MIP Award 2007-08

[benji]

Almost every single one I have seen have shown the players stayed the same. The changes came from changes in usage and role, not from the minutes increase. Hollinger did a study in his first book that showed players who recieved more and consistent minutes preformed better than those who received less and inconistent minutes.

There are obviously exceptions, such as a player who simply is not capable of logging more minutes (Mike Sweetney) or would foul out ten times (Danny Fortson).

Antawn Jamison is pulling down 9.8 rebounds per game, Bynum 9.7 and Reggie Evans 9.4. But Bynum is snagging 22% of rebounds, Evans 21%, Jamison just 14%.

When you look at [something]/game, you can change minutes to change the numerator. When you look at [something]/minute you can only change the [something] or minutes to change the rate.

[---]

Choosing MIP by per game stats is icky...it's how Randolph won.

For me, I go with per game. I think that projecting to do something (ie. per 40 stats, etc.) and actually doing it are very different.

I like looking at adjusted stats to see what guys are doing wel in limited minutes, but they have to earn those minutes and duplicate that success in real time for it to be worth much at all IMO.

[benji]

I think that projecting to do something (ie. per 40 stats, etc.) and actually doing it are very different.

But those stats aren't projecting anything. They're telling you, this is what a player did every fourty minutes. It could be per 36 or per 100. Because people screw this up and think you're saying "if he played 40 minutes, he would get" when it is merely an agreed upon standarization, it is actually just "per minute". It is easier for people to comprehend one saying "per 40 this guy is a 20/10 guy", over "per minute this guy is a .5/.2 guy".

For me, I go with per game.

But why would you do this when you know it's pretty much meaningless?

Why not use totals? Because you know, those are affected by how many games and minutes a player plays. But then, you use per game, which is affected by how many minutes a player plays.

I said this before, when you use a players, say points per game, minutes has a bigger factor on it than either part, I can change minutes and drastically alter the players points per game. If a player gets 30 points per game, playing 35 minutes, and I change him to 15 minutes, now he's getting 12.9 points per game, hardly as impressive. Using per minutes is simply leveling the playing field and getting rid of outside descriptors. (What about possessions? SHUT UP! We're just going this far for now!) People instinctively normalize anyway when they say Bynum is getting ten boards a game, in just twenty four minutes!

When you go to per minute stats, you're eliminating another layer of noise, as shown by the above rebounding comparison. Randolph continues to be the poster child of this.

Code: Select all

Randolph0102: 2.8 ppg 1.7 rpg | 19.2pts 11.6rebs PER: 16.9
Randolph0203: 8.4 ppg 4.5 ppg | 20.0pts 10.5rebs PER: 19.9
Randolph0304: 20.1ppg 10.5rpg | 21.2pts 11.1rebs PER: 19.6

Per minute he was pretty much unchanged, but his minutes changed from 5.8 to 16.9 to 37.9. He wasn't "improving" he was simply playing more minutes. If you ignore the fact his minutes more than doubled a 8.5/4.5 jump to 20/10.5 may look impressive, but if you normalize, his jump from 20/10.5 to 21/11 isn't anywhere near as impressive.

You say he has to "prove" it in "real" minutes, fine. But that doesn't mean he improved when the biggest change was minutes. It also means you might never play a stud because he hasn't "proven" it because you haven't played him major minutes to start. How can a Randolph or Michael Redd "prove" they are 20+ ppg scorers if they're stuck behind a "proven" player like Sheed or Allen? Are they worse players simply because someone established is in front of them? It also assumes a lot about minute allocation, and roster construction.

Our goal is to analyze the players as closely as we can, using per game stats is not doing that. Using per minute is far closer to stripping away the "noise" affecting the numbers.

[---]

I completely understadn what your saying, but what I mean is I can't really call a guy a great scorer because he averages 35 points per 40. I'd have to see him prove it, doing it when he's a big target for the defence, dealing with fatigue, etc.

And no, a player is not worse than another because the player in front is established. I think the players just have to have to prove they are that good (most of the time, then there are examples like JO), and the first step to that is getting the minutes in the first place. Then there's coaches like Scott Skiles and Larry Brown who would probably refuse minutes to LeBron as a rookie.

Per minute stats are great for looking for guys ready to "break out", but personally I can't hold a guy who averages X per 40 and X per game in the same light.

I think it's a really useful way of presenting stats though.

[benji]

he averages 35 points per 40. I'd have to see him prove it

He's proven it though. He has done it. He has averaged 35 points per 40 minutes.

We're stuck in a mentality that if a player does not play 36 minutes a game, he is not useful and his time was meaningless. 48 players last season logged over 36 minutes a game. That's it. That is only a third of all possible starters logged "starter" minutes. (Only six did 40 minutes.)

The goal is never maximizing a five man unit, it is maximizing a twelve man team. Per game stats tell you little useful information beyond how many minutes per game a player played.

[Carmo]

You still have to have players in the game that are going to play more minutes and those players are going to be the players that can have maximum impact playing longer minutes. Just because someone gets 10 points in 12 minutes doesn't mean they can sustain that...they might be completely stuffed after those 12 minutes. I don't think anyone is saying their minutes are useless, just that they can't sustain it for longer than that short time, whereas others can.

[benji]

I will go out on a limb and say 90-95% of players who are in the NBA can play 30+ minutes a game easily, and keep 95% of their production. They are, for the most part, world class athletes, not random joes off the street. It is their job to play basketball for "long" stretches of a time.

I don't know why you are required to have players who will play 36+ minutes a game every game. I would personally rather have a Manu/Brent Barry rotation splitting the time over Melo playing all 48 minutes, ten times out of ten.

[Axel]

Chris Kaman, Andris Biedrins, Kevin Martin

three names that come to mind

[benji]

Why Kevin Martin? Because he's taking more shots?

Code: Select all

Martin0506: 16.3pts 5.5rebs 2.0asts | 54.0% eFG, 60.4% TS, 8.0 reb%, 9.9 ToR, 11.8AsR, 15.6 UsG
Martin0607: 23.0pts 4.9rebs 2.5asts | 53.3% eFG, 61.4% TS, 7.1 reb%, 8.3 ToR, 10.6AsR, 20.8 UsG
Martin0708: 25.1pts 4.6rebs 2.3asts | 49.0% eFG, 59.7% TS, 7.0 reb%, 7.5 ToR, 9.1 AsR, 23.8 UsG

[Andy92_man]

I think that Bynum should win the MIP. He's playing very good. He doesn't play too many minutes, but he is putting up some pretty good numbers. I am looking forward to see if he can keep it up. Also Jordan Farmar is playing really well. But that is just my thoughts!

[1CenT]

without looking at stats,

a few guys come into my mind

Chris Kaman, if u play fantasy basketball, and u didn't get this guy, u must be kicking urself..

Last time i checked he was avging 18 points, 14 rebounds, 2 assists, and almost 3 blocks a game..

Dwight is doing exactly the same thing but hes shooting 59% instead of 49% so hes avging 23 and a half points..

But thats only stats...

I've seen kaman play, he just gets his numbers quietly, Dwight, man hes a huge impact... the whole team keys in on him yet he still scores like that, if only he shot 70% ft.... Dwight this year reminds me of a rookie Shaq..

Ronnie Brewer has been real solid, i didn't see him play last season...

[TheMC5]

I think that Bynum should win the MIP.

Of course you do.

[Raps13]

It's way too early to guess this. Like I mean, as well as some of these guys have played so far, if somebody like Bargnani for example's sake comes in and puts up 25 a game and 10 rebounds for the rest of the year there will be no question. Let's wait a while before we decide this award's winner

[Kbryant8]

Antoine Walker will take it this year

[Indy]

It's way too early to guess this. Like I mean, as well as some of these guys have played so far, if somebody like Bargnani for example's sake comes in and puts up 25 a game and 10 rebounds for the rest of the year there will be no question. Let's wait a while before we decide this award's winner

If Bargnani puts up 25 and 10 he should not win MIP. He was the number 1 overall pick in the draft last year.

[TheMC5]

It's way too early to guess this. Like I mean, as well as some of these guys have played so far, if somebody like Bargnani for example's sake comes in and puts up 25 a game and 10 rebounds for the rest of the year there will be no question. Let's wait a while before we decide this award's winner

If Bargnani puts up 25 and 10 he should not win MIP. He was the number 1 overall pick in the draft last year.

And that has what to do with his hypothetical improvement as a player, exactly? If you go from 11 & 5 to 25 & 10, you're in the discussion for MIP, regardless of draft position.

Note: I do not expect Bargnani to put up 25 & 10 this year. He's a good 2-3 years away from that.

[mvpshaq32]

Well, it looks like we can consider Udrih now.
He is certainly getting a lot more minutes in Sacremento and he is making the most out of it.
That career high 27pts against his former team...and the win.

[Mikki]

Well, it looks like we can consider Udrih now.
He is certainly getting a lot more minutes in Sacremento and he is making the most out of it.
That career high 27pts against his former team...and the win.

Yeah cause Bibby's out.

For now, it goes to Azu.

[benji]

What?

Code: Select all

Azubuike0607: 17.4 pts 5.6 rebs 1.7 asts | 52.3% eFG, 57.2% TS, 12.1ToR, 16.8 UsG, 7.7 Reb%, 13.4 PER
Azubuike0708: 16.9 pts 6.5 rebs 1.9 asts | 47.9% eFG, 51.5% TS, 7.6 ToR, 17.6 UsG, 8.8 Reb%, 13.4 PER

Bynum and Brewer are still the guys. And Felton.

Code: Select all

Bynum0607: 14.2 pts 10.8 rebs | 55.8% eFG, 59.3% TS, 15.4 ToR, 15.0 UsG, 15.8 Reb%, 15.4 PER
Bynum0708: 17.3 pts 15.8 rebs | 58.5% eFG, 62.0% TS, 12.2 ToR, 16.3 UsG, 22.1 Reb%, 21.0 PER

Brewer0607: 15.4 pts 4.4 rebs 1.4 asts 2.2 stls | 52.8% eFG, 56.6% TS, 8.6 ToR, 15.5 UsG, 6.7 Reb%, 15.2 PER
Brewer0708: 18.2 pts 3.6 rebs 3.2 asts 3.5 stls | 55.0% eFG, 63.0% TS, 8.2 ToR, 23.2 UsG, 5.3 Reb%, 20.9 PER

Felton0607: 15.4 pts 3.8 rebs 7.7 asts 1.7 stls | 43.4% eFG, 48.1% TS, 12.0 ToR, 21.8 UsG, 5.6 Reb%, 13.5 PER
Felton0708: 18.3 pts 3.9 rebs 8.3 asts 1.7 stls | 46.9% eFG, 55.1% TS, 12.6 ToR, 23.5 UsG, 5.6 Reb%, 18.7 PER

Hell, Howard.

Code: Select all

Howard0607: 19.1 pts 13.3 rebs 2.1 blks | 60.3% eFG, 61.9% TS, 19.3 ToR, 20.8 UsG, 20.5 Reb%, 21.1 PER
Howard0708: 24.9 pts 15.4 rebs 2.9 blks | 61.2% eFG, 63.7% TS, 13.5 ToR, 23.2 UsG, 21.2 Reb%, 27.4 PER

Even LeBron over Azubuike...

Code: Select all

LeBron0607: 26.7 pts 6.6 rebs 5.9 asts | 50.7% eFG, 55.2% TS, 9.4 ToR, 29.7 UsG, 9.6  Reb%, 24.5 PER
LeBron0708: 30.9 pts 7.6 rebs 8.2 asts | 52.0% eFG, 57.1% TS, 8.3 ToR, 32.8 UsG, 10.9 Reb%, 32.5 PER

[---]

What the hell, Raymond Felton getting love from Benji!?

Looks like my predicition on Feltons breakout is working out, hopefully he can keep it up.

[The X]

nice stat comparison Benji....what's Kaman's PER improvement like compared to Dwight Howard's?

[benji]

What the hell, Raymond Felton getting love from Benji!?

I have no problem admitting I'm wrong, especially since I am massively pessimistic in predictions. He had spent how many years looking like a poor man's Mike Bibby. How was I to know this was the year he'd be a standard-price Mike Bibby? (Well, what I'd pay for Mike Bibby. Not what the Kings paid for him after two good playoff series'.)

Kaman PERs: 13.1, 15.0, 12.9 last three years, 12.9 last year...19.4 this year. 14 and 11 last year to 20 and 15 this year. (Darko, I never thought I would be talking about Kaman as a 20/15 guy...)

[BP8]

Andrew Bynum!!!

[Raps13]

It's way too early to guess this. Like I mean, as well as some of these guys have played so far, if somebody like Bargnani for example's sake comes in and puts up 25 a game and 10 rebounds for the rest of the year there will be no question. Let's wait a while before we decide this award's winner

If Bargnani puts up 25 and 10 he should not win MIP. He was the number 1 overall pick in the draft last year.

And that has what to do with his hypothetical improvement as a player, exactly? If you go from 11 & 5 to 25 & 10, you're in the discussion for MIP, regardless of draft position.

Note: I do not expect Bargnani to put up 25 & 10 this year. He's a good 2-3 years away from that.

Neither do I, I was just using him as an example. One guy who probably won't be in the race for this but should is Carlos Delfino. He's gone from an IR guy to an integral part of the Raptors success thus far.

[benji]

but should is Carlos Delfino. He's gone from an IR guy to an integral part of the Raptors success thus far.

What? Injured Reserve guy? He played in every single game last season. And averaged 17 minutes per game during the regular season.