Invictus88 wrote:Q00 wrote:Uncle Mxy wrote:A 15 PER is "NBA normal". If the PER calculations make sense, then with an entire team of 15 PER people, we should have a record that is somewhere around 41-41. I haven't crunched the numbers fully and don't plan to, but it seems like we ought to have a slightly >15 PER average starting lineup (though unsure how Singler and Kid Can't Play) drag us down). We didn't have significant injury to our starting 5 (unless you count Josh Smith's brain damage). Yet we were clearly nowhere near 41-41. Whazzup??
I think you really have to measure the PER of the opposition against us. That's when you start getting some sense where the suckage sticks out -- Singler can't defend at SG, Moose can't defend at PF but is probably ok at C, Josh Smith is potentially ok defending at SF but can't play offense like an SF should. Of course, this is a team sport and PER doesn't necessarily capture that dynamic. But some stuff really stands out just from the individual PER numbers that passes "the eye test".
Its because you can't just add up any individual stat to determine the worth of a team. That's why the moneyball thing never works.
Commenting on both quotes above:
- Smith was actually fairly terrible when defending against SFs this year -- not 'potentially ok'. Our perimeter defense was plain awful and Smith was a part of it.
- There is a misconception that Smith would magically transform his season if he was 'officially' the 4. He played the 4 a lot last year and was terrible -- partly because he still played offense like he was a 3. That has not and will not change regardless of official position.
- The 'eye test' method of determining the worth of a team obviously didn't work as that is what has been used by Detroit Pistons' management for the last 10 years. The reason I can confidently say this is what they used is because they couldn't possibly have used analytics to build this team -- the numbers wouldn't have allowed it.
Now I'm not saying that we should swing completely to just using stats as the sole basis of personnel decisions; just that they should be weighed equally along with other factors. They are tools you can use in determining value like anything else.
People who are irrationally dismissing the value of stats are knowingly or unknowingly sticking their head in the sand; possibly because otherwise it makes their viewpoints completely illogical.
No GM builds teams based solely on the eye test and completely disregards stats. If you think the Pistons weren't aware of Monroe and Drummonds stats when they drafted them, or Smith and Jennings stats when they signed them, then you are being naive.
I personally love using real raw stats to help evaluate players, and then using the eye test to read between the lines of those stats and see what's right and wrong. However, using unproven formulas that attempt to read between those lines for you to evaluate players and build teams is fantasy stuff. Any GM doing that is clueless and I can't understand why any fan would want their GM doing that. To me that's the illogical viewpoint, and the teams like Houston, Memphis, and Toronto who are considered at the front of the analytical movement all got bounced in the 1st round. So until those teams start actually proving what they're doing works, I don't understand how people can continue to campaign for it. That is what's illogical.
If these guys were winning championships, anyone supporting analytics would have a great argument, but when they are going home in the 1st round, I'm sorry but it just can't be taken seriously at this point in time.
You say teams should weigh analytics and the eye test evenly, but why? Where's the proof that those advanced stat formulas that you believe so heavily in actually work, and aren't just some BS made up by writers at espn? There is no proof, so I don't get why fans are so adament about their teams using it. That is what's illogical to me.
