Saturday, 25 March 2017

MilesR: Challenge VII Stats Round Up: "The Reckoning"

Rejoice my friends, for the we are about to embark upon a magical mathematically themed mystery tour of Challenge VII.  Gather round the digitally enhanced campfire and prepare yourselves for to be astonished by tales of algebraic deering-do, trigonomic terrors and captivating calculus.

First, a word of caution.  We, we happy few, we band of paint-stained brothers and sisters, stand upon a precipice of mathemagical divination.  We can choose to go forward bravely in the darkness of the statistical night or meekly cry out for my aid and hand calculators.  Anyone who fears long division may freely depart, for I would not have them in my company on this day:


Maybe a little over the top?  Nah, I think it's spot on.  Who knew that Batman was at Agincourt?

The mathematical rituals have been performed, the appropriate sacrifices to the dark arithmetical gods have been made and the results have materialized before our eyes.

The simplest answer is Challenge VII broke all the records.  No, that isn't quite right.  Challenge seven treated the records like Godzilla would treat Tokyo after stepping on a giant lego piece.  The terms squashed, obliterated, annihilated may be more appropriate.

Lets us examine some graphs (sigh).  First up compares Challenge VII to it's previous 6 ancestors


Wow - VII's 96 painting participants and 83.2K points are both new records.  VII also set record for number of submission (919), average score per Challenger (867) and many more.

Want to go into more detail on VII?  Well, we're going to do it anyway.

Lets take a gander at the score distribution

Percent of Total PointsIVVVIVII
Top 525%26%20%22%
Top 1041%42%33%35%
Top 2062%63%56%53%

It seems that Challenge VII has continued the recent trend of Challenges of being more socially progressive with the share of score for the top 20 participants declining steadily.  Somewhere Bernie Sanders is smiling.  As a staunch capitalist, I find this trend disturbing and a wake up call to re-double the nefarious efforts of my fellow "financial services professionals".  We've gotten complacent since our triumph of 2008.

Lets look at even more meaningless eye candy graph.


First up is the distribution of sales we painted:

28mm was the dominate scale again this year, but 15mm staged a nice comeback.  I am particularly proud that the unallocated portion of this years trend graph is much smaller than last years, where it was closer to 20%.  How did we eliminate the "Purple Stain"?  By bringing the snow Lord into line and getting him to enter the figure detail for the bonus rounds.  Curt may have started his Challenge hosting career as a care free bon vivant but statistics bring us all to heel in the end.  Prepare to be assimilated.

So what do 919 individual submission look like over 3 months of painting combat?

Very pretty, but not all that meaningful.  Lets try taking that same data and aggregating by day:



Interesting - we see what is referred to by professional painting challenge commentators as the "two humped" camel - a slow start then a binge in January followed by a dip and then a mad rush at the end.  That last day looks pretty hectic - there were 56 rubmissionfor 5,455 points!

Enough of this graph stuff, lets get to some data tables where the real geeking can get going.

It is my great pleasure to present to you the final KPI table of Challenge VII:

(Please note that I do grant a limited purpose non-commercial license to use the KPI table if any of you want a new tattoo to impress your friends, your welcome.) 

AHPC VII: THE STATS ROUNDUP
KPI's:
Weeks Completed24681113
Weeks Remaining1197520
Participants w/ a Submission478196969696
Total Submissions89239417540715919
Total Points7,71121,90636,63248,90563,12783,200
Participation Rate47%81%100%100%100%100%
Average per Submitter164270382509658867
Avg per Submission879288918891
Submitter % of Target Completed16%32%48%65%83%110%
Projected Points
Submitters50,12269,95078,83678,46574,13583,200
Remaining Bouns Rounds9,2616,7875,8514,2551,7500
Non Submitters @ 25%13,3581,6570000
Projected Total Points72,74178,39484,68782,72075,88583,200
Figures Painted
28MM
Inf1,1002,3393,7884,8016,1687,945
Cav / Art69155305425545719
Vehicles20557595138175
15MM
Inf1796971,3891,8062,7063,679
Cav / Art28178285361513817
Vehicles1663132167319467
6MM
Inf4791,0511,2231,4642,0402,112
Cav / Art1118108187195195
Vehicles35214322354471631
Score Distribution
28MM84%64%60%61%61%60%
15MM7%11%6%12%12%16%
6MM4%4%4%3%3%3%
Bonus Round "Bonus Points"0%11%13%9%9%12%
Other Scales4%9%17%15%15%9%
Total100%100%100%100%100%100%

So what nuggets of insight can we derive from this most impressive of data tables?  First off, the collective participants of Challenge VII achieved 110% of their initial target - that hasn't happened before.  We painted a lot of 28mm figures but 15 mm staged a comeback from Challenge VI as the table below points out:

15MM ProductivityVIIVI
Points13,4275,822
Figures Painted
Infantry3,6791,728
Cavalry / Guns817294
Vehicles467225

As with past challenges, the "Point-O-Projector" was eerily accurate in forecasting points at the mid point of the Challenge, forecasting 84,687 points during week 6 and 82,720 at week 8 versus the final score of 83,200

Week 11's abysmal forecast of 75,885 is due to operator error as the knucklehead who runs this spreadsheet forgot to adjust to tempering equation for the shorter duration left in the Challenge.

Don't you hate it when you do that?

Economic Value Add

Ok enough of this statistical mumbo-jumbo and lets get into the real protein and tubers of the statistical roundup - economic mumbo-jumbo.  What is the economic value add of this year's event?

Methodology
First a word on methodology.  This has been a controversial aspect of my statistical work as there are many opinions on the both the assumptions and methodology that have been voiced.  Rest assured, dear stats fans, that I listen very conscientiously to each of your opinions, carefully weight the pros and cons and look deep into my soul to ensure I'm doing the right thing.  I'm very proud to say that those concepts that advance whatever agenda I have are quickly adopted into the methodologies.  Any other changes that accurately contradict my agenda are discarded as fake news.  I think we can all agree that's the right way to approach both tracking miniature painting as well as setting public policy.

So how do we calculate Economic Value Add (or "EVA" as it's known on the streets)

(1) The total points of the completion is translated into the equivalent number of 28mm infantry figures by dividing by "5" - 83,200 points is the equivalent of 16,640 28mm infantry figures.

(2) We then calculate the cost of acquiring and painting the individual figure based on cost per mini, cost of paint and the value of the painters time.  I've made some changes this year  and have increased the costs for all three categories to reflect a higher cost environment.  There is still some controversy of the value of a painters time but I'm worried that raising the hourly rate any higher will just force management to replace painters with robots and we wouldn't want that would we?  The total cost per figure (see table below) is $13.50

(3) The sum total of the cost to create a painted and based mini is then multiplied by an economic velocity factor to reflect the fact that when we replace a painted figure in our collective lead piles we, on average, buy 4 figures for every one painted.  At least I do.

So what's the EVA of Challenge VII?

Ecomonic Value Add (EVA)
Total Points83,200
Equivalent 28mm Inf Minis16,640Economic
Value Add
Cost per Mini$2.50$41,600
Paints & Basing Materials$1.00$16,640
Painting TIme$10.00$166,400
Sub Total$224,640
Velocity4$898,560

Just under $900K - that's big time money and a economic stimulus package in 28mm scale.

BOOM, BABY, BOOM!

Over time, the Challenge has truly become a financial juggernaut in the miniatures industry

EVA TrendlineCummulative
IIIIIIIVVVIVIITotal
Participants6244861728896395
Total Points2,24530,56542,07959,93565,10564,69083,200347,819
Equivalent 28mm figures4496,1138,41611,98713,02112,93816,64069,564
Economic Value Created$14,211$193,476$266,360$379,389$500,000$569,272$898,560$2,821,268

Close the $3MM in EVA created to date.  Mr Campbell, I would expect some compensation form the wargaming manufactures is due to you.

The Hall of Heroes
As with every Challenge we do take some time to recognize the lifetime to date performance of our challengers - the Legendary Painteristas.  Through challenge VII there have been a total of 384 participation events, with many individuals participating multiple times.

RankNameLieftime Points# of Challenges
1DaveD15,1796
2MilesR15,0596
3RayR11,2186
4TamsinP10,3285
5MartinC9,9913
6IanW8,6926
7Millsy8,4395
8FranL8,4074
9KentG7,1752
10GregB6,9466

I'm very excited to announce that we are welcoming a new member into the 10,000 club - Tamsin!  Welcome Tamsin, your membership card and secret decoder ring will be sent to you shortly.  Please remember to use your newly granted social status only for good, not evil.

DaveD still reigns supreme as the "Points-Progenator".  It should be pointed out that MartinC is coming fast on our heels and will likely become the person to attain "10,000 Club" membership the fastest after next years Challenge.

Top ten Individual performances

RankChallengeChallengerPoints
12RayR4,940
22KentG4,724
34AndrewS4,180
45DaveD4,138
57MilesR4,125
63ChrisP4,076
77MartinC4,051
85MarkO3,854
93JamesB3,761
107GrahameH3,611

In term of top ten individual performances, we saw 3 from Challenge VII crack the top ten this year.  Sadly none of us came close nor are we likely to ever exceed RayR's majestic score of 4,940 points from Challenge II.  That's a mark for the ages and all of us need to come to grips with the fact that the Challenge's future statue in Trafalgar Square will likely have Ray's visage starting down upon us like a Elephant Seal baking in the sun wearing a speedo.

Lastly, there's one burning question that I am sure all of you thirst for the answer to - which day of the week reigned supreme during Challenge VII?

Minion# ChallengersTarget PtsActual Pts% Target Met% of Total PointsPts/ChallengerOvr/Undr Index
Monday1617,95020,398113.64%24.52%1,274.841.47
Friday1713,45017,993133.78%21.63%1,058.411.22
Saturday1414,25015,283107.25%18.37%1,091.641.26
Tuesday168,90011,369127.74%13.66%710.560.82
Wednesday1710,4509,31189.10%11.19%547.680.63
Thursday159,9508,11081.51%9.75%540.670.62
Grand Total11,00073773.70%0.89%737.000.85
Grand Total9675,95083,200109.55%100.00%866.671.00

Shockingly, Monday stand out as the best day of the week in just about every category, including self pro-motion.  Who knew?

MONDAY - for the WIN!

Monday looks cooler in a pie chart - cooler cause it's blue.


There you have it, Stats fans.  Another Challenge has come and gone but like a hot branding iron, it left a mark.  As I close this overly long and pompous post, I would like to sincerely thank a few people:

Curt - for organizing this madness and proving that Tom Sawyer was a brilliant management visionary - painting this fence sure is fun....

Sarah - for putting up with Curt and not killing him due to annoyance.  As we've discussed many times, thinking about doing so doesn't count.

My fellow Minioneers - Excellent job and somebody else take Monday - it's too much work.  Sorry about making you wade through my silly spreadsheet of doom

Challenge VII participants - thanks for making this such a fun event.  Special shout out to MatinC for pushing me as my points rival.  I do wonder if he's just playing me like at an auction - just bidding up the ante to see if I'll bite and I always do.   I look forward to our AWI Challenge next year.

Lets go make some terrain now.

Calc-On, Dudes


30 comments:

  1. OK, Miles, step away from the bell curve, put down the moving average, and keep your standard deviation in plain view.

    ...on second thoughts, I don't really need to know about any of your deviations, standard or not. Keep those to yourself.
    ; )

    ReplyDelete
    Replies
    1. I think some things are best not shared...

      Delete
    2. Hey. YOU'RE the one who suggested a commemorative statue to Ray, Painter of Thousands, Bagger (or is that Badger?) of the Sands, and Smuggler of Budgies.

      After that thought, it's time for the old Mind Bleach.

      Delete
    3. I'm liking the sound of this???

      Delete
  2. Love it, one of my favourite posts of the year. Really in depth review of the data. The value part is very interesting and I'm a little annoyed that I missed the 10k by 9 pts. If only I'd known. Thank you

    ReplyDelete
    Replies
    1. Thanks for a fantastic Challenge this year.

      Delete
  3. Excellent, I like the fact it really shows the Friday crew just aimed so low in their target to appear to be successful.

    Roll on challenge 8 ... the stock up project has commenced ... and I will have my new super productive mancave set up ready too!

    ReplyDelete
    Replies
    1. I'm not sure either Martin or I can beat you if you have 3 months to paint rather than the 2 you had this year!

      More Sudan or (gasp) a new topic next year?

      Delete
    2. That was certainly a clear-cut attempt by the Friday crew to manipulate at least one statistic - I suspect the same of the Tuesday crew ;)

      With the Sudan project near-as-dammit complete, what's the next big one? Ancient Egyptians? Classical Indians?

      Delete
    3. Hope it's something complicated but I suspect Zulus, so we are all doomed.

      Delete
    4. Oh yeah of little faith..

      It won't be .. napoleonic.. or anything made up.. I.e 40k ...

      Usuthu!

      Delete
    5. Just once Dave, I'd like to see you essay something in the SF field, even if it's only Sir Garnet Wolseley ordering the troops to form square to receive the charge of a Green Martian horde!

      Delete
    6. Come now Tamsin, such comments are beneath you - especially when Team Tuesday had at least 3 instances of challenge goals being raised in the last month!

      Delete
  4. A truly magnificent elucidation of the numerical and financial factors of the Challenge. It's a pleasant non-surprise (well, given the calibre of the participants involved one should definitely not be surprised) that Monday's Manic Mob ruled the roost :)

    I'm also pleased to find myself the 4th member of the 10k club and look forward to receiving my decoder ring - you can rest assured that it will only be used for the most excellent of purposes. Looking at the table and extrapolating various factors, I suspect that next year might very well see Dave become the founding member of the 20k club.

    The final two days, particularly the Sunday were definitely very busy - that was nearly 10% of the final points and 10% of the submissions in just two days. I think that the self-sacrificial folks who heroically stepped up and helped with minion duties deserve especial praise.

    Just a quick question - in the final economic impact table, are the data for previous years shown on a historical cost basis or adjusted to the same costs as used for this year?

    ReplyDelete
    Replies
    1. The past EVA's were not adjusted upward for the new assumptions. I could come up with a set of excuses siting improper inflation adjusting and restatements but the real reason was I was too lazy to do so.
      :)

      Delete
  5. This is a tremendous exhibition of statistical sorcery, Miles. Bravo! I'm rather dazzled with it all.

    The one stat that really shocked me is the one stating that the Challenge has generated the equivalent of over 69,000 painted 28mm figures. Wow. That just boggles my mind. Thanks for putting together this smorgasbord of stats and erudite analysis.

    Now, who would like the honour, and tremendous fun, of 'painting a fence' next winter? It's an elite club so I only take on the best. Any takers? ;)

    ReplyDelete
    Replies
    1. Both Dave D and I have painted the equivalent of 3K figures each over the span of the Challenge - that's like half his Sudan collection!

      Delete
    2. That's only the stufff that scored!

      Delete
  6. Well done Miles!! Great stats! Some i understood and others........well, I was never any good a Math? Glad to see I'm still top. My score will be beaten one day, hopefully long after the pigeons have pooped all over my statue in Trafalgar square???

    ReplyDelete
    Replies
    1. One day, archaeologists of the future will find your statue and say, "Truly, they were as gods who made this. Slightly unhinged and possibly insane gods, but gods nonetheless."

      Excelsior!
      ; )

      Delete
    2. Hmmm - I'm now inspired on the weathering for the model of Ray's statue that I'll shoehorn in on next years terrain bonus round.

      Delete
    3. Don't worry about the stats you don't understand. You can trust me as I'm a financial services professional and have your best interests at heart*

      * subject to terms and conditions

      Delete
    4. Ah, our old friend the asterisk...!
      ; )

      Delete
  7. Another spate of mathematical madness to demonstrate that trying to grasp all that data only unhinges the mind...almost as badly as trying to explain to it other evidently!

    So the most important yet uncovered issue thus far is - how does the data show that Millsy is, in fact, still a GIMP? :-)

    ReplyDelete
    Replies
    1. I think it's time to give the GIMP award a bit of a rest - Millsy's been a tremendous sport about the teasing but lets not push that envelope any further. Don't worry, I'll be thinking up a special award for someone else...

      Delete
  8. Good work friday crew. Close runners-up on total points painted and smashed the %Target Met. I'm proud of you all.

    ReplyDelete
  9. Well, if you keep this up Miles, I might actually start to like stats! Great stuff mate, thanks for keeping us posted and on our toes this Challenge what would we be without your statistic support?!

    ReplyDelete
  10. Ohh a scatter plot - what do you mean it doesn't tell us much - what new heresy is this man? Great job crunching numbers again Miles - although I may report you to the PETS (people for the ethical treatment of stats).

    ReplyDelete