2018 trading, drafting and list management: players and personnel

[Aprilbr]

From reading the comments and thinking about the Hanners' trade, I believe that football people will focus on it heavily over the coming few years. Essentially it's a salary dump trade given the Swans got so little in return, so history will see it one of 3 ways depending on what transpires from now.

One, if he returns to his form of 2016 and before then it will be seen as a huge win for the Saints. Perhaps one of the most one-sided trades in recent history. Second, if he continues to struggle, then a brilliant trade by the Swans to ruthlessly recognize his decline and unload a player being paid elite salary to deliver mediocre performance. Third, he performs to average levels (no star but a reasonable contributor). Then considered a mutually beneficial trade maybe slightly in the Saints favour. Whatever the outcome it will be watched closely by the AFL community!

Sent from my SM-G965F using Tapatalk

[Ludwig]

Mate you've been spruiking our depth and up and coming midfield for a while. Why the sudden change in heart.

Not really a change of heart. I was just looking for an excuse to use Premiership Spoondow. I highly value the emotional outburst prior to thinking things through.

When I had a look at our current list, I still feel we're doing okay, but a lot of young prospects will have to continue to develop well. There's always an injury here and a form slump there, but it looks fine on paper.

[Ludwig]

Hi, I'm actually one of the authors here. To clarify, May's gonna be 27 in January and history says he's good for missing a half dozen games every year, so that hurts him on an accumulative valuation system. When/if he gets traded for more than that, it's a sign he's getting a premium because of breaking contract, and that the club is thinking more short term improvement to the list than longer term payoffs.

I like your system, but it can't factor in a lot of intangibles, which tends to make some of your ratings appear off the mark. It doesn't mean it's wrong. How would you rate your system based on the historical analysis of previous year ratings of players?

[Ludwig]

Here is the ordered list (Top 18) of the HPN player ratings for the Swans. I does reflect the importance of our younger players producing in the coming years. Just for reference, Will Hayward is rated a pick 3 and Heeney a pick 4.

Hayward, Will
Heeney, Isaac
Florent, Oliver
Papley, Tom
McCartin, Tom
Ronke, Ben
Jones, Zak
Parker, Luke
Lloyd, Jake
Hewett, George
Mills, Callum
Cunningham, Harry
Sinclair, Callum
Newman, Nic
Rampe, Dane
Stoddart, Ryley
Melican, Lewis
Franklin, Lance

[liz]

Here is the ordered list (Top 18) of the HPN player ratings for the Swans. I does reflect the importance of our younger players producing in the coming years. Just for reference, Will Hayward is rated a pick 3 and Heeney a pick 4.

[/FONT][/COLOR]

I know nothing of the methodology but just perusing that list makes me question the validity of the assumptions inherent in the model.

In particular, I believe that if the Swans wanted to trade either Luke Parker or Callum Mills they would receive a higher trade return than for any player above them on that list apart from possibly Heeney. I understand age factors into their residual value, but I still think another club would more happily pay a high trade value for the known, consistently high-level output from Parker than for the guessed improvement of a Ronke or McCartin. And for Mills, age cannot be an explanation given he's younger than most of those above him on the list.

And even though it's just one spot, does anyone really think Newman, a player the Swans could yet delist (or give away for next to nothing) is more valuable than Dane Rampe? Age is sometimes irrelevant.

[Flying South]

Not really a change of heart. I was just looking for an excuse to use Premiership Spoondow. I highly value the emotional outburst prior to thinking things through.

When I had a look at our current list, I still feel we're doing okay, but a lot of young prospects will have to continue to develop well. There's always an injury here and a form slump there, but it looks fine on paper.

Lol fair enough mate. I've often been accused of the same.

I keep changing my mind about our list. For me there is so much uncertainty. Will be an interesting year.

[707]

- - - Updated - - -

I have to share this quote re: Barrett from Trade Talk - AFL.com.au

"Damian Barrett is perhaps the best footy journalist in the country and his Sliding Doors column is always full of nuggets."

Shows what nepotism there is in journo ranks, his Sliding Doors is full of nuggets you flush down the toilet, Barrett is a flog!

[R-1]

That was weird, board ate my reply.

I like your system, but it can't factor in a lot of intangibles, which tends to make some of your ratings appear off the mark. It doesn't mean it's wrong. How would you rate your system based on the historical analysis of previous year ratings of players?

Yeah, any model is going to be a simplification that eliminates intangibles - as they say, all models are wrong, some are useful. In our case, we're modelling for total accumulated value in the future, with value defined as a zero-sum sytem called Player Approximate Value. Every year there's a certain amount of value attributed to players - 300 per team, assigned by key stats and scaled for team strength (so maybe 330 for good teams and 260 for bad ones). Picks and players both have an expected future output of value expressed in these terms, so we can derive the expected values of each thing and compare them. But that naturally leaves a lot out, including:

• Time preference. Clubs would prefer to get their future value at certain times. Immediately if you're a contender seeking a best-22 player (or trying to add quality to escape a rebuild), longer term if you're planning your list profile a few years ahead. There's different discount rates depending on that.
• Clubs ratings of specific drafts and ranges of picks within drafts - average valuations of picks can't reflect if a draft is deep or shallow, or the shape of talent (eg 20 even picks at the top then a drop-off, or the current 7ish elite group), or the kinds of players in it (eg, a poor year for talls), or "draft board" stuff like if there's targeted South Australian locals, or a Queensland club who thinks everyone available at a top 10 pick will just leave. And if a club isn't using a pick live, it's got very little value to them at all.
• Why players have missed games, etc. Injury, form, player quality and other things all just look like players not getting senior games and not producing value. Clubs will make sharply different assessments of worth based on what they think about the medical situation or the chances to fill a role within a team. Basically - our projections of players have their own risk based discount on value, but clubs have good reason to have quite different risk assessments.
• Anything to do with contracts, ie whether someone is in contract or not, and whether there's a cap dump or partial payment of contract or something else happening. Player salary is another trade and list planning currency (actual currency!) and it doesn't convert linearly into these valuations.
• The "eye test". Some players are, subjectively, considered more valuable than the Player Approximate Value assigned them for their actual current form. Some defenders are like that, in particular. If the starting point valuation is an underegging of how clubs are actually thinking, that'll pan out in the projections as well.

So basically, the system can tell you which side is more likely to get more value at the end of the day, based on a system built on averages for picks and player ratings derived from like 30 years of basic player data. We reckon it's a good framework to use as a starting point, a sort of default value. Then as analysts we can look at that comparison to actual club moves, and try to see what else is going on. A lopsided trade usually has an explanation in some of those intangibles.

[KSAS]

WC managed to keep their Gaff and we managed to make one! (re: what we got for Hannebery trade).

[barry]

oooh, I like this lol

I agree 100%, can't bitch nd sook, but like I said, at some point I look forward to an explanation.
One thing to want Hanners gone, another thing to have him go in such a poor deal.
But it's only a poor deal after everything is all said and done.
Best way to look at this is to look at the ins and outs after the trade periods and the draft.
That's what the club will be doing.

The swans (Harley) would not want a repeat of the Mitchell experience with Hanners.

To lose one gun in a bad trade is a mistake, to lose two is incompetence.

[Go Swannies]

The swans (Harley) would not want a repeat of the Mitchell experience with Hanners.

To lose one gun in a bad trade is a mistake, to lose two is incompetence.

I must have missed the gun performances of Hanners in 2018 - can you remind me which ones they were?

And how many players would you have moved on to match the Hawks' offer on Mitchell? Lloyd, Allir?

[Ludwig]

The swans (Harley) would not want a repeat of the Mitchell experience with Hanners.

To lose one gun in a bad trade is a mistake, to lose two is incompetence.

While the HPN trade calculator is being discussed, I thought I would throw in that if we wanted to reverse the trade today, i.e, trade Florent for Tom Mitchell, we would be slight losers. That's right, HPN values Florent just slightly higher than Mitchell. I guess that means that Ollie will win 2 Brownlows in due course.