Few players have risen as fast in the footballing public’s consciousness as former Geelong star Tim Kelly – going from apprentice sparky/WAFL standout to one of the best handful of players period in three short years. The story of Tim Kelly isn’t typical of the era of the One National League and increasing professionalism, but instead of State Leagues hoarding talent and parochial lines being drawn.
Note: This post is part of a series of posts using a valuation method called Player Approximate Value (PAV) to evaluate trades for fairness and balance. Elsewhere, you can read much more about the method and theory behind PAV. You can also read about PAPLEY, the projection method used to derive expected future PAVs, which has been revised for 2019. Expressing the value of players and picks in terms of expected future PAV provides a common currency for comparing them in trades and other movements.
When looking at the player values that PAPLEY spits out for every player, HPN also likes to do a manual sense check to find the nearest comparison players since 1988 to see if the projection makes sense and is likely to hold up. Over the next week you will note several occasions where we disagree (either partially or wholly) with the PAPLEY projection. Sometimes a model can’t know all the key inputs.
For Tim Kelly, only one name pops up with a strong similarity in our model – Darren Jarman. The former North Adelaide midfielder was one of the last top state -based players to hold off from going east, staying at the Roosters for for the first five years in his career. In that time, Jarman was selected by a VFL club twice, first by Melbourne in 1986 (at pick 55) and then by Brisbane as a concessional selection in 1989. After years of resistance, and alleged disagreements with Crows administration, Jarman was traded to Hawthorn for pick 10 in the 1990 draft.
Kelly’s first two years are somewhat similar at face value to Jarman’s, with selection in the All-Australian team, and holding down a critical role in a very good team almost immediately. Jarman produced 15.7 PAV in his first year and 21.2 in his second – a little lower than Kelly’s 19.2 and 23.0, but in the same range.
For the rest of his career, Jarman was able to accumulate 151.5 PAV past his age 25 season, which is the smallest possible sample of one for Kelly comparisons.
The trade
It is also worth noting that Kelly’s PAPLEY projection of 108.8 as listed above above includes a zero season – the age 23 season he played in the WAFL. As HPN doesn’t have the right statistics to calculate PAV from in state leagues (let alone produce a quality comparison chart to convert values to AFL level), it is hard to work out just how good he would have been in the AFL if he had played in it that year.
One was of doing so is imputing what he was most likely to have earned that year, based on other career arcs. To evaluate the trade with a truer projection, we’ve taken this step and Kelly’s value boosts from 108.8 PAV to 121.4 PAV – almost exactly on pick 2. Kelly is a rare case due to his late entry to the AFL and high quality, and in this situation it is fair to use this boosted, imputed value as it is more likely to predict his actual future career.
It’s also probably fair to guess that as a result of the above information that Kelly is worth the upper range of his prediction (above pick 1), and this trade indicates this. The net of picks in and out for the Eagles is about -173 PAV, or an implicit value of Kelly equivalent to pick 1 plus pick 53.
In the rawest sense, Kelly had the best season of anyone traded since Patrick Dangerfield in 2015. Chris Judd’s name has been thrown up this week, but Kelly had a better year in the season immediately before both were traded. Simply put, trades involving players as good as Kelly don’t happen often, so it’s understandable that they might demand a premium value.
West Coast have been forced to pay a fair amount for Kelly, the picks Geelong get back projecting to about 30% more than Kelly’s future output. However, the Eagles won a premiership in 2018 and will likely think they are a strong chance for a 2020 flag, this downgrade in long term value will not bother them much at all.
It also may be worth noting that the pick 24 West Coast gave up in this trade was Sydney’s second round pick, secured by a pretty good win the Eagles had live brokering Sydney’s Nick Blakey shenanigans. Essentially the pick was secured with what would have been pick 54, which also should soften the blow further.
In immediate terms, HPN predicts that the Eagles have gained 12 PAV for 2020 from this trade by turning some (likely) raw draftees into immediate production. Geelong have similarly lost value for 2020. The Cats will reportedly turn one of these picks (or a later pick) into Jack Steven, and potentially other ready made players to hit the market, boosting their prospects a little in the near term.
Essendon’s involvement is a subtle and breakeven value shift of a few places up in the 30s and down in the 50s, both areas of the draft where PAV suggests the movement of value from pick to pick is pretty flat. It is fair enough for Essendon to make this shuffle, but quite unclear which party out of Geelong and West Coast insisted on it rather than just trading picks 33 and 57 with each other directly.
Verdict: Long term win for the Cats, short term win for the Eagles. Essendon hopefully at least had fun getting involved.