2022: Week 6 - Solution

Solution by Tom Prowse and you can download the workflow here. 

This week we take a look at something a little different compared to The Prep School challenges. This challenge is a bit of fun and we look at the likelihood of scoring big with a 7 letter word in Scrabble. 

Step 1 - Input Scrabble Scores

The first table to focus on is the Scrabble Scores sheet. This contains all the information about the scores and tiles within a game of Scrabble. However, all of this information is within the same field therefore we need to split this out so that we have three individual fields instead. 

To do this we can use the custom split functionality to first extract the Points. We could do a split on the ':' symbol, however to make things a little easier we can just return anything before the first occurrence of the word 'points:':

We can then make this a whole number and rename to Points. 

Next is the tile and frequency information. This is a little harder but we want to use a custom split again, but this time return everything after the ':' symbol. 

This has the opposite effect compared to the last split, as it removes the Points information and leaves the rest of the string. 

From here we can parse the each of the letters out by using another split, but this time splitting all values by a ',':

At this stage we should now have 10 different split fields with a range of different letters and their frequency. 

Now to make our lives easier, and so that we don't have to go and create a split on each of these fields, we can pivot the data so that all of the split fields are in a single column. Here you can use the wildcard columns to rows pivot, by using 'Split' in the wildcard search: 

Now they are in a single column, we need to make sure that we remove any spaces so that we can then split by the 'x' value. This time the automatic split should work and we should be left with two new fields - Tile and Frequency. 

Our table should now look like this: 

Step 2 - % Chance of Drawing a Tile

Next, calculate the % chance of drawing a particular tile by using the calculation Frequency / Total Number of Tiles. 

First, we need to calculate the total number of tiles by using a LOD calculation: 

Total Tiles

Then we can use that to calculate the % Chance, rounded to 2 decimal places: 

% Chance 

ROUND([Frequency]/[Total Tiles],2)

After these calculations our table should look like this: 

Step 3 - 7 Letter Words Input

Now it's time to turn our attention to the 7 letter words table in our input. After inputting the table into our workflow, the first calculation we need to make is to calculate the length of each word using the Len function:

Length

LEN([7 letter word])

From here we want to split out each of the letters into a separate fields. Unfortunately, you can't do this natively with Tableau Prep, therefore we need to include our scaffold to help us out.

After inputting the scaffold table, we can then join this onto the 7 letter words table using Length >= Scaffold:

This will allow us to have each word repeated from 1-7 (7 rows). Therefore, we can now use the Mid function to extract each of our letters based on it's position and the scaffold value that we have just joined on. Eg, 1 from the Scaffold will be letter 1, 2 from Scaffold letter 2, etc. 

Letter 
MID([7 letter word],[Scaffold],1)

After this calculation our table should look like this: 

From here, we can now count how many occurrences each letter has in each word. Within the aggregation we want to Group By 7 Letter Word & Letter, then Sum Number of Rows: 

After the aggregation our table should look like this: 

Step 4 - Combine Workflow Branches

Now we've prepared both tables, we need to take with the 7 Letter words branch is to join this onto our first branch. To do this we can join on Letter = Tile:

Note we don't need to make these the same case as Tableau Prep will figure it out!

After the join we now have a single table that looks like this: 

Step 5 - Update % Chance

We can now update our % chance of drawing a tile based on the number of occurrences that it appears. Following the requirements we can use the following calculation to do this: 

% Chance

IF [Number of Occurrences]<=[Frequency]

THEN POWER([% Chance],[Number of Occurrences])

ELSE 0

END

This will either return a 0 or will multiply the % chance by itself based on how many time its occurs within the word.

Our updated % chance field now looks like this: 

Step 6 - Total Score

To calculate the total score that each word would score, we first need to multiply the points by the number of occurrences: 

Score

[Points]*[Number of Occurrences]

Then we can use a LOD to total the points across each word: 

Our table should now look like this: 

From here we can then pivot the data, so that we have a new field for each letter, using a rows to columns pivot and then Sum % Chance: 

Then finally we need to tidy the null values by using the following calculation:

% Chance 

IIF(ISNULL([A]),1,[A])

*IIF(ISNULL([B]),1,[B])

*IIF(ISNULL([C]),1,[C])

*IIF(ISNULL([D]),1,[D])

*IIF(ISNULL([E]),1,[E])

*IIF(ISNULL([F]),1,[F])

*IIF(ISNULL([G]),1,[G])

*IIF(ISNULL([H]),1,[H])

*IIF(ISNULL([I]),1,[I])

*IIF(ISNULL([J]),1,[J])

*IIF(ISNULL([K]),1,[K])

*IIF(ISNULL([L]),1,[L])

*IIF(ISNULL([M]),1,[M])

*IIF(ISNULL([N]),1,[N])

*IIF(ISNULL([O]),1,[O])

*IIF(ISNULL([P]),1,[P])

*IIF(ISNULL([Q]),1,[Q])

*IIF(ISNULL([R]),1,[R])

*IIF(ISNULL([S]),1,[S])

*IIF(ISNULL([T]),1,[T])

*IIF(ISNULL([U]),1,[U])

*IIF(ISNULL([V]),1,[V])

*IIF(ISNULL([W]),1,[W])

*IIF(ISNULL([X]),1,[X])

*IIF(ISNULL([Y]),1,[Y])

*IIF(ISNULL([Z]),1,[Z])

After this calculation we can remove any unneeded fields and our table should look like this: 

Step 7 - Filter & Rank

The final steps are to filter out the words with 0% chance and then rank the remaining words. 

First we can use a filter to exclude 0 from the % Chance field

Then we can create two rankings, Likelihood and Points Rank: 

Likelihood 

Points Rank

After completing both of the ranks, we are ready to output our data! Our output should look like this: 

You can download the full outputs here. 

After you finish the challenge make sure to fill in the participation tracker, then share your solution on Twitter using #PreppinData and tagging @Datajedininja, @JennyMartinDS14 & @TomProwse1

You can also post your solution on the Tableau Forum where we have a Preppin' Data community page. Post your solutions and ask questions if you need any help!