Scoring Anomalies – Round 18

There were some improvements I had been hoping to make to the system this week; unfortunately, I haven’t had the time, so those will have to wait until next week. In the meantime, here’s what SyntheticCoach had to say about the scoring on the weekend:

First up, Dustin Martin:
Synthetic: 83 (93.5) – actual: 104

It’s worth looking at the full score breakdown here:

effectiveKicks: 10 -> 40.0
effectiveHandballs: 14 -> 21.0
clangers: 4 -> -16.0
        of which freesAgainst: 0
groundBallGets: 6 -> 27.0
goals: 1 -> 8.0
behinds: 0 -> 0.0
marks: 3 -> 6.0
        of which contested: 0 -> 0.0
        of which intercept: 0 -> 0.0
tackles: 1 -> 4.0
freesFor: 0 -> 0.0
hitoutsToAdvantage: 0 -> 0.0
goalAssists: 1 -> 3.5
spoils: 0 -> 0.0
otherOnePercenters: 0 -> 0.0
total: 93.5

Critically, there are a couple of stats that are missing from the SyntheticCoach formula, because we don’t have access to them: handball receives (1.5 points each) and gathers from hitouts (2 points each). Martin had 17 uncontested possessions that did not come from a mark on the weekend, so there’s a fair chance that a good chunk of those came from handball receives. Furthermore, he also had 5 clearances, so there could definitely be some gathers from hitouts in his score. With this in mind, it’s quite possible that SyntheticCoach is underestimating his ‘correct’ score.

Another possibility is that he was given favourable scaling – 581 metres gained and 6 score involvements is likely to have such an effect.

All this should be a good reminder to us that although it is an exciting system, and a lot of progress has been made, there is still a lot that SyntheticCoach doesn’t know about. So just keep that in mind when looking at its output.

Moving on to the GWS/Collingwood game, there were a couple of scores that jumped out:

Lachie Whitfield
Synthetic: 97 (118) – actual: 119

Jack Crisp
Synthetic: 84 (102.5) – actual: 66

This was a very high-scoring game from a point-allocation standpoint (Grundy got a raw SyntheticCoach score of 206!), so someone was always going to lose out when it came to scaling. The game was practically over at quarter-time, so that is no doubt what CD focussed on.

Moving on, everybody’s favourite burn-man got a few more points than SyntheticCoach expected:

Shaun Higgins
Synthetic: 78 (84.5) – actual: 98

With 13 non-marked uncontested possessions, it’s quite possible that there were some +1.5 handball-receives in there, to which SyntheticCoach was not privy. And with 637 metres gained and 6 score involvements, it’s also possible that he got favourable scaling.

In yet another high-allocation game for the week, anyone who owns Hawks Ben McEvoy or Daniel Howe (draft league, perhaps?) can feel slightly cheated:

Ben McEvoy
Synthetic: 142 (159.5) – actual: 126

Daniel Howe
Synthetic: 140 (157) – actual: 121

The Hawks won the game, so if CD was going to scale somebody down, you would think it would be the other team, but this doesn’t seem to have been the case, with most Geelong scores lining up well with their synthetic ones.

Finally, in a more easily understood case of unfavourable scaling, Jack Macrae and Josh Dunkley were both slightly short-changed on the weekend.

Jack Macrae
Synthetic: 132 (149) – actual: 120

Josh Dunkley
Synthetic: 142 (159.5) – actual: 127

Once again, this was a high-allocation game; unsurprisingly, players from the losing team generally lost-out.

Were there any scores from the weekend that you would like broken down? Let me know in the comments below!

---

Leave a comment / Scroll to bottom