The AFL assist

official AFL statistics have been revamped this year, they are much more substantial than previous years. In business an old adage is that companies measure what is important to them, stats can be seen as a weak level 'rules of the game' they certainly can be the rules of the contract.

I say this because it is my belief that play will always conform to the game. This is not so remarkable in AFL, NBA or Culing, but it is in business. It can be huge if your companies main reporting metric is 'unit sales' vs 'profitability' one will see sales managers running promotions and campaigns to push as much stock onto customers as possible, the other will see sales managers trying to cut down on inventory, advertising expenditure and closing down smalltime customer accounts.

So too it is curious to observe that the herald sun is reporting among other things 'the assist' - a stat very familiar in basketball. An assist is a pass that results in a shot being made by another player. it is hugely relevant in basketball because such passes are frequent and necessary for an effective offence.

Does it stack up in AFL though? I've gone off the assist in the NBA I used to think it was uncontestably far more valuable than points on a players stat line, but I'd been fooled by that old hidden cost, which I'd disregarded.

NBA - Kobe Bryant last years MVP averages around 25 points, 5 rebounds, 5 assists.
Chris Paul last years runner up averages around 18 pts, 11 assists, 5 rebounds. (i just made these figures up)

Now in my old mindset this meant that if you looked at the offensive value of each player then you'd convert assists into points. since an assist is only counted if the pass results in a scoring shot being taken then you could arguably say that Kobe gets 25 points per game, plus 10 points (via assists) for his team = 35 point effort.
CP3 gets 18 pts plus 22 (via assists) for a whopping 40 pts per game, offensively Chris Paul is the more prolific scorer... oris he?

factoring in opportunity cost, say that Chris paul shot 55% from down town, (3 pointers) and 80% from the field (2 pts) so each shot he takes is statistically speaking worth about 1.6 pts. Without going into what free throws and 'and one' plays would weigh in, then we must accept that for each assist Paul dishes then the opportunity cost of Paul not shooting himself is 1.6 pts. (damn I wish I'd figured out easy numbers).

Kobe is the most lethal shooter in the league, so the opportunity cost is even worse in an assist.

but if we rerun the numbers factoring in opportunity cost its nowhere near as clear cut - Kobe is worth 25 pts per plus .2 (his higher opportunity cost) x 5 assists = 26 pts. CP3 gets 18 pts on his own and .4 x 11 in assists = 22 pts.

Furthermore due to the way assists are measured we only rack up an assist if it succeeds, we can tell immediately how many points result from assists if given the assist number, (except we don't know if they are worth 3pts or 2 something else I didn't factor in) We cannot do the reverse though, and say given Kobe's 25 points we can assume 12 assists, kobe created 12 assist opportunities, because Kobe could bring the ball down himself, score off a fast break, drive the lane etc.

Furthermore whilst kobe will have a stat line of say 8-10 (8 scores for 10 attempts at field goals) 2-3 3pt 3-4 ft telling us his accuracy at shooting under various scoring sub divisions, we don't get how accurate Chris Paul is at assists, you don't see 11-12 assists from 12 attempted assists, you do get turn overs but you don't know if Chris Paul makes 11-30 passes, that is 11 assists and 30 ordinary passes. Turn overs can arise from steals, fouls, travelling and bad passes, so they don't tell us how good someone is at dishing dimes. They simply tell us who dishes out the most. So Chris Paul may get 11 assists per night to lead the league, but perhaps Steve Nash gets 8 out of the only 8 passes he made? we can't see that in the stat line.

So it's interesting to see this in footy. In one sense it makes sense, it is the midfields job to deliver the ball to a full forward or other scoring option in the forward line. At the same time, the accuracy of any given full forward is touch and go. Footy gives you a point (and is perhaps the only sport that does) because they can see what a goal kicker was trying to do.

But marks can be contested and spoilt relatively easily. A full forward like Fevola that can make explosive breaks and convert marks into goals is going to make a midfielder look far more valuable, more so than a point guard that racks up assists because they have Kobe to pass to. VS Nathan Brown kicking to Mathew Richardson, one of the most reliable, least reliable kicks in the league. Nathan Brown might be one of the best kicks for delivering a ball NBA, no look off hand/foot style to a player that is wide open. But if that player is Mathew Richardson then you can kiss that assist goodbye.

More telling I feel is the old 'Forward 50' which measures how many times Midfielders can deliver into the 'danger zone' where your team has the opportunity to score. That means a midfielder can run it in, then boot the goal himself just as easily as kicking it to Fevola on the run.

Furthermore the question of any stat that play might conform to is 'does it improve the game'. Tim Duncan is this question embodied in the NBA, the 'Big Fundamental' plays flawlessly in accordance to the rules, he almost has perfected the pussy art of soccer simulations for the NBA too. And as such whilst Duncan's effectiveness can't be questioned, the rules can because Tim Duncan is the most boring 'superstar' in the league.

AFL has been pretty good at rule changes that cause the game to be faster, rougher and more entertaining, such as the shooting clock to speed up full forwards like Fev and Mathew Lloyd, and extending the mark distance from 10 to 15 meters to prevent boring possession maintaining chipping around for hours.

Stoppage time reduces in AFL it's big selling point over NBA (the last 3 minutes of which are painful set plays and endless time outs). So will the assist, should midfielder's contracts suddenly depend upon them 'help the game'

I don't know. It could see a return to the era of the 80s/90s of big full forwards kicking a tonne every game (Brerton, Ablett, Lockett etc.) It could also see midfielders making unnecessary kick or handball passes to other players trying to rack up assists (and increase the threat of a turnover) which are more valuable to them contract wise than kicking a goal themselves for the team.

Footy is not like NBA though, the teams are bigger, the physical requirements more democratic and roles less defined. In any team their are 4 midfielders where in NBA their function would be performed by the solitary point guard. Their are usually at least 2 big scoring options and other goal sneaks available in the forward line. You have defencive specialist taggers in the midfield and then you have more scoring options in the midfield. it is far less predictable with far more 'moments of truth' in any given possession, particularly since you can tackle wih impugnity (if it's legal) whereas fouls in bball are more like credits to be spent, so possession usually results in a scoring attempt rather than a turnover which is relatively rare in bball.

All that said the AFL assist, I predict will probably just dissappear because it isn't informative. The Brownlow probably won't go to the player we suddenly discover is better at delivering the ball (to a decent kick) than anyone else in the league because the marking and goal kicking are really far more deterministic of the score than the inbound pass.