should be maintained and only the new rows should be shown to be manually mapped for Transaction Type.

-          I was also required to trace the user that uploaded the latest data for a particular Day-Company combination.

In Anaplan was set the below :

1.       Creating 2 numbered lists with the Transaction file structure:

a.       #DL01a Transaction File – LastFile : used to delete and upload only the Last file

b.       #DL01b Transaction File – All Rows : used to merge all the files and all the rows

                        i.      Contains all the Transaction columns

                        ii.      LastFile Flag: not blank for all the rows of the last uploaded file populated from a column that is always populated (like Company or Transaction Date)

                        iii.      UserLoad: text to save the user that uploaded the LastFile. It was used to maintain the values when the Users will be setup with "No access" on the model.

                        iv.      Boolean “Consider Row?”: if true, the row will be considered in calculations

2.       Created a process that launches these actions

a.       Delete ALL rows from list #DL01a Transaction File – LastFile

b.       Import from CSV file into list #DL01a Transaction File – LastFile using the unique properties combinations

c.       Import action to populate the Company List (if new Companies are updated)

d.       Import action to clear “LastFile Flag” in the list #DL01b Transaction File – All Rows for all the rows that have “LastFile Flag” not blank.

e.       Import action to put FALSE in “Consider Row?” in the list #DL01b Transaction File – All Rows for all the rows that have the same Company-Transaction Date combinations imported in the list #DL01a Transaction File – LastFile

f.        Import action to import from #DL01a Transaction File – LastFile into #DL01b Transaction File – All Rows using the same unique properties combinations as import from CSV file action, adding values for “LastFile Flag” and TRUE for the “Consider Row?”

§  Create a module MAP01 Company-Day that has TRUE for the existing combinations in the list #DL01a Transaction File – LastFile

§  Boolean “Filter clear LastFile Data” in a module with #DL01b Transaction File – All Rows that takes from MAP01 the Company-Day combinations present in #DL01a Transaction File – LastFile and used as filter to the import action.

g.       Import action to update the Current User in the list #DL01b Transaction File – All Rows filtering the rows where “LastFile Flag” is not blank

§  Create a module with Users and #DL01b Transaction File – All Rows with property Show ALL users = OFF and filter with “LastFile Flag” is not blank for the import action.

§  If in the model exists the line-item “Current Date” that it is periodically updated, it can be traced also the Upload Date.

In the downstream calculations modules were considered only the Transaction rows with the “Consider Row?” = TRUE and the other rows were ignored.

To maintain the model clean, a periodical launch to delete the rows “Consider Row?” = FALSE can be launched.

I hope it helps if someone has similar requirements 😊

Regards,

Alex

Project Calendar Creator with Dynamic What-If Scenario Modeling

I collaborated with a private label organization to design a robust 17-scenario Project Calendar Creator, capable of dynamically updating the duration days of over 30 milestones based on selected project attributes. This tool leverages historical benchmarks to automatically calculate the optimal duration for each project stage based on attributes like project type, complexity, and resources.

Once the project attributes are selected, the tool instantly generates a tailored project calendar, reflecting the benchmarked duration for each milestone. The calendar also enables real-time tracking of the project’s progress, allowing stakeholders to monitor whether the project is on schedule or facing delays.

In case of delays or other issues, the system supports immediate adjustments, empowering project managers to swiftly reallocate resources or alter timelines to keep the project on track. Furthermore, the tool continuously evolves by incorporating data from completed projects, refining its benchmarks and scenarios, and thus enhancing future project planning accuracy.

This dynamic and iterative approach ensures that each project benefits from the latest insights, fostering an environment of continuous improvement in project management practices.

Best,

Arjun Gandhi

CTO, Master Anaplanner & Certified Trainer

Rock - Paper - Scissors Game

I built a rock-paper-scissors game where you play against Anaplan. You choose your option, Anaplan chooses its option, and the result is shown as a GIF animation.

Attached is a 20-second recording of me playing the game:

If you’re interested in the back-end details, feel free to send a message.

Hi everyone !

What an interesting challenge, hats off to everyone that posted on this challenge and put the time and effort to go well beyond their working hours.

Since I joined the wonderful world of Anaplan I have always been overworked and I have to say time managementhas always been the biggest challenge. So this thread is very impressive.

For a number of reasons I don’t really have a finished project to showcase, instead I am going to describe/draw a couple of relevant and slightly unusual applications where I used Anaplan, hoping it could spark ideas !

1 – Tracker for the Anaplan call or POC tracker

This is something I created when I was working at a previous customer for a couple of use cases. While I do not have any screenshots I can draw a simple mock-up.

• For whom ? Anaplan CoE leaders or for some POC trackers

• Why ? Do you find yourself dealing with the Anaplan Customer Success Business Partner and having to navigate with many different topics, issues, tickets ? The CSBP manage their own agenda using a standard template, which in most cases is great. But did you ever feel you needed to take more of a lead/ownership ?

•Solution: I created an app that served as a tracker to lead the agenda or the monthly/quarterly call with Anaplan for open tickets, Anaplan issues, features and POC deliverables.

I provided access to this app tothe Anaplan Customer Success Business Partner, internal colleagues and used it to track and prioritise all of the above.

At that time we were receiving special help and Support for Anaplan as we were assessing whether it was the right time to migrate from Classic dashboarding to UX. We were managing this through a POC and we really needed a common centralised repository to keep track of everything.

It was a little bit like a centralised Excel or Jira, where we were creating entries, uploading images and links to the Idea Exchange, using text line ítems for commentaries and or the standard comment and tag features and it was working very well.

This was enabling me to lead the cátch-up calls and I believe this was working well for both us at the customer and Anaplan as well.

This is a very simplified mock-up of the page I created :

2 – Model Size Estimator

This is something I started working on and never finished and that I think should/could be an app available for everyone.

•For whom ? Everyone

•Why ? We know that in Anaplan model size is a big topic and we always need to work with estimates.

I see 2 types of estimations : High level or Granular level.

High Level estimates are always needed before a project starts. For new projects there is a reasonably simple Excel spreadsheet that is circulating which is providing a basic high level estimate based on a number of factors

Granular level estimates are usually more of a BAU topic. Perhaps your model has grown a lot and you need to explore options to reduce its size or perhaps you are at the stage where you are asked for new functionality and you can/want to estimate it with more precision.

•Solution:

What I wanted to do was to create one single app that could be used for those 2 types of model size estimates. The high level estimate would have been the easier one to tackle as I could have replicated the Excel design into Excel…I know it’s not good practice but that’s all is needed for it ! 😋

What I did was to start creating the granular level estimator. I only had 0.5 days to do so (while I was consulting on a Project) and I started with something that was shared by Anaplan (thanks Paul G!). I wanted to create something that could have been reused with any models, exporting and importing the blueprint as the baseline and then having the opportunity to interact with it adding new lists, new subsets, time ranges, subsets, new modules dimensioned by any lists, and subsets and being able to change the size of any of these objects and/or central time/calendar settings.

I eventually came close to complete it but I got stuck with the integration of the subsets. I started using the APIs with postmanand Ryan from Anaplan shared helpful tips, however, I had to park this and I never really had the opportunity to pick it up again.

However, this is a random selection of the pages I created.

I hope you find this post and these ideas helpful !

Alessio

Ahoi, Anaplanners! 🌊🌊🌊🌊🌊🌊

This challenge is wonderful, thank you for bringing this up @becky.leung

I am still amazed at how creative our community can be!

So here is my project:

I've recently built the famous Battleship game in Anaplan, and it was both a fun and challenging project! My goal was to bring the classic naval strategy game to life using the Anaplan platform, complete with ship placement, turn-based play, and randomized fun facts about maritime history and marine warfare that appear with each new shot.

Starting screen

This screen serves two roles simultaneously: it functions as a welcome page, ensuring players that everything is ready for a new game, and as a settings dashboard where you can change configurations, reset the game, add new players, and more.

If you scroll down, you'll find ship properties configurations, as the rules can vary slightly in different countries:

Fleet/Ship Placement

One of the most intricate and rewarding parts of building the Battleship game in Anaplan was designing the Fleet/Ship Placement module. The challenge was to create a system that allowed players to position their ships on a grid while ensuring that no two ships could overlap and that all placements adhered to the game's rules.

This was particularly challenging as I needed to ensure that ships were placed correctly on the grid without overlaps, while also allowing for flexibility in positioning. Tests and try-and-errors were infinite 😅. I used a combination of boolean/binary logic and list mappings to handle the various ship sizes and orientations, ensuring a smooth and error-free setup phase.

User-Friendly Interface: The placement process is designed to be intuitive. Players can click and drag across the grid to select the cells for their ship, and a visual confirmation appears when the placement is valid. Once confirmed, the ship is locked in place, and players can move on to the next one.

Compromises had to be made. I initially wanted the user to place all the ships without needing to select the ship type beforehand, but the backend calculations became too complex. So, I decided to change the approach to the current one that you see now.

The "Start the battle" action is locked until all ships are placed for both players and "Ready?" status gets green.

Gameplay

Once both players have set up their fleets, the game moves into a turn-based battle mode. I implemented DCA to make the game quick and responsive. Players can easily track hits and misses, and the game automatically updates the status of each ship as hits accumulate.

In this version of Battleship, the gameplay isn't strictly turn-based in a one-by-one fashion. Instead, the sequence of moves depends on whether the previous player hits or misses. If a player successfully hits an opponent's ship, they get to take another shot. If they miss, the turn passes to the other player. All shots are logged on the bottom section

Player 1 screen:

Player 2 screen:

All shots are logged at the bottom section "Moves" to calculate whose turn it is now, display the right status, and calculate which new fun fact to show.

"Random" Fun Facts

To add an educational twist, each new shot reveals a random fact about maritime history, marine warfare, or board games. This not only keeps the game entertaining but also informative, adding a layer of trivia that players can enjoy as they play.

It's done pretty simply: Since I already have a Shots register I can match each player-coordinate combination with a Facts repository (there are 200 facts in total = a 10x10 field x2 players)

I am really happy with how this game turned out. Of course, there were some compromises but on the other hand, I have discovered really awesome things. For example, the background of the main field is sooo satisfying like it is an ocean with waves 😌

This small project was a great way to explore the flexibility of Anaplan and test my skills,and I hope it inspires others to think outside the box with their own models!

Fair winds! 💨 ⛵