Love the poll idea. I gotta say, though, some of the contracts in these groups are not only not bad, they're exceptional. Hathaway, Coleman, Trenin, Pelech, Tanev are all undervalued by traditional points/TOI measures because of their defensive impacts, for example.
I love people disagreeing with me. That’s why I’m having a vote. Only about 1 in 20 visitors is registering a vote, but the contract nerds subscribing to my feed are highly educated in this field. If you want to make a case, you’re welcome to argue in the chat. The thread is open to free subscribers. https://substack.com/chat/1094421/post/6d11c868-3247-4db8-bf3f-67b295408710
I’m the first to admit that my modelling undervalues defensive players, but it also uses the variables most strongly correlated to actual salary. If there was a reliable defensive variable with a steady correlation to salary, I’d consider adding it. I came close to adding a hits modifier at certain production ranges, but there are also cheap hits available at a discount every summer. Just because there is correlation in some ranges, doesn’t mean spending that money is a good idea. Also, I had trouble getting to 60 bad contracts given the projected cap increases, some just raised red flags. Tanev it’s just age + term = worried. He was worth what they paid him this season for sure.
Right right. According to how contracts are modeled, these are over/undervalued deals. I suppose that's a major benefit of using statistical models that measure defense and underlying performance. They should hopefully provide a better measure of player impact than what front offices currently use. Given your modeling, these good and bad contracts are likely right in line with how NHL teams value guys.
My Expected Free Agent Value model is designed to predict what players will actually get on the open market. It hates the Slavin contract but I did not include it in the tournament. Every summer I predict what everyone’s contract will be (I’d love to have a friendly competition with anyone else who does the same) and the EFA model is my primary influence. If you can point me towards a defensive input that correlates strongly to salary, I’m open to suggestions. There are certainly players who get a premium for defensive impacts, but there’s also lots of cheap replacement level players every summer who play strong defensively.
Love the poll idea. I gotta say, though, some of the contracts in these groups are not only not bad, they're exceptional. Hathaway, Coleman, Trenin, Pelech, Tanev are all undervalued by traditional points/TOI measures because of their defensive impacts, for example.
I love people disagreeing with me. That’s why I’m having a vote. Only about 1 in 20 visitors is registering a vote, but the contract nerds subscribing to my feed are highly educated in this field. If you want to make a case, you’re welcome to argue in the chat. The thread is open to free subscribers. https://substack.com/chat/1094421/post/6d11c868-3247-4db8-bf3f-67b295408710
I’m the first to admit that my modelling undervalues defensive players, but it also uses the variables most strongly correlated to actual salary. If there was a reliable defensive variable with a steady correlation to salary, I’d consider adding it. I came close to adding a hits modifier at certain production ranges, but there are also cheap hits available at a discount every summer. Just because there is correlation in some ranges, doesn’t mean spending that money is a good idea. Also, I had trouble getting to 60 bad contracts given the projected cap increases, some just raised red flags. Tanev it’s just age + term = worried. He was worth what they paid him this season for sure.
Right right. According to how contracts are modeled, these are over/undervalued deals. I suppose that's a major benefit of using statistical models that measure defense and underlying performance. They should hopefully provide a better measure of player impact than what front offices currently use. Given your modeling, these good and bad contracts are likely right in line with how NHL teams value guys.
My Expected Free Agent Value model is designed to predict what players will actually get on the open market. It hates the Slavin contract but I did not include it in the tournament. Every summer I predict what everyone’s contract will be (I’d love to have a friendly competition with anyone else who does the same) and the EFA model is my primary influence. If you can point me towards a defensive input that correlates strongly to salary, I’m open to suggestions. There are certainly players who get a premium for defensive impacts, but there’s also lots of cheap replacement level players every summer who play strong defensively.