Quote:
Originally Posted by fidstar-poker
I didn't say the bench was better than the starters.
You just said they sucked top to bottom.
But numbers seem to say their bench is okay compared to other benches.
no, i never mentioned benches, it's just basketball works, much there's a lot of unexplained stuff in sports such as series effects are other things which have held up over the entirety of the games so we know those effects are real even if we don't know why they exist
a team getting blown out nearly always scores more buckets in garbage time than the other team, it doesn't matter if it's starters vs starters, starters vs subs, subs vs subs - it just always happens over a large enough sample.
so people who get garbage time minutes are always going to have impressive +/- numbers and starters on bad teams are usually going to look terrible
i never said they sucked top to bottom (but they do), i was refuting this, which heavily implied they were better when he was off the court but that's extremely unlikely and i was pointing out that all the starters had terrible +/- and they were so bad they often go blownout so the +/- of the bench would be highly inflated
Quote:
Originally Posted by fidstar-poker
Wizards are +13.4 pp 100 possessions when he's off the court.
for this reason +/- numbers are worthless when looked at in a vacuum, you would need to isolate to only close situations and then you'd have a more meaningful bit of data
no, the nature of blowouts is that the team getting blown out always gets more points
if you take the best team in the nba and have them face the wizards then the spread in the halves and quarters is going to be flat but will be progressively in the favor of the wizards the later you go in the game
take tonight's biggest expected blowout in ncaab -27 Oklahoma
if you just look at the 1st half the the spread is not 27/2= 13.5 but rather -16.5
as you can see looking at alt spreads for 1h, it doesn't even offer below 15
and mind you, this isn't even a blowout, they are simply accounting for the possibility that the final minutes will be meaningless and that the other team will makeup some points late in the game as a result
this is because it's just a known phenomenon in basketball that teams which are up heavily nearly always give points back later in the game
So when wizards play the knicks tonight with a spread of 7.5, if the wiz are able to keep it relatively close then guys like poole will stay in the game, if they get blown out then poole will likely sit most of the 4th quarter
and if you check live odds, when the wizards are down by 20 halfway through the 4th tonight, it won't be the implied math of "ny is better than them by about 2 points per quarter so spread should be 21" but rather "it's a blowout so wiz are going to score more and the live spread will be something like 17" because the wiz will be expected to score 3 more points than the knicks in those final minutes regardless of who is on the floor