Hi Guys,
didn't know quite where to post this. Here is the latest post from my newsletter:
you can check out my other posts here:https://breakinglines.substack.com/
In 2011 the Los Angeles Angels signed Albert Pujols to what was then one of the richest contracts in baseball history. Pujols had put together a flawless first 11 seasons of his career, collecting 3 MVP awards and 2 World Series Titles along the way. Despite being age 32 the acquisition of the future Hall of Famer was a major coup. Or so it seemed.
After joining the Angels Pujols’s production almost immediately began to decline. Soon his contract would be considered one of the worst in all of Major League Baseball. As the graph below shows, once he joined the Angels his WAR began to decline in unison with his rapidly rising salary. His production dwindled to that of a replacement level player leading fivethirtyeight to label him the worst player in baseball. At the time he was the sixth highest paid player in the league.
When a player is eligible for free agency, we know there is a risk of The winner’s curse. There is also a tendency to value a player based on their historic performance. The risk is that their past performance may not reflect their future production. This can lead to overpaying a declining player. The Angels thought they were getting an MVP calibre player. Instead they ended up paying heavily for a player whose production could’ve been provided by a reserves league replacement. This is a list management catastrophe.
Developing models for projecting player performance may provide teams with a degree of alpha relative to their competition. Simply using the “eye test” is not sufficient and open to subjectivity and bias. Combining the two approaches would provide teams with a powerful edge with list management decisions. Such models would be instructive for teams trying to determine what a fair contract offer for a player may be moving forward. This potentially minimises the risk of overpaying for a player who is about to commence their decline.
Team data scientists could collate historical data of a particular position, for example inside midfielders and perform an analysis on how such players progress over their career. Variables could be identified as predictors of an imminent decline in performance. For instance, it’s possible tackles or clearances or “hard ball gets” are an indicator for future decline in performance at a certain age as they are likely correlated to extra wear and tear on the body. Perhaps specific surgeries prove to be good indicators. With a large enough sample size it’s possible such markers of future performance could be identified. Teams could then look at a player’s performance data and correlate it to multiple past players who played the same position and tracked similarly in terms of specific variables (tackles, possessions, clearances, kilometres run) and look at how their career progressed as they aged. Whilst such models would contain significant margins of error, they may provide a useful guide to future performance.
I put up a poll on Twitter last week asking people to vote on which player they would rather recruit:
The dilemma for clubs is the potential to overpaying for a younger player with growth left in him versus paying for an established player whose best days could be behind him.I wonder if the Adelaide Crows’ reluctance to re-sign Brad Crouch is a result of their concerns in how he is likely to perform over the term of his future contract. The potential compensation pick on offer has also played a part in their decision making. It’s possible however they have done a projection analysis and determined that the club would be overexposed in offering the lengthy deal Crouch was seeking. I suspect a healthy Crouch will perform very well for St Kilda over the next season or two. What will his contract look like relative to his performance during his age 30 and 31 seasons though?
Essendon reportedly offered a lengthy contract to Jye Caldwell for what has been quoted at around $450,000-$500,000/yr. The consensus seems to be this is a gross overpay by the Bombers. My counter to this concern would be that Caldwell is likely to grow into this contract. If he develops into the player that many project him to be then it is likely that whilst the contract may appear to be expensive now, it may end up being a very team-friendly deal soon thereafter.
The AFL average player salary is around $350,000-$400,000/yr. If all Caldwell becomes is a best 22 midfielder, then his contract will still be around fair value for the Bombers. If he develops into one of Essendon’s better midfielders then the Dons may have got themselves a bargain at the back end of his contract. There is obviously a risk still as it’s possible that he does not develop as expected.
The narrative a week ago was that Collingwood may be looking to offload Adam Treloar in order to facilitate some key re-signings and potentially recruit Jeremy Cameron. Cameron has now committed to Geelong and reports have suggested that the Pies are still potentially looking to move on from Treloar, perhaps for reasons not related to his salary. He reportedly has a heavily back-ended contract that runs through till the end of the 2025 season. It would be very easy for a List Manager to be seduced by Treloar’s numbers. His midfield production is Pujols-like in terms of elite volume. But is he on the precipice of an Albert Pujols like decline? With significant hamstring injuries in recent seasons, any team looking to recruit Treloar should project what his value add to the team will look like not just next season but 3-4 seasons from now.
didn't know quite where to post this. Here is the latest post from my newsletter:
you can check out my other posts here:https://breakinglines.substack.com/
In 2011 the Los Angeles Angels signed Albert Pujols to what was then one of the richest contracts in baseball history. Pujols had put together a flawless first 11 seasons of his career, collecting 3 MVP awards and 2 World Series Titles along the way. Despite being age 32 the acquisition of the future Hall of Famer was a major coup. Or so it seemed.
After joining the Angels Pujols’s production almost immediately began to decline. Soon his contract would be considered one of the worst in all of Major League Baseball. As the graph below shows, once he joined the Angels his WAR began to decline in unison with his rapidly rising salary. His production dwindled to that of a replacement level player leading fivethirtyeight to label him the worst player in baseball. At the time he was the sixth highest paid player in the league.
When a player is eligible for free agency, we know there is a risk of The winner’s curse. There is also a tendency to value a player based on their historic performance. The risk is that their past performance may not reflect their future production. This can lead to overpaying a declining player. The Angels thought they were getting an MVP calibre player. Instead they ended up paying heavily for a player whose production could’ve been provided by a reserves league replacement. This is a list management catastrophe.
Developing models for projecting player performance may provide teams with a degree of alpha relative to their competition. Simply using the “eye test” is not sufficient and open to subjectivity and bias. Combining the two approaches would provide teams with a powerful edge with list management decisions. Such models would be instructive for teams trying to determine what a fair contract offer for a player may be moving forward. This potentially minimises the risk of overpaying for a player who is about to commence their decline.
Team data scientists could collate historical data of a particular position, for example inside midfielders and perform an analysis on how such players progress over their career. Variables could be identified as predictors of an imminent decline in performance. For instance, it’s possible tackles or clearances or “hard ball gets” are an indicator for future decline in performance at a certain age as they are likely correlated to extra wear and tear on the body. Perhaps specific surgeries prove to be good indicators. With a large enough sample size it’s possible such markers of future performance could be identified. Teams could then look at a player’s performance data and correlate it to multiple past players who played the same position and tracked similarly in terms of specific variables (tackles, possessions, clearances, kilometres run) and look at how their career progressed as they aged. Whilst such models would contain significant margins of error, they may provide a useful guide to future performance.
I put up a poll on Twitter last week asking people to vote on which player they would rather recruit:
The dilemma for clubs is the potential to overpaying for a younger player with growth left in him versus paying for an established player whose best days could be behind him.I wonder if the Adelaide Crows’ reluctance to re-sign Brad Crouch is a result of their concerns in how he is likely to perform over the term of his future contract. The potential compensation pick on offer has also played a part in their decision making. It’s possible however they have done a projection analysis and determined that the club would be overexposed in offering the lengthy deal Crouch was seeking. I suspect a healthy Crouch will perform very well for St Kilda over the next season or two. What will his contract look like relative to his performance during his age 30 and 31 seasons though?
Essendon reportedly offered a lengthy contract to Jye Caldwell for what has been quoted at around $450,000-$500,000/yr. The consensus seems to be this is a gross overpay by the Bombers. My counter to this concern would be that Caldwell is likely to grow into this contract. If he develops into the player that many project him to be then it is likely that whilst the contract may appear to be expensive now, it may end up being a very team-friendly deal soon thereafter.
The AFL average player salary is around $350,000-$400,000/yr. If all Caldwell becomes is a best 22 midfielder, then his contract will still be around fair value for the Bombers. If he develops into one of Essendon’s better midfielders then the Dons may have got themselves a bargain at the back end of his contract. There is obviously a risk still as it’s possible that he does not develop as expected.
The narrative a week ago was that Collingwood may be looking to offload Adam Treloar in order to facilitate some key re-signings and potentially recruit Jeremy Cameron. Cameron has now committed to Geelong and reports have suggested that the Pies are still potentially looking to move on from Treloar, perhaps for reasons not related to his salary. He reportedly has a heavily back-ended contract that runs through till the end of the 2025 season. It would be very easy for a List Manager to be seduced by Treloar’s numbers. His midfield production is Pujols-like in terms of elite volume. But is he on the precipice of an Albert Pujols like decline? With significant hamstring injuries in recent seasons, any team looking to recruit Treloar should project what his value add to the team will look like not just next season but 3-4 seasons from now.
