September 2018 update on my fountain pen database

It’s been a while since I put on the nerd-hat on and talked about my pen database and I expect there are tens of you just waiting to hear what mad method I use now to select my next pen, nib and ink for rotation.

I should go back to square one to set the scene for what is to come.

One of the things that sets me apart from the average fountain user is that I use a database to select my next pen, nib and ink (PNI) combination to come into rotation.

I am a Mac user at home so the relational database application I use is FileMaker Pro, of which I am still using version 16 (version 17 is available but due to cost and lack of enhancements that appeal to me I am holding out till version 18).

I am quite the archivist so I have backups of all my fountain pen database versions dating back to early 2011. Sometimes I open an older version just to look what variant of logic I was using then to choose my next PNI.

My current selection method is thus:

For the longest time the pen was the starting point for my algorithm but these days I rotate through three selection methods, pen as starting point, nib as starting point and lastly ink as starting point.

What defines the pen, nib or ink to use? The algorithm for this is as straightforward (i.e the least number of variables) as it has been in years.

For each pen, nib and ink I calculate a historic days between times in use value for a maximum of the last six times in use. I also use a current days since last used value (DSLU) to give a maximum of seven values.

I then calculate the average and median of each of these sets of values and derive a final single value as the highest of either current DSLU, the average or the median.

The above image is the top section of my pen calculation table. D0 is current DSLU, D1-D6 are the historic days between uses, AvD is the average of D0-D6, MnD is the median of D0-D6 and AD is the single final figure. Way is the selection method that was the highest. As you can see the current days since last used score dominates the score for the top six pens.

Nibs tend to be more variable with what value dominates. The top two are based on median, as is the fourth, with third and fifth being based on average. It is not until the sixth item, current DSLU is the value.

The reason for this is that until recently there was a mismatch between the number of pens that would use these nibs and the number of nibs available to be used and so the historic times between use are across the board higher than standard for the M250 model of nib. It especially affects the median calculation and I expect to have rapid repeat use of some of these nibs till the lower values start to take effect.

Inks are exclusively dominated by DSLU (D0). Because of this dominance I added a factor to the ink selection to try to counter this. I gave each ink a number between 1 and 7 to represent the days of the week (DOW) with 1 being Sunday and 7 being Saturday. When I select an ink for rotation I only select from the pool of inks that have a DOW number that matches the day of selection.

You may notice that the ink table is currently dominated by Noodler’s ink and/or vibrant ink colours (reds, oranges, pinks and, violets) this is because I have recently decided not to count historical uses of inks in pens that were not Pelikan. This decision have basically nullified quite a lot of uses of my most “dangerous” inks.

The concentration of these inks at the top of my ink list will exacerbate the repeat use of my M250 nibs as many of these are used in M200 series pens which are the only ones I will allow to use with the more “dangerous” inks.

Currently I calculate (you don’t need to see the details) the next seven pen, nib and ink combinations to enter rotation and display them in a one to a page layout.

I select the first entry as the next combination to use but I like to see what is coming up. Not all will eventuate as listed as items do reorder as time passes.

I think that is enough for now.

Current rotation methodology

As a matter of interest my current rotation methodology is seven pens each in rotation for 14 days.

Reasons for this methodology:

  1. My favourite pen case, the Franklin-Christoph Pen Roll Seven in Napa leather holds seven pens.

  2. It means I change a pen out every two days which allows me a one night on and one night off from composing a blog post.

  3. 14 days in use allows ten working days so that each of my pens gets an even chance of being used at the office where by far the largest amount of fountain pen use occurs.

  4. Seven pens is a good in use compromise for me, enough to allow a range of nib widths and ink colours but not too many that I feel choice paralysis.

A new method of selecting pen, nib, and ink combination

I seem to be constantly thinking about the algorithms that drive the pen, nib and ink (PNI) selection process in my fountain pen database. I had always considered the pen to be the starting point for the process. I wanted to challenge that assumption so I rewrote my algorithms to provide three PNI options - one with a pen as the starting point, one where a nib was the start and lastly where an ink was the start.

I randomly* select the Priority method to use on a particular day by interrogating the last digit of a millisecond timestamp taken during the startup script of the initial time I open the database each day. 

Option 1 is Pen Priority

Screen Shot 2018-08-12 at 7.29.21 pm.png

My current Pen option and the one randomly selected for today  - M620 Piazza Navona with M600 14K fine nib, inked with Diamine Blue Black.

Option 2 is Nib Priority

Screen Shot 2018-08-12 at 7.30.44 pm.png

The M101N 14K medium nib is driving this option - this nib is limited to M101N gold trim and M120 model pens.

A quirk with this method lies with the selection by nib above which requires a pen that is way down the list for rotation. Pelikan M120 Iconic Blue is currently number 32, so not due back another month. It should even itself out in time.

Option 3 is Ink Priority

Screen Shot 2018-08-12 at 7.31.14 pm.png

In this case the Ink Priority and the Pen Priority are selecting the same combination of PNI. This due to both the pen and ink having the same rating for pen/ink use.

While I was at it I also challenged the assumption that the Days since last used (DSLU) value should be the main contributor to the score that pen, nib and ink starts with before I add modifiers to determine the next item into use.

I had previously created tables that record the duration of out of use between times in use of my pens, nibs and inks - this equates to a historic days since last used for each occasion.

Screen Shot 2018-08-12 at 7.32.46 pm.png

I had analysed this data by recording the average and the median values for each. (A bit of maths definition here - average is the sum of the values of the items in the sample divided by the count of those items in the sample, median is the value of the middle item of the sample, calculated by sorting the values and then recording the value that is at the position of count/2). Difference between two values gives an indication of how skewed the sample is.

Screen Shot 2018-08-12 at 7.33.52 pm.png

The real question was how many uses of history is appropriate for each type of item. A part of my normal stats for my database I record the median times used (TU) for each of pen, nib and ink so why not use this as the number uses of history to select. The current TU values are pens 14, nibs 12, and inks two. So I decide to bring into the history table the current DSLU score and include that with the number of historic figures to get an average and median score. I tested this new method and found that using the simple average or median score of all would make the current DSLU score meaningless. I played around with a few algorithms and finally decided that the current DSLU had to continue to play a direct role.

Screen Shot 2018-08-12 at 7.33.33 pm.png

So the end result is that I have replaced the current DSLU with current DSLU plus two other possibilities depending on the individual DSLU vs median value of the entire category of item. So what is the median value of the entire category? Well its another stat that I track. Currently for pens it is 53 days, nibs 60 days and inks 378 days. So my current formula to determine the end result to sort of replace DSLU is thus.

If the items current DSLU is less than the median DSLU for that category the score it gets is the minimum of current DSLU, Average DSLU or Median DSLU, correspondingly if it is greater it is the maximum of current DSLU, Average DSLU or Median DSLU.

The result of this calculate seen in the below pen table as the field AD15 only makes up ¾ of the total (sCalc) the rest is made of a value that represents how it rates in usage in the last 12 months (DU12) and a value that represents how it rates in use with the seven pens in currently in rotation over its last seven times in use. Yes I like the number seven.

Screen Shot 2018-08-12 at 7.34.39 pm.png

For nibs I run with the calculation as 100% of the total.

Screen Shot 2018-08-12 at 7.35.51 pm.png

For inks I add three other values to the result of historical usage calculation, the first being a value that represents how it rates in usage in the last 60 months, secondly I have a rating relating usage by ink brand and third a rating that relates to usage by ink colour. These last two are attempt to balance usage between brands and colours.

Screen Shot 2018-08-12 at 7.35.23 pm.png

All in all I am happy with this new direction of pen, nib and ink selection. Time will reveal if there are some biases in my algorithms that I have not foreseen.

*I truely do not know if the millisecond timestamp does give me a true random number.

New nib vs ink writing sample template

On the weekend I created a new MS Word writing sample template which I will use to document my nib and inks as they pass through my pen rotation. I print the sample template out on A4 sized 52gsm Tomoe River paper and then scan the results to get the image below. 

SNI027025_20180308.jpg

 

Out of interest I did a calculation of how many nib and ink combinations exist in my collection and sadly found that at my current rate of pen rotation it would take me 163 years to get through them all.

Pen, nib and ink rotations 2017

Whilst I continue to look back at the year that was 2017. I thought I would share some stats around my pen rotations for the year.

The simplest number is that I used 213 different pen, nib and ink combinations during the year.

Pens

By pen manufacture it was Pelikan 191 and the remaining 22 comprised of Waterman 16, Sailor 2, Aurora 1, Bexley 1, Lamy 1 and Stipula 1. From May I was Pelikan only.

Looking strictly at the Pelikan pen models, the use ratios mirror my ownership ratios. The most used was my M600 series pens where 24 pens were used 79 times. Most pens got used between 3 and 4 times during the year.

Inks

By Ink manufacturer it was Noodler's with 52 then Robert Oster Signature with 37 and thirdly  Diamine with 33. I used inks from 23 different manufacturers during the year. All usage charted below.

InkManChart.png

By Ink colour. Again it trends with the total number of inks I own of each colour.

InkColourChart.png

In 2018 I intend to exceed the 213 pen rotations of 2017 as I am changing pens out on a 10 day usage timeframe (rather than 12 days) which will give me around 253 rotations. This is mainly to reduce the time it takes to get around my ink collection which now stands at 438 bottles.

Fountain Pen Database revisited

I last blogged about my fountain pen database in July 2015. I thought now over two years later would a good time to revisited the subject. The 2015 post is here. Rereading that post I was surprised how much had changed in the algorithm I use to determine the pen, nib and ink combination to use. I seemed to be always tweaking the algorithm (its almost a hobby in itself).

An interest note from the 2015 post - I said my collection at the time was 90 pens, 116 nibs and 279 inks - my current collection is 56 pens, 65 nibs and 426 inks! A 147 ink increase in 26 months...

Though many details change some things remain the same - the primary factor for the pen, nib and ink is still days since last used (DSLU). I could simply use this value to select my next pen each time but that is a bit boring so I have endeavoured to spice things up.

So how do I currently calculate things?

For a pen the base value (sCalc) is derived from the DSLU value modified by:

  • subtracting the number of days this pen has been in use in the last 12 months (DU12M).
  • subtracting the number of pens of the same model already in use (PIU).
  • subtracting a calculated value that relates to number of inks that have a higher rating than highest DSLU pen (IRS).
  • adding a calculated value if this pen is able to use the current highest DSLU nib (TNBS). 

This is the simplest calculation for deriving highest rated pen I have had in years. The sCalc formula is expressed in the database as: 

Round(If(Status =1 or Status =3 ;-999;DSLU - DU12M - PIU - IRS + TNBS ) ;0)

The pen selection table

FPSS_20171001_PC.png

For a nib to calculate its base value the DSLU is modified by:

  • subtracting the number of days this nib has been in use in the last 12 months (DU12M).
  • subtracting the number of nibs of the same model already in use (NIU).

The sCalc is expressed in the database as: 

If(Status = 1 or Status =3;-999;DSLU ) - DU12M - NIU

The nib table

FPSS_20171001_NC.png

For an ink to calculate its base value the DSLU is modified by:

  • subtracting the number of days this ink has been in use in the last 5 years (DU60M).
  • subtracting the number of inks by the same manufacturer already in use (IIUM).
  • subtracting the number of inks of the same colour grouping already in use (IIUCC).
  • adding a calculated value if this ink has been used less times than the number of years it has been in my collection (DOYTU). 

The sCalc is expressed in the database as: 

If( Status=1 or Status =3;-999; Round(DSLU + DOYTU - DU60M - IIUM - IIUCC ;0))

The ink table

FPSS_20171001_IC.png

What do the colour fills in IC column signify? They are simply a representation of the far right column in the table above. Dark Green is rating 1 (the safest), light green is a 2, the yellow a 3, orange a 4 and not shown is red for a 5 (Two inks score a 5 - the two Noodler's Baystate inks)

Next pen, nib and ink combination table

FPSS_20171001_Next1.png

To derive the next pen, nib and ink combination I take the top ranked pen (Pelikan M620 Piazza Navona), select the three top ranked nibs that can be used with it and then select the top four inks that can be used with each combination of pen and nib. In the database the outcome of these series of calculation is shown in the above screenshot (I have only included the screen of the nib I used - a M600 14K Medium). You will notice that the two top ranked inks (Noodler's American Eel Red Rattler and El Lawrence) from the previous ink table are missing from the selection screen, Why? They have a rating number (4&3 respectively) greater than the pen rating of Pizza Navona (2) so are banned from use in this pen.

Currently in rotation table

FPSS_20171001_PIU.png

As you can see from the above pens in use screen I chose to not select the top ink suitable (Lamy Blue-Black) but rather the second one (Noodler's Walnut). Why? Well I had just inked up Pelikan 4001 Blue-Black in the M815 Wall Street LE and it and the Lamy are almost identical and a brown ink looks good in the semi-transparent body of the M620 Piazza Navona.

I have left out a lot of detail of how 'the sausage is made' because the post would be more tedious than it is already. I hope this gives some further insight into my pen database madness. :)

A pen database and my gamification of pen rotation

I am an outlier in the world of fountain pen users in that I use a database to algorithmically select which pen, nib, and ink combination should next enter rotation.

Why? Well, my day job is very much IT focussed and I get to "play" with databases a fair amount of the time so it was not unusual that in 2010 when I came to dealing with my burgeoning fountain pen collection I turned to a technological solution to manage it. To balance my strictly PC/Microsoft work world I am a 100% Apple product user in my personal life, so the solution I chose revolved around an Apple product - Filemaker Pro. Unfortunately this is a business-grade database solution and with that comes with a business grade price (I just updated to version 14 for the princely sum of AUD$250). In hindsight I could have achieved a similar result by using a spreadsheet application especially to record data around my purchases but I think I would struggle to get the same benefit on algorithmic selection of the next into rotation. My current collection of 90 pens, 116 nibs, and 279 inks give rise to a very large number of possible combinations. 

The Database

The image below is my In-use page from my database. Yep only three pens in rotation at a time.

The same three on the one per page version just for added bling...

So how does it all work?

Well each of my pens, nibs, and inks are assigned an UID (unique identifer) for me it is a 3 digit number starting at 100 i.e my first item is 101 etc. 

One of my writing samples images below showing the UIDs 156, 166, and 381. The other three values are the date in DD/MM/YYYY style (Sorry, but not really sorry, US readers) :)


I have master tables for each of the pens, nibs and inks where I record data about them. Data recorded for pens include all the normal purchase fields like when bought, from whom, price paid. I also record particulars about the pen such as what nib came with it, new or used and an image of each pen. I also give each a rating in the range of 1-5 (where 1 are my precious, expensive pens and 5 are ones that are for any use) to allow me ban it from being used with certain inks (the reason for this will come apparent later).

A fragment of my pen master table.

The nib table is similar with purchasing data and image, though data about the nib itself is expanded to include whether it is factory or a custom grind and rating is given out of ten to include my feelings about the writing quality of the nib. 

A fragment of my nib master table

Lastly inks get pretty much the same general fields as pens and nibs but have some other fields relating to colour such as scan of each ink's swab and image of the bottle with label and most importantly they also get a rating 1-5 of how dangerous this ink maybe to my pens. Why? - Well I own inks that have been known to stain - Noodler's Bay State Blue tops this list so it gets a rating of 5, 1s are normally only given to inks from manufacturers that also make expensive pens (my theory is that they are unlikely to sell an ink that will cause damage to their pens), so 2s are allocated to safe inks from non-pen manufacturers, 3s and 4s are given to ones with super saturated colours or special formulation from non-pen manufactures - lots of my Noodler's ink fall into this category. 

A fragment of my ink master table.

So for a pen to be used with an ink the pen rating has to be equally or higher than the ink rating or if you look at it from the ink side an ink has to have a rating equal to or less than the pen. Those inks that fail this test will never be used with that pen.

This is going to get way more nerdy but stick with it. :)

So I have my master tables for each of the pen, nib and ink. I also have a table that records the history of each combination. The includes fields holding date into use, date out of use, status, duration, comments on the combination etc. Currently this table holds 864 records dating back to October 2010. 

To allow me to use an algorithm to propose future combos I created a few other reference tables. One holds pens and nibs which lists all 1,136 valid combinations of my pens and nibs , another is pens and inks which lists all 25,389 valid combinations of pen and ink, and lastly a tables that holds the  32,643 valid combinations for nibs and inks. These tables extract data from my main history table to show if a combination has been used, if so, how many times, and how long ago.

I have three other reference tables to how the current status of my pens, nibs and inks. These tables also act one of the building blocks to get to the next combinations.

The major factor determining items for the next combination into use is the DSLU (days since last used) value of each of the pens, nibs and inks. Originally I used DSLU as the sole factor but that became a bit boring as my pens would return to use in pretty much the same order.

So begain a lot of tinkering with other factors to give some variation to the DSLU order, I won't go into the long and much more boring history of what I tried. I wil simply (LOL) explain what I do now to gamify my rotation.

I try to apply a weighing factor so I don't get too much of any one ink colour or items for the same manufacturer of pen or ink into use at the same time.

Current I do this by looking at the last 17 combinations into rotation and applying a ranking where the oldest is a 1 and the three in use at all given 17, with those in-between ranked along the way.

I sum these scores and writing the totals to three other reference tables (Colours, Ink Man, Pen Man). I then subtract these score from the maximum of those scores so that I get a range of positive values (SC column).

I then link these scores (SC) back to the respective current status tables, using Pen Man for the pens and the other two for the inks. Within these tables I apply a range of other 'fudge factors' just for futher variation. These include adding the number years owned and also a number out of 12 that respresents how close the item is to its anniversary of entering my collection. 

I am constantly trying to strike a balance between those pens and inks that have been in my collection a long time and those that are new. Lately I have noticed that my Pelikan pens were being overly penalised because they make up the single largest portion of my pen collection so I have added a factor to the pen calculation to give a bonus if DLSU is greater than 250 days. At the moment most Pelikan's return to use on average after 262 days compared to 191 days for other makes. Inks however are taking more than 600 days to return to use, and this extends by 30 days with each new ink I purchase. 

Calculations happen and the proposal for the next pen, nib and ink combination page is displayed below. Actually it is three pages as I currently select only one pen, but allow display of the top three nib choices each with the top four ink choices.

So the next pen is my Lamy Studio Violet Limited Edition with top nib choice of a stainless steel oblique medium nib, with Noodler's Dostoyevsky as the top ink. Interesting enough Dostoyevsky, an ink that has had one previous use has pushed ahead of the Organics Studio ink that has yet to have its first use. The reason for this is most likely found in the ink manufactures table where Organics Studio has a score (SC) of 22 and Noodler's of 49. The 27 difference is enough to overcome the bonus a never used ink gets in the calculations.

So that's it in a nutshell :)

Thank you if you have made it to here. You probably think I am mad and you aren't the first...

If you think that it all seems way too controlled and lacks free-will, you again will not be the first. I sometimes agree, so now and again I disregard my database suggestion and ink up a combination of pen, nib and ink that I really, really want to try.