Scoring Anomalies – Round 12

Written by The Salamander on June 11 2021

Last week, I said the task for this week was to try and weight some real scores, using the ‘who did what when’ data I now have access to. That data is laid out a bit differently to the usual, non-time-series data I feed into SyntheticCoach, so there are a couple of things I need to rejig before I can get that to automatically produce raw scores based on that data, but I can still put scores together manually and go from there. It’s a little tedious, but perfectly doable.

I’ve picked out Fremantle’s Luke Ryan and Andrew Brayshaw, at the top second-bottom of the above Scaling Anomalies chart respectively, since that gives me a useful starting point: one got weighted up, and the other down, so if I can find a set of weightings that works for both, I’ll (hopefully) be on to something.

Here’s how the game went:

The Fremantle vs Western Bulldogs score worm.

The first task here is to divide the game up into chunks, so that each can be given a particular weight. The two obvious ways to do this would seem to be quarter-by-quarter, or everytime the score changes. The former is less work, so I’ll start there and see how things go.

The task is to find weightings for each quarter such that, when the weighted raw scores for the game are scaled to 3300 points, Brayshaw is at 120 points, and Ryan is at 133. There are some very slight differences in the raw scores produced by the two different sets of data – the usual data gave Brayshaw and Ryan 163 and 133 respectively, while the time-series data has them at 165 and 134.5. This is because the two data sets label some things quite differently, so I clearly have a little bit of work to do to get them to have equal definitions of things. But it’s close enough for the time being.

Once the raw scores are weighted, they will be multiplied by the game’s multiplier (3300 / total raw points) to get the scores for the match to sum to 3300. SyntheticCoach has the multiplier for this game at around 0.84, but the real number is probably slightly different, as it will be based on 3300 divided by total weighted raw points, but it shouldn’t be that far off. In any case, you would need all the weighted scores for the match to calculate what that multiplier should be, so we’ll stick with 0.84 as being a reasonable approximation for now.

This means we need to find weightings that get Brayshaw and Ryan’s weighted raw scores to around 143 and 158, respectively.

The Bulldogs kicked away in the last quarter, so the logical place to start is by weighting that quarter down. Giving it a weighting of 0.66 gets Brayshaw’s score right to where it needs to be (he was prominent in the last quarter), but it doesn’t do anything to get Ryan’s raw score up where it should be:

Q1: 1, Q2: 1, Q3: 1, Q4: 0.66 – Brayshaw 144 (almost spot-on), Ryan 126 (way too low).

Adjusting all three of the other quarters up is clearly not right:

Q1: 1.25, Q2: 1.25, Q3: 1.25, Q4: 0.66 – Brayshaw 169.5, Ryan 153

Weighting Quarters 2 and 3 up and leaving the first down is a bit better, but still not right:

Q1: 1, Q2: 1.25, Q3: 1.25, Q4: 0.5 – Brayshaw 148, Ryan 143

It might be time at this point to take a look at when each player was most prominent. For Ryan, that was the second quarter. Since he got weighted up significantly, that suggests that the second quarter was weighted highly:

Q1: 1, Q2: 1.5, Q3: 1, Q4: 0.5 – Brayshaw 144.5, Ryan 149

That’s getting us in the right direction, but it’s still not enough. Pushing the weighting for the second quarter even higher gets us closer still:

Q1: 1, Q2: 1.75, Q3 1, Q4: 0.5 – Brayshaw 150, Ryan 163

But it’s not until we weight the third quarter down slightly, and the fourth a little further down, that things start to really take shape:

Q1: 1, Q2: 1.75, Q3: 0.9, Q4: 0.4 – Brayshaw 140, Ryan 157

That gets the numbers very close to working, at least for these two samples I’ve picked out, but weighting the second quarter highly but not the third? Looking at the worm above, I’m not sure I buy that. It’s as close to the truth as I’ve been able to get this week, but I’m not convinced we’re quite there yet.

It would seem I do, yes.

But there’s always next week. In the meantime, you can look up any player you like in the table below to see how their score was affected by scaling on the weekend; for an in-depth breakdown of their scoring, see the spreadsheets attached at the bottom of this post.


Name Club Raw Linear-Scaled ActualDiff
Callum Coleman-JonesRICH96.5768610
John NobleCOLL72.559634
Liam BakerRICH95.5759217
Lachie SchultzFRE4538424
Tom DoedeeADEL96.57876-2
Harry McKayCARL4352
Max LynchCOLL725846-12
Hugh McCluggageBL108.59187-4
Daniel RioliRICH12.5105-5
Jordan RidleyESS10280800
Jack SilvagniCARL74607111
Bayley FritschMELB6151598
Mitch HannanWB83.57063-7
Jack HigginsSTK1351071125
Andrew BrayshawFRE163137120-17
Adam CerraFRE89.57567-8
Noah BaltaRICH104.58274-8
Paddy DowCARL8367714
Hunter ClarkSTK32.52624-2
Oscar ClavarinoSTK5342453
Charlie SpargoMELB89.5758813
James BellSYD0000
Zac BaileyBL128107103-4
Mitch CrowdenFRE796652-14
Brandon StarcevichBL76.5647713
Aaron NaughtonWB87.573818
Callum L. BrownCOLL66.554617
Brayden SierCOLL33.52721-6
Zac LangdonWCE117.59594-1
Liam RyanWCE133.510812820
Tobe WatsonFRE16.514184
Tim EnglishWB91.57769-8
Tom De KoningCARL57.547481
Jordon ButtsADEL4234439
Josh DaicosCOLL1199694-2
Nathan MurphyCOLL564537-8
Oscar McInerneyBL73.56156-5
Luke FoleyWCE84.56867-1
Nick CoffieldSTK50.540411
Liam HenryFRE62.552575
Keidean ColemanBL6252520
Sam WalshCARL164.5133124-9
Bailey J. WilliamsWCE1311110
James RowbottomSYD108.586893
Rhylee WestWB16.514140
Bailey SmithWB645446-8
Lachlan ShollADEL65.55351-2
Max KingSTK4334351
Chayce JonesADEL66.55445-9
Sam WicksSYD50.54036-4
Jack PetruccelleWCE35.529323
Luke EdwardsWCE75.561632
Isaac QuaynorCOLL88.572742
Joel AmarteySYD78.562664
Matthew CottrellCARL2923318
Will HamillADEL28.52321-2
Ned McHenryADEL96.57874-4
Brayden HamESS86.568735
Tom McCartinSYD1128880-8
Nick MurrayADEL0000
Xavier O'NeillWCE705750-7
Riley Collier-DawkinsRICH44.53527-8
Liam StockerCARL8065716
Riley ThilthorpeADEL8770766
Kysaiah PickettMELB74.5627311
Harry SchoenbergADEL675450-4
Oskar BakerMELB0000
Harrison PettyMELB68.55753-4
Trent BiancoCOLL90.573752
Trent RiversMELB89.57573-2
Deven RobertsonBL1089076-14
Jaxon PriorBL6958613
Luke JacksonMELB756360-3
Caleb SerongFRE9277847
Jack MadgenCOLL7359623
Ronin O'ConnorADEL231915-4
Harry EdwardsWCE109.58983-6
Caleb PoulterCOLL1229989-10
Harrison JonesESS22.51817-1
Cody WeightmanWB79668014
Justin McInerneySYD7257592
Ryan ByrnesSTK57.545483
Chad WarnerSYD47.538391
Josh TreacyFRE85718817
Logan McDonaldSYD5342420
James JordonMELB93.578824
Matthew OwiesCARL5847558
Jay RantallCOLL22.51814-4
Nik CoxESS1269995-4
Archie PerkinsESS4838380
David MundyFRE130.510997-12
Eddie BettsCARL8065727
Shannon HurnWCE120.59892-6
Grant BirchallBL847068-2
Lance FranklinSYD11591954
Marc MurphyCARL7863707
Paddy RyderSTK130.5103101-2
Bachar HouliRICH11087958
David MackayADEL65.55346-7
Jack RiewoldtRICH66.5526311
Jarryn GearySTK74.55950-9
Scott PendleburyCOLL135109108-1
Josh P. KennedySYD7962642
Shane EdwardsRICH90.5718312
Dayne ZorkoBL1461221220
Ed CurnowCARL645242-10
Cale HookerESS96.576804
Trent CotchinRICH886966-3
David ZaharakisESS6652553
Rory SloaneADEL12198980
Nic NaitanuiWCE1451171214
Daniel RichBL11697981
Jordan RougheadCOLL1088780-7
Taylor WalkerADEL1361101166
Stefan MartinWB2924262
Dylan GrimesRICH927371-2
Jake MelkshamMELB352927-2
Steele SidebottomCOLL1391131174
Travis ColyerFRE68.55747-10
Michael WaltersFRE584942-7
Liam JonesCARL133.510897-11
Brad SheppardWCE4.54-1-5
Steven MayMELB116.5971069
Jack ReddenWCE13310896-12
Taylor DuryeaWB1301091123
Michael HibberdMELB67.556626
Josh ThomasCOLL86.57068-2
Dane RampeSYD90.57165-6
Mitch RobinsonBL7663641
Max GawnMELB171143138-5
Josh CaddyRICH84.56762-5
Dion PrestiaRICH12296993
Dyson HeppellESS106.584884
Jimmy WebsterSTK907165-6
Luke ParkerSYD117921019
Alex KeathWB7361621
Tom LiberatoreWB110.5931029
Andrew GaffWCE137111110-1
Jamie CrippsWCE74.560633
Jack DarlingWCE82.567670
Dustin MartinRICH122.5971025
Josh BruceWB635352-1
Tom McDonaldMELB12210211210
Ryan LesterBL58.549512
Nat FyfeFRE72.561709
Will Hoskin-ElliottCOLL104.585916
Brodie SmithADEL1441171192
Jack NewnesCARL67.555561
Robbie FoxSYD5947481
Brody MihocekCOLL4637458
Harry CunninghamSYD13210497-7
Elliot YeoWCE1119087-3
Rory LairdADEL185.5150144-6
Lachie NealeBL12710684-22
Lincoln McCarthyBL63.55348-5
Hayden CrozierWB10891943
Jarryd LyonsBL157.5132128-4
Jamie ElliottCOLL1441171203
Sebastian RossSTK114.590999
Jack CrispCOLL151122120-2
Jake StringerESS60.54844-4
George HewettSYD113.590933
Andrew PhillipsESS11994973
Anthony McDonald-TipungwutiESS60.548491
Jason JohannisenWB49.5425715
Nathan WilsonFRE1169789-8
Marcus AdamsBL71.560622
Tom HickeySYD15712413612
Brad CrouchSTK188.5149141-8
Luke DunstanSTK1098679-7
Joe DaniherBL86728816
Lachie HunterWB117981013
Kamdyn McIntoshRICH745848-10
Tim MembreySTK137.510912819
James StewartESS65.55239-13
Nick VlastuinRICH8567714
Zac WilliamsCARL108.58810315
Paul SeedsmanADEL15312413612
James AishFRE978171-10
Caleb DanielWB123.510499-5
Nathan BroadRICH614838-10
Mason WoodSTK19.515161
Jake LloydSYD135.51071081
Adam SaadCARL1219897-1
Jack MacraeWB153.51291290
Sam DochertyCARL102839411
Bradley HillSTK115.59182-9
Patrick AmbroseESS32.52622-4
Jack MartinCARL806561-4
Jack SteeleSTK195154140-14
Blake AcresFRE927765-12
Dom SheedWCE133.510899-9
Jack BillingsSTK60.54843-5
Christian SalemMELB81.56864-4
Alex Neal-BullenMELB7966682
Callum WilkieSTK1261001022
Marcus BontempelliWB1731451494
Nick HindESS1007975-4
Jake KellyADEL5141487
Daniel McStayBL4336360
Reilly O'BrienADEL172139137-2
James HarmesMELB1078986-3
Nic NewmanCARL10787903
Christian PetraccaMELB160134130-4
Angus BrayshawMELB7563696
Jayden LaverdeESS755956-3
Jake LeverMELB51.543441
Darcy MooreCOLL9073818
Marc PittonetCARL927573-2
Jackson NelsonWCE82.56759-8
Callum Ah CheeBL74.562631
Alec WatermanESS6249578
Connor BlakelyFRE5143430
Shane McAdamADEL95.577803
Isaac HeeneySYD66.553552
Kyle LangfordESS149.51181268
Tom BarrassWCE117.59581-14
Darcy CameronCOLL5847536
Charlie CameronBL75.56353-10
Patrick CrippsCARL102.58381-2
Rory LobbFRE5344451
Aaron FrancisESS2923252
Jack GrahamRICH157.5124117-7
Jason CastagnaRICH123.5971058
Zaine CordyWB59.55046-4
Brayden MaynardCOLL114.5931007
Zach MerrettESS180142137-5
Anthony ScottWB8168757
Jayden ShortRICH104.58278-4
Roarke SmithWB5445494
Jordan DawsonSYD1159189-2
Bailey WilliamsWB685756-1
Dan ButlerSTK67.55348-5
Jake AartsRICH116.59289-3
Dougal HowardSTK7358679
Marlion PickettRICH1038178-3
Sam WeidemanMELB42.535350
Darcy TuckerFRE1028680-6
Darcy ParishESS212.5168166-2
Tom ColeWCE70.557570
Rhys MathiesonBL17.5157-8
Jacob WeiteringCARL103.58482-2
Callum MillsSYD137108100-8
Ben KeaysADEL149121109-12
Eric HipwoodBL78.566682
Shai BoltonRICH151.51191190
Mabior CholRICH13210493-11
Jordan De GoeyCOLL85.569701
Jayden HuntMELB534434-10
Jack SinclairSTK132.510599-6
Harris AndrewsBL137.51151172
Matt GuelfiESS27.52219-3
Jake WatermanWCE74.560611
Clayton OliverMELB10991943
Bailey DaleWB11294962
Tom PapleySYD93.574806
Brennan CoxFRE7765672
Mason RedmanESS108.586959
Will HaywardSYD7962675
Andrew McPhersonADEL4032386
Ryan GardnerWB33.528302
Josh RothamWCE987975-4
Griffin LogueFRE716054-6
Andrew McGrathESS12.5109-1
Oliver FlorentSYD947473-1
Taylin DumanFRE83.57067-3
Josh BattleSTK47.53837-1
Sean DarcyFRE138116112-4
Tom HighmoreSTK1915172
Will SetterfieldCARL1159382-11
Luke RyanFRE13311213321
Jamaine JonesWCE74607414
James RoweADEL37.530322
Darcy FogartyADEL735958-1
Tom SparrowMELB91.57675-1
Zac FisherCARL715853-5



Raw: SuperCoach points, added up according to the formula put out by Champion Data, before any scaling.

Linear-Scaled: Raw points, multiplied by a multiplier (3300 / total raw points for the match) to get the scores in each game to add up to 3300.

Actual: Actual SuperCoach points, as awarded by Champion Data.

Diff: Difference between actual and linear-scaled points. In practice, this is how favourably or unfavourably a player was scaled – this is ultimately what I’m setting out to investigate.


Download the full breakdowns:

Round 12 Full Breakdown (Stat-By-Stat)

Round 12 Summary (General Categories)


Leave a comment / Scroll to bottom

8 thoughts on “Scoring Anomalies – Round 12”

  1. Great right up, I know many people post quarter by quarter SC scores. Would it be worth using the those scores to help determine scaling each quarter?


    1. I hadn’t actually thought of that… it could definitely be worth looking into. CD seems to scale throughout the match and the re-do it afterwards, so I’m not quite sure how far it will get me, but it could be a good angle to poke around from. Thanks!


  2. F*cking knew Neale was robbed! Thank you, Sal!

    Swear his stats were worth atleast a ton.


  3. Great article and I genuinely applaud you for trying to decipher the riddle that is supercoach scoring but I think that you might be hunting unicorns in a 5th dimension.

    I’m no statistical expert but I am mildly autistic with borderline savant qualities when it comes to looking at data. Try running your weighting data alongside supercoach ownership in 6 game blocks. I can see a pattern



Leave a Reply

Your email address will not be published. Required fields are marked *