Giving Back and Net Rating

[bakahump]

I will admit that Basketball stats are a blind spot for me. And I may not be applying this correctly....But....

We talk alot about "PP getting hunted and is thus unplayable" (in the playoffs anyway). That infers to me that EVERYTIME down the court on Defense then PP gives up points. (obviously EVERYTIME is hyperbole but bear with me).

But I wonder is there Value in PP being just good enough to help his team score EVERYTIME offensively despite being hunted and giving up points on defense EVERYTIME.

Again hyperbole. But if PP on the floor gives up 1.26 points per possession (or whatever) but is part of an offense that scores 1.20 ppp.......is that really unplayable? for 10-20 mins?

In other words.......is there value in getting a @0 net rating for 10 mins a game? 20?

Obviously -1 Would be slightly worse.....but over say 15 mins a game not the end of the world. -2 worse still but still not the end of the world. Granted at some point it would be too much for the Rested Starter that PP facilitates to overcome.
Basically.....is there a Basketball equivalent to an innings eater. And how far below avg would they have to be to be a real detriment.

Especially for THIS Celtics team who has such a large positive net rating with the starters. JT,JB,KP etc etc can help cover alot of PPs warts while in there with him and helping to maintain a 0 Net rating or by being rested and scoring +5 or +10 over when they are back on the floor. Because the celtics are "so good at scoring" (1.2ppp in the example) they can ....."allow" 1.26 or 1.3 over 10 mins and it not be horrible.

Maybe PP is a bad example.
Maybe I am misapplying Net Rating.

But essentially "How badly would PP need to be hunted to make him the negative we all believe he would be". Can he "Eat innings" and still provide (Playoff) value.
Finally how does 3pters factor in? PP would obviously be a "better defender when defending the 3pter" then in the post....or on the drive. So Allowing Butler to score 2 point post ups "consistently" might be offset even more by the fact that PP might facilitate (by virtue of his own shot or because the team is a very good 3pt shooting team) More 3s to counter those consistent Jimmy Butler 2s.

Maybe I am way off.....but figured you guys would have some insight.

 

[TomRicardo]

Oh boy is this a loaded gun.

To start, basketball is more about trying to make the best of finite amount of minutes then trying to cover an vast amount of innings like baseball so no there is not a basketball equivalent to inning eater per se. Now to contradict myself a bit, you are always looking for bench players that can give you a solid amount of minutes. That said quality is rated higher than quantity more substantially than in baseball.

Also while position less basketball is all the rage, bench players fill roles within a team. Generally bench players are really good at a skill or two and you can hide their limitations with the on court players while other starters rest. Sam Hauser is more valuable to Boston than say Denver.

Pritchard is a fantastic bench guard in taking up usage. His shooting is down but at career levels or a bit below it he gives guys like Tatum, Holiday, and White a blow even when they are on the court. He also keeps the ball out of Brown's hands except to score which is huge. As long as you have two of Tatum, Brown, White, and Holiday on the court, you can survive Pritchard on the court defensively. Right now he is the 8th guy on the team and the most realistic point of improvement to your playoff rotation if you are the Celtics.

 

[bakahump]

@TomRicardo
Thanks for the response Tom. And I follow all that. But there seems to be a ongoing mantra that (someone like) PP is "unplayable" in the playoffs. That they will be "hunted mercilessly". And I have witnessed games where it was PP getting switched onto PP.....Backs him down.....makes post move, 2 pts.
But what I dont seem to hear much about is on the other end.

Granted PP is no great shakes. He is maybe Average offensively especially in THIS offense (for all the reasons you mention.....less Brown is more Brown....decent shooting......keeping White/Tatum/Brown etc fresh).

But if PP is -5 over 20 mins.....If I understand correctly.....thats basically 3 buckets. (and maybe 40-50 possesions.)
What I really dont hear much of is the "plus" side. What did PP (and Co) do on Offense? If he can trade (again my example) 1.2pts per possession for 1.5 Points (which seems like it would be DOING GREAT For the Opponent). Over 10 possesions (say 20secs per possesion??) thats only 3 pts/3 mins. So a minus 15 which seems terrible.

So maybe I/we need more data???
Avg possesion time?
Points per possession with PP on the court, for and against?
(is there a stat that measures this??)
Does this change Radically during playoffs?


I guess I am just trying to determine what is "Unplayable". and does a player of PPs caliber (7/8/9th man) fit that criteria. Or are we just "GAWWW Butler over PP AGAIN!!!!"

 

[bakahump]

Maybe as simple as MPs/NETRating?

20 mins -7 doesnt seem terrible for a backup.
2 mins -10 does.

 

[DavidTai]

What are his actual ratings in the playoffs? for all I know he's a bad match-up with Miami specifically?

 

[bakahump]

I guess thats the question. But its been pretty much taken for granted here and elsewhere that PP (or at least player 8,9,10 on your roster) are unplayable in the playoffs. Specifically if they are smaller guards. I have no doubt its partly true.....I mean the celts (and others) have spent years trying to be entirely switchable. I just wonder HOW much is it true.

Using +- He is +60 total over 39 playoff games. or +1.53 in 11 minutes per game
Some of those are huge games with 7 +10s or Better.
Some are stinkers with 7 games -7 or worse
and some are "net neutral" with 11 between -2 and +2

Miami specifically (because Butler) he is 14 games
6 Net Neutrals (between -3 and +4)
1 real stinker at -8 (4 mins) and a -6 (24 mins)
and 2 good to very good games (+39 in 23 mins and +7 in 13)
2 DNPs
Then a couple -4 (6 mins) and +4 (12 mins)

The "most recent PP" in 2022 we had 7 games where he was
-1
+39
+7
-6
-4
-1
-3
Averaging 15 mins in the series
In 2023 we had 7 games against Miami
-3 (11 mins)
DNP
+4 (12 mins)
0 (1 min) (a must win so you could understand the extreme bench shortening)
-8 (4 mins....bad game which we still won and controlled from the outset)
DNP
-1 in 3:45

Basically PP will have a stinker +- game about once per series, but be "AVG" to well above average every other game.

Is that good management by the coach? Probably in part.
And the logs certainly show against Atlanta for instance he doesnt play (less then 30 seconds in BOTH games).....So Trey probably has eaten his lunch in the past and he was being protected.

Golden State Game 2 for instance.
-7
But in 18 mins.
Not great but when compared to the rest of the team
JT -36 (34 mins)
JB -10 (27)
Al -15 (27)
Smart -12 (25)
TL -6 (14)
DW -17 (30)
(now I am certainly not saying Nesmith who was a +10 in 10 mins should have started lol......game context matters....)

But for that game I fail to see how our "#8" 2nd guard off the bench was "unplayable" while everyone else was considered fine.

So again my thought.....and it could totally be off base.....is that a -4 or so PP in 10-15 mins a game is a "valuable" (err Passable?) "minute eater" on this team even in the playoffs. And I think his being hunted is over played a bit. OR...At least the NET RESULTS of him being hunted.....is overblown a bit.

 

[TomRicardo]

Maybe as simple as MPs/NETRating?

20 mins -7 doesnt seem terrible for a backup.
2 mins -10 does.

Net Rating is really not useful to evaluate a player casually especially over small sample sizes. It is better to look at different lineups. If you are looking at Pritchard you would see that Pritchard is best with Tatum and his worst with Brown.

 

[TomRicardo]

But for that game I fail to see how our "#8" 2nd guard off the bench was "unplayable" while everyone else was considered fine.

So again my thought.....and it could totally be off base.....is that a -4 or so PP in 10-15 mins a game is a "valuable" (err Passable?) "minute eater" on this team even in the playoffs. And I think his being hunted is over played a bit. OR...At least the NET RESULTS of him being hunted.....is overblown a bit.

Yea this isn't a good use of net rating. it is much better to look at overall lineups or trends over season. This is like using batting average and RBIs over a playoff series to determine a line up.

 

[bakahump]

Thank you.

But isnt the issue with that is Regular season and Playoffs are "two different things" because of intensity?
I understand that you can see "Hey when Baka plays with tatum he is +2.....when he plays with mere mortals he is -1078" but trends....at least to common convention seem to go out the window No?

 

[PedroKsBambino]

The NBA is about matchups, especially in the playoff. And so the reality of net rating and other similar metrics is that (in particular for the playoffs) they are inextricably linked to three things:

1. Who the opponent has on the floor and the resulting matchups
2. Which teammates are on the floor
3. To a lesser degree than either of above, but still real: what role the player is asked to play at both ends

With PP, the concern is primarily about item 1: will they play a bigger/strong PG and play PP off the floor defensively, either on switches or posting/driving? That isn't as big a factor in regular season---less focus on matchup play; more minutes against bench/inferior players; more need to play bench guys. I personally think it's at least partially true of him, but we'll see.

A net rating is perhaps best thought of as an estimation of true talent level as a teammate---across the season, and different matchups/lineups/roles, what does the person tend to contribute? The problem with the playoff use is that the impact of those matchups/lineups/roles goes up so the relevance of the single "across all variables" number goes down.

The NBA playoffs are sort of like asking "how well does a player hit a fastball" vs a fastball pitcher rather than looking at simply his overall batting average. And so the single number is there, and a real approximation of something, but not necessarily the right answer to who should play.