BuzzerBeater Forums

BB Global (English) > little DMI research project

little DMI research project

Set priority
Show messages by
From: Cuzittt

This Post:
00
114729.3 in reply to 114729.1
Date: 10/7/2009 7:40:15 AM
Overall Posts Rated:
00
9, 6402, 69400, 6
9, 5133, 47300, 4
9, 9661, 78400, 2
9, 6814, 64100, 4
9, 5393, 60500, 6
8, 10265, 54500, 2
8, 4492, 29700, 5
8, 5823, 22700, 4
8, 6120, 29500, 3
8, 3443, 20300, 5
8, 9775, 65600, 3
8, 4734, 22700, 7

In the vain of more data is better... these are all of my guys who are at 8 or 9 Game Shape.

This Post:
00
114729.5 in reply to 114729.4
Date: 10/7/2009 8:13:58 AM
Overall Posts Rated:
129129
Sorry forgot that one...Stamina=3

I want what all men want...I just want it more.
This Post:
00
114729.6 in reply to 114729.1
Date: 10/7/2009 8:38:00 AM
Overall Posts Rated:
383383
(913991)
Gameshape 9
Stamina 8
Salary $ 36 390
DMI: 413600

I list him cause his stamina dropped from 9-8 last tu, dmi dropped by 1000. He didn't play on training position last week.

Perhaps this helps.

This Post:
00
114729.7 in reply to 114729.1
Date: 10/7/2009 10:39:27 AM
Overall Posts Rated:
225225
Please post the following (only!): game shape, visible salary, DMI, stamina, optional: player id as link
Imaginary example (I don't have any real ones on my team):
GS, Salary, DMI, Stamina, playerId

If the real relationship is DMI= (GS coeff)* SALARY + (some vector of other skill factors) simply looking at stamina might be insufficient. It is quite possible that skills enter the DMI calculation both as logs (in the salary formula) and as linear values, separately.

"I don't know half of you half as well as I should like; and I like less than half of you half as well as you deserve."
This Post:
00
114729.9 in reply to 114729.8
Date: 10/7/2009 11:27:31 AM
Overall Posts Rated:
225225
Possible, but I doubt that free throws impact DMI. The only other contender then seems to be experience, and I'm afraid we'll get too little diverse experience data to check this anyway, since only the small sample of observations that have a game shape with very low sublevel yield useful information for this research. But stamina does seem to play a role. Of course the DMI formula could be arbitrarily complex to tease us, but let's get there when a simpler approach has failed ;-)

I am not talking about free throws only -- any player skills can enter the DMI calculation as a linear combination. Again, I provided a simple example: if DMI contains the the salary (as predicted by your calculator, and based on current skills), which is a linear combination of skills, there is nothing stopping the "extra factor" from being a sum of any (or all) of the player skills times an appropriate coefficient.

Therefore, if you omit any skills just because they enter salary already, you might be biasing the results.

Last edited by GM-kozlodoev at 10/7/2009 11:28:07 AM

"I don't know half of you half as well as I should like; and I like less than half of you half as well as you deserve."
This Post:
00
114729.11 in reply to 114729.10
Date: 10/7/2009 11:35:41 AM
Overall Posts Rated:
225225
yes I understood you. but then the observation about DMI at proficient game shape would look odd. everything's possible though, especially with a deliberately meaningless index, but I'd like to stick with the simple approach for now. anyway what we need right now is not so much more theory but more data :-)

I think we have agreed that there is ample evidence that in proficient GS,

DMI = 10*SALARY + something,

and that you're looking for the something. How will it look odd if something = B*(sum of all/some player skills), where B is a coefficient?

I understand you need data, I just want to make sure you're collecting the right one.

Last edited by GM-kozlodoev at 10/7/2009 11:36:21 AM

"I don't know half of you half as well as I should like; and I like less than half of you half as well as you deserve."
This Post:
00
114729.12 in reply to 114729.1
Date: 10/7/2009 3:05:07 PM
Overall Posts Rated:
33
Thanks for the project. Here are some of my players. I want to know if it validates or refutes your theory.

GS, Salary, DMI, Stamina, playerID
9, 12772, 105100, 2. (2938541)
9, 4940, 44400, 3, (5710281)
9, 4648, 35700, 1, (9212182)

6, 1536, 2700, 2, (10598285)

Most of my players exceed your minimum thresholds, mediocre GS having DMI exceeding 14% instead of less than 10%. These examples look like they meet your criteria. The following is known to exceed the threshold.
7, 13794, 60700, 6, (4163456)
6, 3941, 8400, 6, (10589740)

This Post:
00
114729.13 in reply to 114729.10
Date: 10/7/2009 3:22:50 PM
JMDCeltics
II.3
Overall Posts Rated:
32683268
8, 5071$, 26100, 4
8, 3883$, 23800, 3
8, 2739$, 19000, 5
8, 3563$, 21200, 7
8, 6522$, 26100, 2
7, 2820$, 8900, 7
6, 3405$, 7900, 3
9, 4384$, 35800, 2
9, 4097$, 45900, 8
7, 3912$, 9900, 2
8, 6633$, 42000, 5
9, 5325$, 50200, 6
7, 6478$, 18700, 1
6, 3010$, 7300, 5
8, 6492$, 27100, 1

Thanks for your work

Advertisement