The Predictive Power Of Point Differential (Kings = -0.8)

[kingsboi]

To be clear, my point wasn't that I expected Philly and Dallas to bounce back and be better than the Kings THIS season but rather within a few seasons.

Teams tend to be horrible, get a top pick or two and then start getting good again. Teams that aren't the Kings anyway.

Not to make an excuse for their miserable drafting, but it would be nice if the Kings had the luxury of drafting in the top 3 or 5 for once...where their chances of messing up the pick diminishes a bit more IMO, even though nothing is ever given in the draft.

 

[funkykingston]

Not to make an excuse for their miserable drafting, but it would be nice if the Kings had the luxury of drafting in the top 3 or 5 for once...where their chances of messing up the pick diminishes a bit more IMO, even though nothing is ever given in the draft.

I don't feel like doing the research but it sure feels like for a team that has generally been so bad that the Sacramento Kings haven't had a lot of really high picks. Say top 5 or better.

I remember Pervis Ellison at #1

Billy Owens at #3 - traded for Richmond

Tyreke at #4 and Boogie at #5

Are there top 5 picks that I'm missing?

EDIT: Forgot about Thomas Robinson. I think that's all.

 

[kingsboi]

I don't feel like doing the research but it sure feels like for a team that has generally been so bad that the Sacramento Kings haven't had a lot of really high picks. Say top 5 or better.

I remember Pervis Ellison at #1

Billy Owens at #3 - traded for Richmond

Tyreke at #4 and Boogie at #5

Are there top 5 picks that I'm missing?

That about sums it up as far as I know. But in order to attain Boogie, the Kings had to have a top 5 pick and that was a consensus pick too so there was no way they could of messed that one up even if they want to. Hell, looking at it, he developed into more than I ever thought he could become in spite of all the turmoil which speaks volumes on his skill set and work ethic. That said, this team desperately needs a top 5 pick IMO...no way around it really because drafting 7-10 every year is not getting it done, no matter how many all-stars were passed up by this team.

 

[hrdboild]

To be clear, my point wasn't that I expected Philly and Dallas to bounce back and be better than the Kings THIS season but rather within a few seasons.

Teams tend to be horrible, get a top pick or two and then start getting good again. Teams that aren't the Kings anyway.

In that case, I fully agree

 

[New Era]

Well, we are now at -2.8 which looks to put us around 33-36 wins on the year, so I guess this thread/Bricks point is useless now.

 

[Spike]

Well, we are now at -2.8 which looks to put us around 33-36 wins on the year, so I guess this thread/Bricks point is useless now.

Let's see what the next two games bring.

 

[Bricklayer]

Well, we are now at -2.8 which looks to put us around 33-36 wins on the year, so I guess this thread/Bricks point is useless now.

-1.9 actually, with an asterisk, which still leaves us ahead of the current holder of the #8 seed

 

[funkykingston]

-1.9 actually, with an asterisk, which still leaves us ahead of the current holder of the #8 seed

We're at -2.1 just like Washington and Indiana - two other mediocre teams likely to miss the playoffs.

You don't get asterisks because your coach decided to rest a guy or due to injuries. Your stats are your stats.

 

[bajaden]

I don't feel like doing the research but it sure feels like for a team that has generally been so bad that the Sacramento Kings haven't had a lot of really high picks. Say top 5 or better.

I remember Pervis Ellison at #1

Billy Owens at #3 - traded for Richmond

Tyreke at #4 and Boogie at #5

Are there top 5 picks that I'm missing?

EDIT: Forgot about Thomas Robinson. I think that's all.

Ellison is the only number one pick we've had. One year we had the worse record in the league, and three teams jumped ahead of us and we ended up picking fourth. That was the Indian burial ground using all of it's power. Thank God were out of that building. Petrie didn't do bad considering. I think Peja was a seventh pick, and Kevin Martin was down around 21 or 22. Hedo, I don't remember, but my guess is around 11. Don't feel like looking it up. Even Casspi, who took a while, was a late first round pick. Douby aside, I think one of Petrie's worse picks without Maloof's interference was Donte Greene. I didn't like him at Syracuse and he proved out to be everything I thought he would be.

 

[funkykingston]

Ellison is the only number one pick we've had. One year we had the worse record in the league, and three teams jumped ahead of us and we ended up picking fourth. That was the Indian burial ground using all of it's power. Thank God were out of that building. Petrie didn't do bad considering. I think Peja was a seventh pick, and Kevin Martin was down around 21 or 22. Hedo, I don't remember, but my guess is around 11. Don't feel like looking it up. Even Casspi, who took a while, was a late first round pick. Douby aside, I think one of Petrie's worse picks without Maloof's interference was Donte Greene. I didn't like him at Syracuse and he proved out to be everything I thought he would be.

Peja was #14, one pick after Kobe (who Vlade was traded for)

I believe Hedo was 16th or 18th

 

[Bricklayer]

We're at -2.1 just like Washington and Indiana - two other mediocre teams likely to miss the playoffs.

You don't get asterisks because your coach decided to rest a guy or due to injuries. Your stats are your stats.

If we end up just sub-.500 and missing the playoffs by 2-3 games like an Indiana, that's a step forward for us.

And yeah, its bit of an asterisk because we chose to give up that differential to the Rockets. If +/- is predictive, then the game yesterday only applies if Cousins gets hurt. And he always has, so watch out. But then again, he may not, in which case the Rockets game is just a mulligan. Not the loss, but the margin. We would have almost surely lost anyway on the homecourt of a team that has won 7 in a row, but it was an artificial uncompetitive margin.

 

[whoopdeedoo]

If we end up just sub-.500 and missing the playoffs by 2-3 games like an Indiana, that's a step forward for us.

And yeah, its bit of an asterisk because we chose to give up that differential to the Rockets. If +/- is predictive, then the game yesterday only applies if Cousins gets hurt. And he always has, so watch out. But then again, he may not, in which case the Rockets game is just a mulligan. Not the loss, but the margin. We would have almost surely lost anyway on the homecourt of a team that has won 7 in a row, but it was an artificial uncompetitive margin.

Do you have a second account? If not maybe you should consider making one. It feels like it's gotten to the point where some are finding things to disagree about just because you wrote it. I don't see how someone can justify ignoring 24 games of data which was collected when your best player played every one of those games, and your second best player missed one game just because they BOTH miss one game at the same time which greatly influenced the data.

 

[peacelate]

Yesterday's game is not an asterisk. Instead of losing by 20 pts, the Kings decided to lose by 30 pts. They knew they had no chance of winning the game so decided to rest their star player.

 

[Bricklayer]

Yesterday's game is not an asterisk. Instead of losing by 20 pts, the Kings decided to lose by 30 pts. They knew they had no chance of winning the game so decided to rest their star player.

Therefore yesterday's margin IS an asterisk.

Not the result, but the margin, which is what this thread is about.

 

[kupman]

Maybe we should add a second asterisk for those games when were getting beat soundly for 44 minutes until our opponent let up on the gas pedal with about 4 minutes to go making the score look closer than the game actually was.

 

[funkykingston]

Do you have a second account? If not maybe you should consider making one. It feels like it's gotten to the point where some are finding things to disagree about just because you wrote it. I don't see how someone can justify ignoring 24 games of data which was collected when your best player played every one of those games, and your second best player missed one game just because they BOTH miss one game at the same time which greatly influenced the data.

It's not about ignoring 24 games of data, it's about NOT ignoring that 25th game.

The Kings chose to rest Cousins. I don't even disagree with that decision but you don't get to say that the stats don't matter for that game because you chose to rest him.

Up until this season my views have aligned with Brick's much more often than not but this season it seems like he's squinting very hard to try and see this team as better than it is.

And because we disagree on the value of improving just slightly. To me, missing the playoffs by 2 or 3 games instead of by 6-8 games isn't progress this season. Normally it might be but with Gay, Collison, Lawson, McLemore and Casspi set to be free agents and Afflalo (and perhaps Tolliver) almost certain to be cut loose as well, I don't see the value in an incremental improvement.

To me it just represents likely losing a first round pick in an offseason where they already have to replace (or re-sign those willing to return) 6-7 players, all of whom have been contributors this season.

Being slightly better isn't an accomplishment if they take a step back next season. It's a treadmill of mediocrity and why the Kings have been awful for so long while other teams bottom out and then get good again while the Kings try really hard to win 36 games instead of 33.

 

[whoopdeedoo]

It's not about ignoring 24 games of data, it's about NOT ignoring that 25th game.

So explain to me why a game where the best two players on the team are missing is a valid data point to include. Everyone knows without Cousins, and to a lesser extent Gay that the team is dreadful. For the purpose of this thread, to predict how the Kings season will go relies on at least Cousins playing the majority of games. If he doesn't play the entire premise, in relation to this years Kings, is worthless. Do you understand that and it's the reason the Houston game is inconsequential to the original post?

 

[macadocious]

So explain to me why a game where the best two players on the team are missing is a valid data point to include.

Because everyone has a game or two like that, so everyone has those games included as well.

 

[whoopdeedoo]

Because everyone has a game or two like that, so everyone has those games included as well.

Everyone in the league has a player they rely on as much as Boogie? Hardly.

 

[macadocious]

Everyone in the league has a player they rely on as much as Boogie? Hardly.

You're kind of making funky's point for him, aren't you?

If we can't even be moderately competitive without our best player, how good are we? Boogie is going to miss 5-20 games every year because of how he plays. So you better be able to withstand that.

 

[funkykingston]

So explain to me why a game where the best two players on the team are missing is a valid data point to include. Everyone knows without Cousins, and to a lesser extent Gay that the team is dreadful. For the purpose of this thread, to predict how the Kings season will go relies on at least Cousins playing the majority of games. If he doesn't play the entire premise, in relation to this years Kings, is worthless. Do you understand that and it's the reason the Houston game is inconsequential to the original post?

Because that's part of who the Kings are as a team. It's part of the overall stats. It's not like you'd look at the Warriors' point differential and say a blowout against the Mavs doesn't matter because Dirk is out and they are awful.

Point differential is a cumulative stat. Games where you rest guys count. Games where guys are injured count.

The Kings ARE awful without Cousins. But that's part of why they keep winning 27-33 games every year.

 

[peacelate]

The more one has to cherry pick stats, back-to-backs, road games, altitude, the more I think you are just putting lipstick on a pig.

 

[whoopdeedoo]

You're kind of making funky's point for him, aren't you?

If we can't even be moderately competitive without our best player, how good are we? Boogie is going to miss 5-20 games every year because of how he plays. So you better be able to withstand that.

No, I am not making his point for him. You are convoluting arguments. The entire thread is predicated on point differential that was calculated with our best, and most important player playing. When you take that out of the equation it's a completely different discussion. The discussion is then something like how good are the Kings without Cousins, I assume you can see how that is something much different than point differential as a predictive measurement.

 

[peacelate]

No, I am not making his point for him. You are convoluting arguments. The entire thread is predicated on point differential that was calculated with our best, and most important player playing. When you take that out of the equation it's a completely different discussion. The discussion is then something like how good are the Kings without Cousins, I assume you can see how that is something much different than point differential as a predictive measurement.

If you are going to do that, you have to asterisk the Kings win vs. the Mavs where they won by 31 in which Dirk (their best player) was missing.

 

[macadocious]

But that's not how point differentials for the season work. They aren't based on fully healthy rosters. So there is no reason to exclude games Cousins doesn't play. Do we asterix the Jazz since Favors and Hayward have missed a lot of time? Since his first season in the league, cousins hasn't played more than 74 games.

 

[whoopdeedoo]

If you are going to do that, you have to asterisk the Kings win vs. the Mavs where they won by 31 in which Dirk (their best player) was missing.

Go ahead, if you think the Dirk who said he might be retiring this summer has anywhere near the impact on his team as Cousins take out the data from that game. That's ok to do. The stats are just stats, we as analysts have to give them context.

 

[Bricklayer]

If you are going to do that, you have to asterisk the Kings win vs. the Mavs where they won by 31 in which Dirk (their best player) was missing.

anymore I think you want Dirk out there just so you can score more. Dude's a pylon.

Key point stands. Yesterday meant almost nothing so long as Cousins stays healthy. We CHOSE that margin of defeat. The only game all year long where we failed to compete at that level was the early Milwaulkee game. We're talking predictive value here, not aesthetics. If we sat Cuz the first 5 games of the year and we lost each one by 30, then he got healthy for the 6th game and you predicted another 30pt loss...

Let's put it this way, if we started putting money on games from here on out, and you placed your bets assuming we were a team that periodically gets beat down by 40 points, and I placed my bets on the assumption we were a team that plays teams tight, pretty soon I'd have all your money and we'd have to put you to work around here to pay your debts. Which might actually be a good thing, since we've got a lot of tasks piling up in the corners.

 

[Bricklayer]

But that's not how point differentials for the season work. They aren't based on fully healthy rosters. So there is no reason to exclude games Cousins doesn't play. Do we asterix the Jazz since Favors and Hayward have missed a lot of time? Since his first season in the league, cousins hasn't played more than 74 games.

You take that when it comes.

Although it should be noted that a more advanced way to look at +/- is in fact to take into account who is healthy and not. At least if you are trying to predict the results of any single game.

But since there is absolutely no reason that HAS to happen, you can't factor it in in a reasonable way. Over the course of a season Cousins could miss 0 games, 10 games, 20 games, all with radically different effects on +/-. But the only knowable fact is that thus far he has missed no games he had to miss. We gave the Rockets our queen before even starting the game last night. That renders it a fairly useless game for predicting anything until and unless Cousins goes down in the future.

 

[peacelate]

anymore I think you want Dirk out there just so you can score more. Dude's a pylon.

Key point stands. Yesterday meant almost nothing so long as Cousins stays healthy. We CHOSE that margin of defeat. The only game all year long where we failed to compete at that level was the early Milwaulkee game. We're talking predictive value here, not aesthetics. If we sat Cuz the first 5 games of the year and we lost each one by 30, then he got healthy for the 6th game and you predicted another 30pt loss...

Let's put it this way, if we started putting money on games from here on out, and you placed your bets assuming we were a team that periodically gets beat down by 40 points, and I placed my bets on the assumption we were a team that plays teams tight, pretty soon I'd have all your money and we'd have to put you to work around here to pay your debts. Which might actually be a good thing, since we've got a lot of tasks piling up in the corners.

I'd actually win because I would use 25 games worth of data to make my prediction, not exclude games that I know the Kings were going to get blown out in anyways.

 

[hrdboild]

Guys... this is the same point I was making the day after the Laker game. Blowouts are going to happen and that means day to day the point differential could change quite a bit, especially this early in the season. Whether we're in "good shape" or not for the playoffs then depends dramatically on where you take the sample from. After we win by 24 points the situation looks pretty good. We lose the next game by 34 and we're back to sucksville. It's just one level of abstraction removed from overreacting to every win and loss as a barometer for our playoff chances, which we do anyway. It really only matters over a long period of time and it doesn't change the fact that one stat and one stat alone is the sole criteria for making the playoffs: wins. We could lose all of our games by 1 point and you can tell yourself we were this close but realistically, you still need to get those Ws more often than you get the Ls to have a chance.

The only way this matters is where you stand on the 5 year plan. If you think we need to make the playoffs this season to advance one step down the road toward keeping DeMarcus than trying to prove that we will make it despite our record right now has implications on whether we seek out trades for our impending free agents or stay the course. I don't think we need to make the playoffs this season but I do think we need to get something in return for Rudy, Darren, Omri, and Ben if we can't re-sign them because we are so asset-deficient and disadvantaged in the free agent market that we can't afford to bleed away what little talent we do have for nothing. And unless you know that we will make the playoffs or at least come close to it, our odds of keeping all of them (save Restricted FA Ben -- who we probably don't want to re-sign regardless) go down dramatically.

So... playing the odds, we have a chance of making the playoffs if we can convert more of those close losses into wins. Simply maintaining our current performance will not be good enough. Overall you can't say that injuries have hurt us much so far but we're starting to enter a period where they could -- if Rudy is out for awhile we're missing our #2 scorer. Maybe other guys step up and we get better without him but regardless of the specific circumstances, a realist would expect us to suffer more games lost to injury in the final 57 games than we've experienced so far in the first 25. Then you also have the odds of DeMarcus missing games in the last month due to accumulation of technicals (looking pretty likely). I know there are fans here that hate to hear this and that's perfectly fine -- I can understand where you're coming from -- but I always factor in the 5 year plan and the odds of missing the playoffs right now (better than 50%) plus the odds of losing our pick to the Bulls if we stay the course (dicey) plus the impending free agency situation plus the strength of the draft this year and the possibility of adding the starting PG we desperately need with a top 10 pick all adds up to a clear mandate for me of what we should do right now and it involves a lot of trades.

We're not good enough right now-- the numbers show it. We haven't been good enough for 10 years. There are some signs of improvement for sure and you expect to get better down the line as your players get used to a new coach, but pragmatically it's unwise to stake your future on low odds. Or maybe that's just me. I'm not a gambling man. Someone else could tell you that you lose 100% of the poker hands you don't play -- which isn't wrong either. And they're probably more fun at parties than I am. They're probably also broke.