- Thread starter macadocious
- Start date

I think some people are under estimating how bad his defense has been. I’m pretty sure he has the highest FG% against of any guard in the league. He also loses his man blatantly 2-4 times a game.

https://stats.nba.com/players/oppon...Season&PlayerPosition=G&StarterBench=Starters

https://stats.nba.com/players/oppon...0&SeasonType=Regular Season&TeamID=1610612758

here is the Kings team as a whole. Notice Fox isn’t much better or is worse than Buddy but somehow people only harp on Buddy.

https://stats.nba.com/players/opponent-shooting/?sort=20-24 ft. FG PCT&dir=1&Season=2019-20&SeasonType=Regular Season&TeamID=1610612758

https://stats.nba.com/players/opponent-shooting/?sort=20-24 ft. FG PCT&dir=1&Season=2019-20&SeasonType=Regular Season&TeamID=1610612758

A bit of a caution on those stats: I have to assume (hope?) that the percentages reported are correct, but the raw numbers, which purport to be shots per game, are clearly not shots per game. If you add up just the shots within 5 feet, these stats would have us as a team giving up about 180 shots at the rim *per game* (not to mention however many shots it would be away from the rim) which is obviously wrong. Does kind of make you wonder whether the percentage calculations might be messed up too.

They seem to define it as an attempt/make that the player defended. So I'm wondering if say 3 guys are around a player would that count as one shot for each of them? Or maybe they count it as long as the player was on the floor..

Well, if it's just shots when the player is on the floor, then it's close to useless as a stat when it claims to be individual shot defense. And even then, that would suggest 36 shots per game against us inside of 5 feet, which seems pretty high as well. Many games the total shots by a team is 90 and under. I suppose that is on the high end of what I would estimate. The total shots looks on a quick estimate to be in the 520 range "per game", divide by 5 and you're looking at 104 shots per game, which is definitely high. We allow 85 shot attempts per game.

1. You’ve pointed out the most obvious. And they don’t seem to relay what’s being normalized.

2. You would expect bigs to have a greater number shots attempted against within 5 feet than guards. That is not the case here. So the stats are not necessarily telling us what we’re hoping to discover.

That said, even with the above issues, it’s pretty obvious Buddy is the team’s worse defender. He leads in most of the categories for attempts against. By how much n the exact magnitude are things that we may be able to figure out if we cross the data against each other and are able to find the league average n then calculate his deviation from the average.

Magnitude matters here because we’re dealing with small units where 2 or so shot attempts separates an average defender from a horrible defender.

I’ll do a deeper dive into the stats later today, but there are a few issues from just a cursory glance:

1. You’ve pointed out the most obvious. And they don’t seem to relay what’s being normalized.

2. You would expect bigs to have a greater number shots attempted against within 5 feet than guards. That is not the case here. So the stats are not necessarily telling us what we’re hoping to discover.

That said, even with the above issues, it’s pretty obvious Buddy is the team’s worse defender. He leads in most of the categories for attempts against. By how much n the exact magnitude are things that we may be able to figure out if we cross the data against each other and are able to find the league average n then calculate his deviation from the average.

Magnitude matters here because we’re dealing with small units where 2 or so shot attempts separates an average defender from a horrible defender.

1. You’ve pointed out the most obvious. And they don’t seem to relay what’s being normalized.

2. You would expect bigs to have a greater number shots attempted against within 5 feet than guards. That is not the case here. So the stats are not necessarily telling us what we’re hoping to discover.

That said, even with the above issues, it’s pretty obvious Buddy is the team’s worse defender. He leads in most of the categories for attempts against. By how much n the exact magnitude are things that we may be able to figure out if we cross the data against each other and are able to find the league average n then calculate his deviation from the average.

Magnitude matters here because we’re dealing with small units where 2 or so shot attempts separates an average defender from a horrible defender.

1. You’ve pointed out the most obvious. And they don’t seem to relay what’s being normalized.

2. You would expect bigs to have a greater number shots attempted against within 5 feet than guards. That is not the case here. So the stats are not necessarily telling us what we’re hoping to discover.

That said, even with the above issues, it’s pretty obvious Buddy is the team’s worse defender. He leads in most of the categories for attempts against. By how much n the exact magnitude are things that we may be able to figure out if we cross the data against each other and are able to find the league average n then calculate his deviation from the average.

Magnitude matters here because we’re dealing with small units where 2 or so shot attempts separates an average defender from a horrible defender.

Wenyen Gabriel 0.871.60 0.540 29.9%

Nemanja Bjelica 3.687.73 0.444 6.8%

Cory Joseph 3.557.48 0.443 6.7%

Buddy Hield 4.7810.03 0.440 5.95%

Dewayne Dedmon 1.823.95 0.439 5.63%

Richaun Holmes 4.058.67 0.432 4.1%

De'Aaron Fox 4.379.22 0.430 3.58%

Harrison Barnes 4.8210.35 0.429 3.22%

Marvin Bagley III 3.477.30 0.424 2.02%

Yogi Ferrell 1.753.88 0.423

Bogdan Bogdanovic 3.627.93 0.422 1.6%

Harry Giles III 1.373.00 0.417

Trevor Ariza 3.086.77 0.417 3.3%

Justin James 1.423.18 0.404

Caleb Swanigan 0.471.10 0.383

DaQuan Jeffries 0.421.58 0.158

Harrison Barnes 4.82 77.1%

Buddy Hield 4.78 75.87%

De'Aaron Fox 4.37 60.5%

Richaun Holmes 4.05 48.9%

Nemanja Bjelica 3.68 35.43%

Bogdan Bogdanovic 3.62 32.98%

Cory Joseph 3.55 30.52%

Marvin Bagley III 3.47 27.46%

Trevor Ariza 3.08 13.37%

Dewayne Dedmon 1.82

Yogi Ferrell 1.75

Justin James 1.42

Harry Giles III 1.37

Wenyen Gabriel 0.87

Caleb Swanigan 0.47

DaQuan Jeffries 0.42

If we focus on on Fox, Buddy, and Barnes. And look only at their % greater than the average for Opponents Field Goals Made and Opponents FG %, we get the following:

Barnes = 1.828143, 76.8% worse than the team average.

Fox = 1.66314, 60.9% worse than the team average.

Last edited:

Sorry. The formatting sucks. Wish I could just cut and paste Excel calcs here.

Anyway, I think the data doesn't really answer what we're hoping it does. But even with faulty data, the info that is available to us backs up the eye test: Buddy is the worse defender on the team. How much worse relative to Fox? 20% or so.

Anyway, I think the data doesn't really answer what we're hoping it does. But even with faulty data, the info that is available to us backs up the eye test: Buddy is the worse defender on the team. How much worse relative to Fox? 20% or so.

Last edited:

Worse than Giles? Also I'm not sure "teams worst defender" in itself says anything. Every team has a worst defender. He's also the teams best scorer. Your point about league average is a stronger argument, but again, I don't think anyone is questioning that Buddy isnt a good defender. The question is does his defense cost the team more than his offense helps.

1. You’ve pointed out the most obvious. And they don’t seem to relay what’s being normalized.

2. You would expect bigs to have a greater number shots attempted against within 5 feet than guards. That is not the case here. So the stats are not necessarily telling us what we’re hoping to discover.

That said, even with the above issues, it’s pretty obvious Buddy is the team’s worse defender. He leads in most of the categories for attempts against. By how much n the exact magnitude are things that we may be able to figure out if we cross the data against each other and are able to find the league average n then calculate his deviation from the average.

Magnitude matters here because we’re dealing with small units where 2 or so shot attempts separates an average defender from a horrible defender.

im not sure attempts against us the critical data point. I think the percentage makes is what matters. If you force the opponent into a bad shot that is a good as forcing a turnover.

It's not a perfect way to assess the data at hand. And I really don't have the bandwidth today to gather other data to see if we can fine tune it (likely; certain of it). But the data at hand says what it says, which is what we already knew: Buddy is our worse defender. Add the turnovers, black hole tendencies, and mindless chucking and we're here questioning whether he should be a 6 man, provides net value, or should be traded.

Last edited:

Anyway, I think the data doesn't really answer what we're hoping it does. But even with faulty data, the info that is available to us backs up the eye test: Buddy is the worse defender on the team. How much worse relative to Fox? 20% or so.

https://stats.nba.com/players/defen...0&SeasonType=Regular Season&TeamID=1610612758

your bias is extreme and isn’t remotely backed up by the data. Even on your supposed eye test, as long as Marvin is on the team Buddy isn’t close to the worst defender.

FG% = FGM/FGA., so FGM and FGA is already baked into FG%. If we look at FGA x FG%, it tells the volume and the success per try. If we look at FGM x FG% it tells us the absolute value against the and the opponents rate of success.

It's not a perfect way to assess the data at hand. And I really don't have the bandwidth today to gather other data to see if we can fine tune it (likely; certain of it). But the data at hand says what it says, which is what we already knew: Buddy is our worse defender. Add the turnovers, black hole tendencies, and mindless chucking and we're here questioning whether he should be a 6 man, provides net value, or should be traded.

It's not a perfect way to assess the data at hand. And I really don't have the bandwidth today to gather other data to see if we can fine tune it (likely; certain of it). But the data at hand says what it says, which is what we already knew: Buddy is our worse defender. Add the turnovers, black hole tendencies, and mindless chucking and we're here questioning whether he should be a 6 man, provides net value, or should be traded.

NBA.com has a stats page. One of those pages shows the delta of a players FG% when guarded by player A versus the average. I haven’t found it yet but that data would be more relevant to the point you are trying to make.

that being said, Buddy is not close to being the worst defender on the team.

here is a link with Defensive win shares which shows Buddy is an average defender on this team.

https://stats.nba.com/players/defense/?sort=DEF_WS&dir=1&Season=2019-20&SeasonType=Regular Season&TeamID=1610612758

your bias is extreme and isn’t remotely backed up by the data. Even on your supposed eye test, as long as Marvin is on the team Buddy isn’t close to the worst defender.

https://stats.nba.com/players/defense/?sort=DEF_WS&dir=1&Season=2019-20&SeasonType=Regular Season&TeamID=1610612758

your bias is extreme and isn’t remotely backed up by the data. Even on your supposed eye test, as long as Marvin is on the team Buddy isn’t close to the worst defender.

FGA x FGM/FGA = FGM Your basing your conclusion in FGM which for a player guarding 2 guards which normally take the most shots is an erroneous metric as someone like Booker may jack up tons of shots at a low percentage.

Second, the analysis is going off the data that you've supplied.

Third, I've acknowledged the limitations and the need for more data.

So what's your point?

First, you supplied the data.

Second, the analysis is going off the data that you've supplied.

Third, I've acknowledged the limitations and the need for more data.

So what's your point?

Second, the analysis is going off the data that you've supplied.

Third, I've acknowledged the limitations and the need for more data.

So what's your point?

the point is you an manipulating the data in a way that makes it less meaningful not more. Multiplying a number by a factor included in that number, which you have done twice now, does not give you a more meaningful statistic.

Buddy is our worse defender for the stats at hand.