Re: SimNation Sim Hangout Tonight taking question (Da_Czar & OG)
Very good points.
I think maybe the solution might be to ditch the concept of an "OVR" rating all-together.
For example, we know in the real world that stats like points per game tend be overvalued in free agent deals--perhaps less so now but still the case. Defense tends to be undervalued and underpaid, largely because it's harder to measure at this point in time. But you still have to construct a roster properly--if you only invest in defense, you may end up lacking shooting and totally crush the spacing of your offense. There are tons of non-linear effects here.
What I'd like to see is a handful of gestalt ratings that cover basic skills: playmaking, scoring, rebounding, and defense at a minimum: perhaps more nuanced would be outside scoring, inside scoring, playmaking, perimeter D, and interior D. Some of these skills may be more scarce at a given time than other skills, and certainly more scarce when combined and depending on player size. For an obvious example, Lebron's ability to defend multiple positions AND serve as a primary playmaker AND score everywhere is what makes him so indescribably valuable. Draymond Green's value comes from his combination of freakish defensive ability across multiple positions, solid playmaking, and league-average 3-point shooting. It isn't that he's an amazing shooter, it's that he's just good enough of a shooter that he won't kill your spacing as a big man and that he can take and make wide-open threes at a respectable clip. If he shot the ball like MKG, he wouldn't be nearly as valuable, though he'd still be worth something.
I'd love to see the AI have the ability to value the combination of these different ratings together, and have the AI behavior toward these players be driven by 1) relative scarcity and 2) team fit. For example, a team might favor a lower usage 3 and D wing player as a starter over a better playmaker (with a higher OVR on the old scale), recognizing that they already have a capable, high-usage playmaker on the floor who will dominate the ball, and what they need is someone to provide spacing and play defense. They would then bring that other playmaker off the bench to spark the reserve unit for when their star needs a breather. A team with limited shooting ability from its big men might be tempted to overpay for a stretch 4 in free-agency, particularly if there are limited stretch 4's available. And in my dream world, there would be levels of GM competency, such that some teams might have GMs who still insist on overpaying for something like points per game and fail to notice the value of a 3 and D wing at a discount.
Just some thoughts. It seems like the actual game engine is getting incredible work to become more life-like--it will be interesting to see how they are able to develop the AI in MyLeague and MyGM. I think they are taking some steps, e.g. with teams being able to be in rebuild or "win-now" mode. I'd like to see GMs with different value ratings for certain things, e.g. valuing defense vs. shooting vs. playmaking vs. all-around versatility. A rebuild could work differently depending on 1) the GM and 2) the team/market. One variation would be the Hinkie model of hoarding picks, tanking, keeping cap space clear (e.g. they will fill open roster spots for the league minimum, rather than overpaying for random veterans), and trying to pounce when the timing is right with a big FA signing or a blockbuster trade package, all the while hoping they draft a superstar with their slew of picks. Another would be how a big-market team like the Knicks or Lakers might try to rebuild, which would involve a bit of tanking, clearing a lot of cap space, and then going all-in for a free-agent splash. When it comes to drafting, some teams would be more likely to go for "best available" and others for "best fit," and this might be in part determined by whether they are rebuilding or not. Some teams may always go for "best available," it would just depend.
I don't think it would take much, though I could be wrong, to make the AI in myGM or myLeague seem incredibly nuanced--this would give the league itself a ton of life and realism.
Very good points.
I think maybe the solution might be to ditch the concept of an "OVR" rating all-together.
For example, we know in the real world that stats like points per game tend be overvalued in free agent deals--perhaps less so now but still the case. Defense tends to be undervalued and underpaid, largely because it's harder to measure at this point in time. But you still have to construct a roster properly--if you only invest in defense, you may end up lacking shooting and totally crush the spacing of your offense. There are tons of non-linear effects here.
What I'd like to see is a handful of gestalt ratings that cover basic skills: playmaking, scoring, rebounding, and defense at a minimum: perhaps more nuanced would be outside scoring, inside scoring, playmaking, perimeter D, and interior D. Some of these skills may be more scarce at a given time than other skills, and certainly more scarce when combined and depending on player size. For an obvious example, Lebron's ability to defend multiple positions AND serve as a primary playmaker AND score everywhere is what makes him so indescribably valuable. Draymond Green's value comes from his combination of freakish defensive ability across multiple positions, solid playmaking, and league-average 3-point shooting. It isn't that he's an amazing shooter, it's that he's just good enough of a shooter that he won't kill your spacing as a big man and that he can take and make wide-open threes at a respectable clip. If he shot the ball like MKG, he wouldn't be nearly as valuable, though he'd still be worth something.
I'd love to see the AI have the ability to value the combination of these different ratings together, and have the AI behavior toward these players be driven by 1) relative scarcity and 2) team fit. For example, a team might favor a lower usage 3 and D wing player as a starter over a better playmaker (with a higher OVR on the old scale), recognizing that they already have a capable, high-usage playmaker on the floor who will dominate the ball, and what they need is someone to provide spacing and play defense. They would then bring that other playmaker off the bench to spark the reserve unit for when their star needs a breather. A team with limited shooting ability from its big men might be tempted to overpay for a stretch 4 in free-agency, particularly if there are limited stretch 4's available. And in my dream world, there would be levels of GM competency, such that some teams might have GMs who still insist on overpaying for something like points per game and fail to notice the value of a 3 and D wing at a discount.
Just some thoughts. It seems like the actual game engine is getting incredible work to become more life-like--it will be interesting to see how they are able to develop the AI in MyLeague and MyGM. I think they are taking some steps, e.g. with teams being able to be in rebuild or "win-now" mode. I'd like to see GMs with different value ratings for certain things, e.g. valuing defense vs. shooting vs. playmaking vs. all-around versatility. A rebuild could work differently depending on 1) the GM and 2) the team/market. One variation would be the Hinkie model of hoarding picks, tanking, keeping cap space clear (e.g. they will fill open roster spots for the league minimum, rather than overpaying for random veterans), and trying to pounce when the timing is right with a big FA signing or a blockbuster trade package, all the while hoping they draft a superstar with their slew of picks. Another would be how a big-market team like the Knicks or Lakers might try to rebuild, which would involve a bit of tanking, clearing a lot of cap space, and then going all-in for a free-agent splash. When it comes to drafting, some teams would be more likely to go for "best available" and others for "best fit," and this might be in part determined by whether they are rebuilding or not. Some teams may always go for "best available," it would just depend.
I don't think it would take much, though I could be wrong, to make the AI in myGM or myLeague seem incredibly nuanced--this would give the league itself a ton of life and realism.
Comment