Momentum. It's something that you're almost guaranteed to hear mentioned at least once in any sporting commentary, and it's a topic that continues to pique my interest. Whether or not it exists remains a legitimate subject for debate, but detecting it, if it does exist, has been elusive.

MoSHBODS has tipped the Tigers to win by about 16 points on Saturday, but Giants supporters shouldn’t be too concerned about that, because MoSHBODS’ record in Grand Finals since 2000 isn’t particularly impressive, as you can see from the table below.

In the previous blog, we created a quantile regression model that allowed us to estimate, in-running, a home team’s victory probability, and to create in-running confidence intervals for the home team’s final margin.

We evaluated that model based on a variety of performance metrics calculated using a 50% holdout sample from the original data set, which included games spanning the 2008 to 2016 period.

But nothing really measures a model’s performance better than a completely fresh data set from a non-overlapping time period, and in this blog we’ll be running the same metrics, but for games spanning the 2017 to 2019 period (up to and including the first week of the 2019 Finals). That’s 616 games entirely unseen by the model.

I’ve created in-running models before, for the projected final total of a game in progress, as well as for the projected final margin and probability of victory.

For today’s blog I’m going to revisit that earlier model I built to project the final margin and estimate the home team’s probability in-running, with a view to being clearer about how the model was built, and how we can assess its efficacy.

A few months back I had a first look at incorporating player data into predictive models, and found that we could knock about 0.4 points per game off the mean absolute error (MAE) of game margin predictions across the 2011 to 2018 seasons by valuing players solely on their Super Coach (SC) scores.

In today’s blog post, the fourth in a series that started with this one, we’ll take the self-organising map that we’ve been using in Parts 2 and 3 and rework it to provide one answer to the question of how many distinct position types there are. The AFL Ratings site implicitly posits 7 distinct types, but the data might suggest otherwise.

This is the third in a series of blog post that started with this one where we created a self-organising map of current AFL players based on publicly available game statistics, and overlaid the positions to which those players had been assigned on the AFL Player Ratings site.

There seems to have been some interest in yesterday’s blog post where we created a self-organising map of current AFL players based on publicly available game statistics, and overlaid the positions to which those players had been assigned on the AFL Player Ratings site.

The dynamic and free-flowing nature of AFL, along with the wide-ranging abilities of some of its players, can make it difficult to categorise any single player as strictly and always playing in a defined position.

For quite a few years now I've used essentially the same methodology to analyse the AFL Fixture for the upcoming season (see this post from 2015, and, probably more relevantly, this one from the year before if you'd like the details).

With the 2018 AFL season completed, we’ve now a little time to think about how team and player data might be combined into a predictive model, and to estimate how much better such a hybrid model might perform than one based on team data alone.

In an earlier post, we built a predictive model for game margins using teams’ individual and shared experience data (gathered from the Fitzroy R package) to supplement the MoSHBODS forecasts.

Back in 2012 I investigated the question of whether or not the number of days' rest that a team received in the Finals had any significant effect on its subsequent performance.

This year's Finals scheduling has made this topic highly relevant again, so it seems timely to update that analysis to include data from the intervening years and to incorporate some of the improvements I've made to estimating team ratings in that same period.

Some of you will already know about the fantastic Fitzroy R package, which makes available a slew of match and player data spanning the entire history of the V/AFL. A number of people are already doing interesting things with that data, and this will be the first in what I expect will be a series of blog posts from me doing my own investigations and analyses with that data.

Back in October of 2017 we looked at the then-upcoming season of the AFL and estimated the strength of schedule for all 18 teams based on the MoSHBODS Ratings and Venue Performance Values (VPVs) that prevailed at the time.

For some time now I've opined that I don't like accuracy as a forecaster performance metric because it's "too lumpy", and that I prefer, instead, mean absolute error.

Today we'll use player data from the fitzRoy package in R to investigate the first names of players in the 15,402 games completed to the end of the first week of the 2018 Finals.

I've been talking here on MoS and elsewhere about doing some analysis of player data for quite some time but, until now, have lacked a key ingredient for that analysis (viz, the data). That's just changed.

For the past four years now I've used essentially the same methodology to analyse the AFL Fixture for the upcoming season (see this post from two years ago, and, probably more relevantly, this one from the year before if you'd like the details).

Since the first VFL Grand Final of 1898 (there was none in the competition's first year, 1897) there have been 120 more, 117 of them providing an outcome and three finishing in draws and necessitating replays.